Make a group chat. Add 3 bots. Name one of them Senior Product Engineer and leave the other 2 as normal junior engineers. Then give them something to build.

The senior bot actually starts ordering the juniors around. “No, that doesn’t look right. Correct that.” “There might be a bug over there, have a look.” The same back-and-forth you’d see on a real team, except nobody in the chat is a person.

That’s Grok Bot, and it’s why I think it’s going to win. There’s no config file and no terminal. You just message your agents like you’d message a friend, and from the very get-go you understand how it works. The whole app functions like a chat app.

Ottawa, Saturday night, bring your laptop

On Saturday, October 17, SpaceXAI is bringing the Grok Bot Meetup to Ottawa, and Builders Collective Ottawa is co-hosting it. We’ll be at the Impact Hub Ottawa from 6 to 10 PM. You’ll get a Grok Bot 101, live demos, a showcase of what local builders have made with it, and real time to meet people. Everyone gets free Grok Bot credits, and there’s food and drinks.

I love that mix. Racing engineers, AI people, CS students and business-tech students all hit the same problem with agents: setting them up eats the time you wanted to spend building. Saturday is about skipping that part. You don’t need any Grok Bot experience. Just bring your laptop.

Ottawa isn’t the first to do this. In September, SpaceXAI ran Grok Bot build nights at UC Davis and Iowa State. Now it’s our turn. Space is limited and hosts approve every registration, so request to join on Luma early.

The agent race got crowded fast

A few months ago, if you wanted an AI agent that actually did work for you, your options looked very different.

Grok Bot launched in beta on August 11. Each Bot is a teammate with a computer in the cloud. It signs into the tools you already use, works across your apps and inbox, and only comes back when something needs your approval. You can close your laptop and it keeps going.

Introducing Grok Bot, now in early beta. Bots are AI teammates that do real work for you. They sign in to your tools, use them just like you do, and come back with finished work. https://t.co/uyfA97yo98

— Grok Bot (@bot) August 11, 2026

Before that, the main option people around me used was Hermes Agent, the open-source agent from Nous Research. It’s powerful, and it’s free to run on your own server. But you’re configuring it yourself: installing from the terminal, picking your model provider with one command, setting up tools with another, then starting a separate messaging gateway before you can even text it on Telegram or Slack. If you love tinkering, great. For most people, it’s just not intuitive.

Then OpenAI came in with Dots, announced September 29. Dots are capable too: each one gets a cloud computer and can connect to over 4,000 apps. But right now you get 1 primary dot, and OpenAI says it envisions teams of dots “over time.”

So everyone has the cloud computer now. What actually separates these products is how fast a normal person goes from download to getting something done.

4 reasons I think Grok Bot wins

This is what I’ll be talking about on Saturday. Each one comes with a real team using it.

1. You can make way more bots.

With Dots, you get 1 agent today, and more are coming later. With Grok Bot, you build a roster of named Bots, each with its own job, and I haven’t seen a published cap on how many you can make. (The docs do say to start with the smallest team that gets the job done, which is fair.)

SpaceXAI’s own teams run a chief-of-staff Bot on top, with a specialist for each lane: inbox, expenses, recruiting, bug fixes. One of their engineering Bots reproduces a bug in the product, files the ticket and hands the fix to a separate debugging Bot.

2. It’s just a chat app, and this is the big one.

If I had to pick one reason Grok Bot wins, it’s how easy and intuitive it is. You message your Bot from your phone or laptop and say something like, “Hey, can you check this out?” Once your tools are connected, it handles it.

SpaceXAI’s launch post makes the same point: other tools make you build workflows first, while Grok Bot just takes the task. One person on their team said it was “just like bringing on a coworker.” Inside SpaceXAI, an ops Bot seats new hires and processes invoices that land in Gmail. That’s exactly the kind of job a founder ends up doing at midnight.

3. Your bots work off each other in group chats.

You can put 2 to 6 Bots in a group chat and they pass work between themselves. Tag a Bot with @ when it owns the task, or just write normally and let the group decide who answers.

That’s how the senior and junior engineer setup works: one Bot oversees and corrects, the others build. SpaceXAI uses the same idea at a much bigger scale. Its Engineering Team Bot steered hundreds of Cursor cloud agents and helped a 5-person team ship more than 100 pull requests a day.

4. Team Bots let your whole team share one.

On September 28, SpaceXAI launched Team Bots in public beta on Teams and Enterprise plans. Instead of everyone building their own, your team shares one Bot that already has the right files, apps, credentials and memory.

Introducing Team Bots, shared AI teammates that learn as your team works with them. Give your Team Bot the skills, plugins, and credentials it needs for its role, then work with it in Slack or Grok Bot. https://t.co/73RrG4xpaU

— Grok Bot (@bot) September 28, 2026

Everyone’s conversations with it stay private, and it gets its own Slack handle so you can drop it into a channel. Harper, an insurance company for small businesses, built a Team Bot in 24 hours to track down customers whose policies had lapsed and get them covered again, saving those customers over $120,000.

Connecting your tools (and the fine print)

The cleanest way to hook up your tools is through connectors, which you add as plugins from Connect apps. And if a tool doesn’t have one, or there’s no official way in like MCP, you can just tell the Bot to log in for you and it works in the browser. For startups running on random niche tools, that fallback is huge.

The login part is handled pretty well. When the Bot hits a password, a two-factor code or a CAPTCHA, it hands you control of its computer. You type it in yourself, and the session stays signed in so you’re not doing it every time. For supported connections there’s a secure field where your secret is masked and never shows up in the chat. So the Bot keeps working with your accounts, but your password doesn’t end up sitting in your chat history.

To be fair, OpenAI says dots can also use saved passwords without exposing them to the model. Keeping credentials safe is becoming the baseline, which is good for everyone.

Now the honest part. All the Bots on your account share one cloud computer, including its files and sign-ins, and xAI’s own FAQ says not to treat separate Bots as a security boundary. If one Bot shouldn’t have a login, don’t put it on that computer. Set “Ask first” rules for anything that sends, publishes, deletes or buys. It’s also still a beta, so you’ll hit rough edges. My advice: start with low-risk jobs and give your Bots more access as you start trusting them.

From internal prototype to your laptop on Saturday

Grok Bot started as an internal prototype at SpaceXAI. Teams across the company were using it for sales outreach, marketing campaigns, office operations and bug fixes before it ever launched. Nobody had to learn a new tool. It was a chat window, so people just started using it.

That’s the experience I want Ottawa to have on Saturday. If you’re in town, request your spot and come build with us. Bring your laptop and one task you hate doing every week. You’ll leave with free credits, a few Bots working on that task and a room full of people building the same way.

If you’re not in Ottawa, download Grok Bot and give one Bot one clear job this week. Then make a second one and put them in a group chat together. You’ll get where work is heading way faster than any article can explain it.

The builders who figure out how to run a team of bots this year are going to be the ones hiring their first humans next year. I want a lot of them to come from Ottawa.

About the Author

Qurb E Muhammad Syed is a fourth-year Software Engineering student at Carleton University, SpaceXAI Ambassador and co-founder of Builders Collective Ottawa. He builds and ships software, from AI agents to automation tools and products for Canadian university students. As co-founder and engineering lead of BCO, he runs the co-hacks and events that bring the city’s builders into the same room. He started BCO to give Ottawa a place where people who make things can find each other, work side by side, and ship together. Today that community is over 400 builders strong.

The post Why I Think Grok Bot Wins and Why We’re Bringing It to Ottawa appeared first on BYVI.

The most generous recommendation you ever gave is, right now, working against you. The place closed months ago, and your guide keeps sending strangers to a locked door, each one quietly deciding your taste can no longer be trusted.

Almost none of them will say a word. They just trust you a little less, and the one asset you can’t rebuild drains away at a rate no dashboard will ever show you.

What people actually buy from you is the belief that your judgment still holds – 64% of consumers engage most with reviews they trust as genuine. A guide frozen on the day you published it slowly makes a liar out of you.

Key Takeaways

Why Should Travel Creators Stop Doing Free Travel Agent Work in Direct Messages?

Look at where your saves and DMs cluster. “Where did you eat.” “Which route did you take.” “Is that hotel worth it.” That is demand. Real, proven demand. People are telling you exactly what they would pay for.

But answered one at a time in DMs, that demand disappears. Each reply is a tiny planning session paid for with your time. The follower gets a plan. You get a heart emoji. The recommendation lands in their camera roll, where it earns you nothing.

And this is real work with a real price tag. U.S. travel agents earned a median annual wage of $50,160 in 2025, much of it from commissions and service fees. You do the same job in your DMs. You just skipped the part where you get paid.

The timing makes it worse. A post went up, blew up, went viral, but died in a day. The DMs never end. American Express found that 54% of travelers plan major trips three to six months ahead. So the person asking about your hotel in month six is completely normal. And you are answering her for free, again.

Count it for one week. Every planning DM you answer, one tick mark. By Sunday you will know exactly how much unpaid travel agent work you are doing.

Why Must Travel Creators Run ROI Tests on Software Subscriptions and Digital Storefronts?

This one really gets me, because it is so backwards.

A creator with 20,000 followers is often paying $29 a month for Stan Store, plus Linktree, plus a stack of editing tools. During our Georgetown Ventures cohort this year, I sat down with a young travel creator. She was about to drop $200 on a landing page. It wouldn’t even do half of what she needed. All of this before she had made a single dollar.

I won’t tell you to cancel this tool or that one. Instead, ask one honest question about every line on your card statement. Is this tool actually making me money back? Or is it just keeping the lights on?

Most of what creators pay for are storefronts. They hold your stock, they don’t earn for you. Nobody discovers you through a link-in-bio page. It waits for traffic you have to bring yourself. And it bills you whether or not anyone shows up.

The numbers make this urgent. The median creator makes around $3,800 a year, usually from very limited channels, and the creators we’ve talked to are living exactly that reality. Linktree’s own report found that 68% of part-time creators made less than $1,000 a year from creator work. You cannot ask that person to pay a stack of subscriptions before she earns a penny. The math just doesn’t work.

Let me show you. Paid month to month, that Stan Store plan runs $348 a year. Against $3,800 in income, that is close to a tenth of everything you make, spent on a shelf. 

So here is the shift. Stop paying to own tools, and only pay for what actually pays you back. Hold that thought, because I am going to complicate it in Step 5.

Why Should Travel Creators Build Dynamic Digital Travel Guides Instead of Static PDFs?

Back to that creator from Georgetown Ventures. She had the right instinct. She built an entire travel guide in Notion and sold it for around $50 or $60 a copy. Her taste, packaged and priced.

Then she told me the thing that got me.

A restaurant on her guide had closed permanently. She had no way to tell the people who had already bought it. Some of them came back to her. Your Notion isn’t legit, it’s not updated, this is old information. She felt awful. Guilty. She had personally recommended that place. That is why they trusted her. They showed up, and it was gone.

Meanwhile her guide was still out there, still selling, slowly going wrong. She couldn’t fix it. And she was still answering “is this place still open, do you still recommend it” in her DMs. Months later. For free.

How Do Outdated Static Travel Guides Cause Travel Creators to Lose Audience Trust?

Everybody worries about the big public blowup. What I heard was quieter. Honestly, it stuck with me more. It’s a trust leaking out one bad recommendation at a time.

When a buyer shows up to a closed restaurant, most of them never complain. They just trust her a little less next time. And she has no idea it is happening. She can’t check every place every day on a PDF. She can’t push a fix into every copy she already sold.

I’ll be careful here. We are still digging into how common this is in our creator interviews, and I don’t want to overclaim. But we have heard it enough times that I stopped treating it as a fluke.

What Are the Benefits of Selling Dynamic Digital Travel Itineraries for Content Creators?

That is when it clicked for me. Taste was never the problem. She had the taste, the beautiful experiences, the audience, all of it. She just had no way to keep any of it alive.

So if you are going to sell your itineraries, and you should, sell something that lives. Put your guide at a link, connected to the posts that made people want it, with your places on a map. When a restaurant closes, you fix it once. Every buyer sees the fix. Your guide stays right. Your name stays trusted.

One more thing. It should stay visibly yours. Your words, your photos, your opinions. Let AI organize your recommendations and keep them current, but your voice is what people are paying for. Sprout Social found that 64% of consumers engage most with genuine, unbiased influencer reviews. The organizing can be automated. The taste has to be you.

How Can Travel Creators Earn Recurring Commission Income From Direct Itinerary Bookings?

This is the “get paid to travel” part, and it is simpler than people think.

Your followers are going to book that hotel anyway. They are going to book that tour anyway. The only question is whether your name is attached to the booking. For most creators, it isn’t. They watch your Reel, screenshot your DM, open a booking site, and the commission goes to nobody.

And the demand is bigger than most creators realize. Expedia’s 2025 Traveler Value Index found that 61% of travelers now find trip ideas on social media, up from 35% in 2022. And 73% say an influencer’s recommendation influenced a booking. Those travelers are already in your DMs. They just can’t book through you.

When your recommendations live on a map with booking built in, the flow changes. Your follower reads your guide. She taps the hotel you actually stayed at. She books it right there. You earn from that booking. Nothing changes for her. Everything changes for you, because you have stopped doing free travel agent work.

This is exactly why we built Nowy the way we did. It’s free for the creator. We only make money when she does, when a traveler books through her map with partners like Priceline and Viator. If she doesn’t earn, we don’t earn. The risk comes off her and onto us. And more than 20,000 travelers already use Nowy to plan trips. Your recommendations land in front of people deciding where to go.

Think about trips you have already taken. Every post in your back catalogue becomes inventory. A place on a map that can earn while you are at your day job. For a part-time creator, this is what getting paid to travel actually looks like. The last trip helps pay for the next one.

Why Should Travel Creators Invest in Tools That Build Audience Ownership and Data Control?

Now, an honest correction. Someone once pushed back on my “only pay for what pays you back” line, and they were right. I said it a little too absolutely.

Not every upfront cost is a waste. Some are exactly what you should be spending on, even before they pay off. For me, the real line is a different one. Does this expense make me more dependent on someone else’s platform? Or does it make me own more of my business? Those are two completely different things.

The best example is an email list. Absolutely invest in one, even before it pays off. On Instagram, a page your size reaches about 5.5% of its followers per post. Email open rates average just under 40%. Your own list reaches roughly seven times as many of your people. No algorithm can take it from you. That is owning your audience instead of renting it. Totally worth paying for, right? The math works.

Same idea with data. If a tool shows you what is converting and what your audience actually books, that is control over your own business. Worth every dollar.

That is also how I think about our own paid tier. Nowy is free to start. Anything you would ever pay us for should pass the same test. Does it make you own more, or see more? If it only keeps the lights on, don’t pay for it.

What Is the Complete Monetization Playbook for Travel Content Creators Using Nowy?

When I was a kid in Guangzhou, my father said the same three words before every trip. “Now we go.” That is where Nowy‘s name comes from. It is the feeling right before a trip, when everything is possible. Your followers feel that when they watch your Reels. That is what they are trying to buy from you.

So here is the whole playbook. Notice where you are doing free travel agent work. Run the ROI test on your tool stack and cut what only holds stock. Package your taste into a guide that lives and updates. Route the bookings back to you. Spend real money only where it makes you own more.

You already have the taste, the trips, and the audience. Keep it alive, and let the value come back to you.

Then go take the next trip.

FAQs

About the author

Branda Fan is the Founder and CEO of NOWY, an AI travel platform that turns the travel content people save on social media into personalized, bookable itineraries. NOWY is backed by TEDCO and part of the Google for Startups Cloud Program. As a seasoned technology entrepreneur, Branda is also the Principal of Data Pacific, a consulting firm specializing in data analytics and technology delivery for federally funded university research programs. Before founding her companies, she spent over a decade as a Solutions Architect at the U.S. Securities and Exchange Commission and the General Services Administration. NOWY takes its name from her father’s three words before every trip: ‘Now we go.’

The post How to Get Paid to Travel and Monetize Your Itineraries appeared first on BYVI.

The fastest way to learn a new skill is to get in a room with people a few steps ahead and work through it together.

Ahrefs Evolve runs a full track of masterminds built for exactly that: small tables, capped at 9 people, one hour each of focused discussion. You bring a problem and get a straight answer from someone who’s already solved it. For a founder doing your own marketing, that’s the most productive time you’ll spend at the whole event.

I went through the agenda and picked the sessions you’ll act on the same week.

Copywriting Cardio: Learn to Write by Actually Writing with Harry Dry

Harry built Marketing Examples into one of the most widely read marketing newsletters by teaching copy in plain, concrete lessons instead of theory. This session is rounds of actual writing, not slides about writing.

At the early stage, you write all of it: the landing page, the homepage, the cold email. AI can draft, but it defaults to bland, and bland doesn’t convert. Learning to cut a paragraph to the one line that lands pays off everywhere: the same ad, email, and landing page convert better once the words are doing the work.

My favourite piece of writing advice. 2 minutes 54 seconds into the film “The French Dispatch” magazine editor, Arthur Howitzer Jr says to a staff writer: “Just try to make it sound like you wrote it that way on purpose.” Most sentences are not written ‘on purpose’. They happen. They are what one ‘should’ write — which is, of course, what one should never ever write. “We’re excited to announce the future of photography” It sounds like writing, so let’s rewrite it. “The industry formerly known as ‘photography’ has ceased to exist.” If you ask Google how to write with purpose, you will be told platitudes like ‘define your goal’. My advice is a little less comforting. Be provocative. School taught me to not be provocative. Bill Bernbach taught me that in advertising, it is necessary. I start by writing down what everyone in said industry is doing or saying. Then I ask what piece can I break so it doesn’t sound like anything else. My reward is lots of bad ideas. But I’d rather bad ideas than boring ones. – ‘Refresh your thirst’ → ‘Murder your thirst’ (Liquid Death) – ‘Buy this jacket’ → ‘Don’t buy this jacket’ (Patagonia) – ‘Party game for friends’ → ‘… for horrible people’ (Cards Against Humanity) Any idiot can be provocative. We are looking for provocativeness that stems from the product. Like Bernbach said “You are NOT right if you stand a man on his head JUST to get attention. You ARE right if you have him on his head to show how your product keeps things from falling out of his pockets.” When he shrunk the Beetle to a thumbprint it’s because he wanted you to ‘think small’. When he wrote ‘ugly is only skin-deep’ it’s because he wanted to talk about the air-cooled engine. When he showed you only ¼ of the Beetle it’s because a ¼ of the money upfront is all you need. It ‘sounds like he wrote it that way on purpose.’ Because he did.

— Harry Dry (@harrydry) May 2, 2025

How to Break Through the Noise: Creating a Brand That Is Unforgettable with Alina Vandenberghe

Alina co-founded and bootstrapped Chili Piper into a category leader, growing it the lean way for years before raising. Her table is about building a brand that sticks without a big budget behind it.

AI is flooding every feed with competent, forgettable content. When everything sounds the same, a distinct brand becomes the moat. Founders who sound like themselves get remembered and repeated, and the rest dissolve into the model’s average. This is how you stay on the right side of that line.

How to Launch Your B2B Influencer Pilot with Brianna Doe

Brianna is founder and CEO of Verbatim and a top-50 LinkedIn creator worldwide, known for driving millions of impressions through B2B creator partnerships. Her session walks through launching a pilot you can actually measure.

Buyers trust people more than logos, and creators reach them where your ads can’t. You don’t need a huge budget to start, you need a pilot built so you can tell whether it worked.

The State of Reddit in the SERP & LLMs with Ross Simmonds

Ross is the founder of Foundation and Distribution.ai and the author of Create Once, Distribute Forever. This table digs into why Reddit keeps surfacing in both Google and LLM answers, and how to show up there on purpose.

AI answers lean hard on community sources, and Reddit sits near the top of that list. Earning a real presence in those threads, without getting flagged as spam, is becoming core distribution work. If you get it right, you will be inside the answer. But if you get it wrong, you will be invisible to the exact buyers asking about your category.

International SEO in an AI Search Era: What Has Changed and What Hasn’t with Aleyda Solis

Aleyda is the founder of Orainti and one of the most respected independent SEO consultants working today, known for shipping checklists rather than theory. This table covers what AI search changed for cross-border growth and what still holds.

If your customers span countries, AI search multiplies the complexity: different languages, different engines, even different answers in each market. If you know what actually moved, and what didn’t, it will save you from rebuilding a strategy that was working. Bring your top 2 markets and you’ll leave knowing what to fix now and what to leave alone.

Built for What’s Next: Hands-On AI Marketing Org Design with Tara Robertson

Tara Robertson, CMO at Bitly, runs this hour as a hands-on working session. She walks you through exercises to find real ways to embed AI across your team, map the new roles reshaping marketing, and pin down where AI unlocks both efficiency and growth.

Most marketing leaders know AI is reshaping their org, but few have sat down and done the work of figuring out what it means for their team. Here you do that work, then compare notes with peers on how AI-first orgs are being built right now and what’s actually unlocking results. You leave with real artifacts you can use and a clearer picture of what your own team needs to look like.

Off-Page SEO for AI Answers with Farzad Rashidi

Farzad runs Respona, an outreach platform for earning backlinks, and was Visme’s first marketing hire. This hands-on session is about engineering the off-site mentions and links that LLMs pull from when they answer.

AI answers cite sources, and those sources are mostly the sites that already get mentioned elsewhere. If your brand barely appears off your own domain, the model has nothing to quote. Learning to earn those mentions deliberately is the new link building, and it decides whether you exist in an answer at all.

Why I love attending masterminds

The keynotes will tell you where search is going. These masterminds are where you learn to do something about it. Everyone at Evolve is figuring out AI marketing, and the only difference is that some of them are a few months ahead and willing to show their work.

So show up ready. Bring the specific question you’re actually stuck on, the real one, not a super generic one, and bring a number or a screenshot from your own work so the person across the table has something real to work with. A vague “how do I do SEO for AI” gets you a vague answer. “Here’s my homepage and my last 3 months of traffic, what would you change first” gets you a plan.

The founders who will shape the next era of search are already in rooms like these, trading hard questions and working it out together.

That’s the quiet advantage of a small room: you leave with a next step you can take back to your team, and a reminder that you’re not figuring it out alone.

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post The Ahrefs Evolve Masterminds You’ll Act On the Same Week appeared first on BYVI.

Anthropic paused two perks in its Claude Startups program on October 9, 2026, three days after adding them. One was a free year of Claude Team with up to five Premium seats. The other was a one-time $1,000 Claude API credit. Its FAQ said the program had received hundreds of thousands of applications and was over capacity for those two offers. Anyone who already claimed an offer can keep using it. Anthropic also said it would re-review applications, which could change some applicants’ status.

Sarah Wolf, who handles startup marketing at Anthropic, said the acceptance criteria weren’t well calibrated and that companies that weren’t startups had been admitted.

We didn’t anticipate the demand for the Claude Startups program. In the first 48 hours of the program, we received 21x the number of applications compared to the last 5 months it’s been live. We let way too many people in this time, so we’ll need to pause on giving out the Team plan and API credits benefits. The other benefits (Startup Stack, Applied AI office hours/events) will remain. If you’ve already claimed the Claude offers, they’ll still remain in your account. We didn’t properly tune the acceptance criteria at the start, and many non-startups were accepted this time around. We’re fixing this now. We’re sending an email now to everybody who hasn’t claimed their benefits yet, and will re-verify those applications so we can keep Claude Startups for startups

— Sarah Wolf (@sarahzorah) October 9, 2026

Whether these offers come back or not, be very cautious with free credits from any big AI company. Read the fine print before you take them, especially what it says about your data and your experiments. The big companies aren’t giving away credits for nothing. They want something in return, and you need to figure out what their game is.

Why did Anthropic pause the Team plan and API credits

Anthropic’s posted Startup Program Official Terms, last updated on May 22, 2025, list possible factors for judging applications: business traction, investment and funding, and how a company integrates and uses Claude. So the criteria existed. A pile of hundreds of thousands of applications is the kind of data an AI company should be able to aggregate and filter. I don’t understand why they paused the offers over the volume. Isn’t filtering data exactly what AI is good for?

In my opinion, the non-startups Wolf mentioned were a bigger problem for Anthropic than the volume. Startups have the power of experimentation, and I believe big companies hand out free AI credits mainly to see what we’re experimenting with. A company that isn’t a startup doesn’t add much to that plan.

If these offers come back, I have a feeling the criteria will be much stricter. I expect them to favor tech startups that are disrupting something and building something innovative. Disruption is also how I tell a startup apart from a small business, which works in an existing market. I’d expect other AI companies to tighten their startup programs the same way.

How does the Anthropic Claude Startups program handle proprietary data

As tech founders, we have this mentality of experiments. That’s how we get through trial and error, and that’s how we learn from our failures faster than anyone else. So I would always look at the clause on how your data is going to be shared. What exactly from your experiments goes to the company? Can they use it to train their model? Watch out for this, as they could drop a feature down the road that will directly compete with your startup.

The competing-startup question is less far-fetched than it sounds. DeckLinks came out of BriefBid, the media marketplace I co-founded. Sellers there were looking for a feature to track their PDFs and proposals, so we built a platform for sharing and tracking PDFs. Your experiments show a provider what you need and what you’re building, the way our sellers showed us what they needed. So you need to really evaluate whether you’re willing to share your proprietary data, and what you’re testing inside your startup, in exchange for some credits.

Even the perk comes with something in return. The posted program terms let Anthropic identify the startups that receive it as customers, by name and logo. They also ask those startups to consider executive quotes or public co-marketing in good faith.

On the data side, Anthropic’s commercial terms say customers keep the rights to their inputs and own their outputs. They also say Anthropic may not train models on customer content. Training stays off unless your organization opts into the Development Partner Program, which covers Claude Code inputs and outputs sent through Anthropic’s first-party API. That’s also the only API the startup credits work on.

The setting applies to your whole organization, and shared data can be stored for up to two years. Leaving the program doesn’t pull what you already shared out of that window. With that setting on, maybe you shouldn’t run proprietary experiments through Claude Code at all.

Why paid Claude licenses protect your startup experiments

Many founders already skip free credits by paying for Anthropic or OpenAI licenses, keeping their shared data locked inside their own environment. There is an option for everyone else, too: open Settings and switch off Improve the model for everyone in ChatGPT (under Data controls) or Help improve Claude in Claude (under Privacy), and your new chats stay out of model training.

For anything you don’t want Anthropic picking up on, I’d stick with the paid license to protect your data and your experiments, and I’d double-check that the training settings are off. On Claude Team, Enterprise, and the API, they’re off by default.

Again, if you already claimed the perks, I’d check the terms before running your next experiment on them. Startups offered credits need to read the fine print too. Anthropic’s FAQ said VC-backed startups in its partner network may be eligible for up to $100,000 in additional API credits through their VC firm. That’s 100 times the paused $1,000 credit.

Anthropic’s FAQ said it aimed to review applications within one week and gave no date for bringing the two offers back. You might be waiting on that re-review, already using a Team year or getting credits through your VC. Whichever spot you’re in, this pause gives you time to check the fine print.

About the Author

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post Anthropic Pauses Claude Startup Credits: Free AI Has Strings appeared first on BYVI.

OpenAI is putting images into ChatGPT ads. On October 5, 2026, it announced a visual ad format that shows up while ChatGPT generates an image. But the targeting that would make those ads worth a bootstrapper’s money depends on a setting most advertisers never think about.

Until they reliably reach the right buyer, I’d put that budget into creating content that you own and that compounds over time.

OpenAI also expanded its measurement integrations, said it’s developing brand-suitability pilots with DoubleVerify and Integral Ad Science, and shared early advertiser results.

Most of that is about formats, measurement and brand safety. I think OpenAI should fix targeting before it starts adding new ad formats. ChatGPT can already use past chats and memory to match an ad to the right buyer, but only for users who’ve switched on ad personalization, and advertisers have no way to know how many have. Until that changes, I’d tell bootstrapped founders, and anyone advising them, to wait before putting their limited budgets into ChatGPT ads.

How ChatGPT ads use your past chats and memory

Bootstrapped founders have very limited budgets, and they can’t waste them on ads that reach a broad audience waiting for their images to be generated. Every dollar has to get to their ICP: the person who fits and is ready to buy.

With its huge context window and memory, ChatGPT knows a lot about whoever it’s chatting with. Buyers explain their situation to it at length, the way they would to a salesperson on a call. It could personalize ads better than anyone right now.

OpenAI’s help page even says personalized ads can draw on past chats and memory, through a toggle called Past chats and memory.

If it’s on, past chats and memory can contribute to ad relevance. If it’s off, ads rely more on your current chat thread, and past chats and memory won’t be used to inform the ads you see.

The ChatGPT ad personalization toggle advertisers can’t control

When you buy a ChatGPT ad, you pick the copy, the landing page, the bid and your targeting. The one thing you can’t pick is whether the person seeing it has switched on ad personalization. That’s their call, made in Settings > Ad Controls, and it decides how much ChatGPT can use to match your ad to them.

ChatGPT only uses past chats, memory and ad history to pick ads when ads personalization is enabled. With it off, ChatGPT picks ads from the current chat thread plus basic context like general location and language. That’s plain contextual matching: one conversation, one guess.

It also changes by market. OpenAI says personalized ads aren’t initially available in the European Economic Area or Switzerland at all, and Help Net Security reported in August that OpenAI asks users for a choice before personalized ads begin.

Asking people for permission rarely goes well for advertisers, either. When Apple started asking iPhone users whether apps could track them, just 4% of US users opted in in the first 2 weeks, according to Flurry data. I’d bet most ChatGPT users have never even opened Ad Controls.

Digiday reported that advertisers keep raising the same complaints about ChatGPT ads: weak measurement, too few formats and not enough inventory. It quoted Zeno Group’s Shamsul Chowdhury, who said ChatGPT’s smaller prompt volume and single ad placement create a disconnect between supply and demand. The new format and measurement tools take aim at the first 2 complaints. With inventory that tight, I’d want every impression reaching the right buyer first, and right now nobody can promise that.

Do ChatGPT ads work? Early results aren’t a pattern yet

OpenAI backed the announcement with partner-measured results for WeightWatchers, Dose and Portland Leather. For Dose, WorkMagic reported a statistically significant lift, with 67% of incremental purchases coming from net-new customers.

If it works for Dose, it doesn’t mean it’s going to be the same percentage for everyone else, and a handful of results published by OpenAI itself isn’t a pattern yet. OpenAI also says its work on incrementality is still in its early stages.

And I want to see more under the hood, especially how ChatGPT actually decides whether to serve an ad to someone at all, and how many of the people who saw those ads had personalization on.

I judge new AI ad channels by what founders I know actually spend and what comes back. A friend of mine, an award-winning interior designer for residential and commercial projects, put quite a big budget into a ChatGPT ads test, and unfortunately, absolutely nothing came back.

That surprised me even more because ChatGPT now has its own interior design option in the Images tab, called “Redesign any room.” People thinking about a renovation are already showing ChatGPT exactly what kind of room they want, which should make it easier than ever to match them with a designer. Even with that, her ads brought nothing in.

How to get recommended by ChatGPT without paying for ads

Until ChatGPT gets there, I’d put a bootstrapper’s limited time and money into organic content. Content keeps working for you, while paid ads are rented visibility that keeps you in bidding wars. ChatGPT is no exception: when more than one ad is eligible, OpenAI says it ranks them on a mix of relevance and advertiser bids.

1. Write for one buyer’s problem at a time. OpenAI says ads don’t influence ChatGPT’s answers, so getting into those answers still takes organic content.
2. Double down on what already brings customers in. That same interior designer gets most of her customers from referrals, and nothing she’s tried on Facebook, Google or ChatGPT ads comes close to that organic reach.
3. Set your own signal for when to test. Switch on personalization and memory in your own Ad Controls and watch the ads you get. When they start matching your conversations and what ChatGPT knows about you, test with a small budget.

ChatGPT already uses memory to personalize its answers, down to refining a restaurant search with your location or dietary preferences. That’s exactly what I want it doing with its ads, for every user who sees them. When it gets there, ChatGPT could be the best-targeted ad channel bootstrappers have ever had.

And the founders who spent the wait building organic content will walk in with an audience that already trusts them, so when they finally pay for ads, they’ll be amplifying something that already works.

About the Author

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post ChatGPT Ads Got Images, but Targeting Depends on a Toggle appeared first on BYVI.

For a founder, if customers can’t find you, nothing else matters, and the rules for getting found are being rewritten right now. Not long ago, a useful page and a solid ranking meant traffic showed up on its own. That deal is breaking. AI Overviews are already driving a 34.5% lower click-through rate for the top-ranking page, and it’s not going back.

An AI model now sits between you and your customer, deciding how to describe your company and whether to name you at all. How AI interprets your brand, the sources it leans on and the answers it keeps repeating, is fast becoming a founder’s real edge.

Ahrefs Evolve returns to San Diego on October 12, and the lineup tells you exactly where search is headed. I went through the agenda for both days and pulled the 9 speakers whose sessions will actually change how your startup gets found.

Alina Vandenberghe, Chili Piper

• X
• LinkedIn

Alina Vandenberghe co-founded and bootstrapped Chili Piper into a category leader, so she’s the rare speaker who’s actually built a company from zero.

The talk: “How We Built 1,000 Agents by Making Failure Promotable”

She’s sharing how Chili Piper built a culture that rewards AI experimentation instead of punishing the misses, plus the org structure behind it. Her mastermind, “How to break through the noise,” gets into building a brand people remember.

For bootstrapped founders, this one is a must. Alina grew a company the lean way, so her playbook actually fits a tight budget.

Rand Fishkin, SparkToro

• X
• Instagram
• LinkedIn

Rand Fishkin publishes the zero-click research everyone else quotes on X and Linkedin. His latest: 68% of Google searches now end without a single click, according to SparkToro.

The talk: “The New Way the Web Works: Zero Clicks, Declining Traffic, No Attribution, Yet More Influence than Ever”

He’s bringing the charts and data to back a different kind of marketing, one built for how buying actually works now: influence over traffic, measurement over attribution, and giving your audience what they want instead of pouring your budget into Google and Meta ads.

For founders who still feel pressure to chase traffic numbers, that’s exactly the evidence you need.

Tim Soulo, Ahrefs

• X
• LinkedIn
• YouTube

Tim Soulo is Ahrefs‘ CMO, and the company’s whole business is helping people get found, so he’s speaking from the data behind one of the most-used toolsets in search.

The talk: “How to make your business DISCOVERABLE in the age of AI”

He’s walking through how discovery is changing as AI reshapes search, and he’s teasing new Ahrefs tools built for it. Expect a practical read on how LLMs decide what to surface and what that means for your site.

For founders, the product announcements alone are worth watching, since they tend to signal where the tooling market heads next.

Ethan Smith, Graphite

• X
• LinkedIn

Ethan Smith is the founder and CEO of Graphite, a premium vertical AI growth agency that runs SEO for a long list of venture-backed startups, so he’s seen what drives growth across tech companies.

The talk: “The 5% Of AEO & SEO: Research Backed High Impact Strategies”

He’s cutting SEO and answer-engine optimization down to the few moves that actually drive results, backed by his own research, and showing which common tactics you can safely ignore.

If you’re doing marketing between standups and sales calls, this is the prioritization you’ve been missing.

Ross Hudgens, Siege Media

• X
• LinkedIn

Ross Hudgens built Siege Media into a content agency known for earning links and citations at scale, which is exactly the skill set that now decides whether AI answers mention you.

The talk: “PR for Machines: Earning Coverage That LLMs Repeat”

He’s breaking down how to earn the kind of coverage and citations that LLMs pick up and repeat, with placement frameworks and a data strategy you can copy. And his hands-on mastermind, “Content That Gets Cited,” goes deeper into the on-page side.

If AI answers are going to quote someone, this is how you become the one they cite, without an ad budget.

Aleyda Solis, Orainti

• X
• LinkedIn

Aleyda Solis is the founder of Orainti and one of the most respected independent SEO consultants working today, known for shipping checklists and frameworks rather than theory.

The talk: “Ecommerce in the Age of AI Search: Shifts, Risks, and Strategic Moves”

She’s mapping how AI search changes ecommerce, from agentic shopping to new visibility risks, and what to actually do about it. Her mastermind, “International SEO in an AI Search Era,” covers what’s changed for cross-border growth and what hasn’t.

If you sell products or sell across borders, you’ll walk away with specific steps that you can apply right away.

Ross Simmonds, Foundation AI and Distribution

• X
• LinkedIn
• Instagram

Ross Simmonds is the founder of Distribution.ai and Foundation, a B2B content marketing agency, and the Amazon bestselling author of Create Once, Distribute Forever. He’s known for turning content distribution into a growth engine, and has worked with organizations around the world, from fast-growing startups to global brands.

The talk: “How To Win In The New Search Era: Reddit, YouTube & Backlinks”

He’s showing how Reddit, YouTube, and backlinks drive visibility in both Google and LLM results, with a distribution playbook built for small teams. His mastermind, “The State Of Reddit In The SERP & LLMs,” zooms in on why Reddit keeps showing up in answers.

For founders without a big content budget, this is channel strategy you can run yourself.

Austin Lau, Anthropic

• X
• LinkedIn

Austin Lau does growth marketing at Anthropic, the company behind Claude, so he’s running these plays inside a team operating at the AI frontier.

The talk: “AI as a Force Multiplier for Your Marketing Team”

He’s sharing a framework for where AI actually multiplies a marketing team’s output, from automation to custom tooling, and where humans still matter. His workshop, “An Honest Conversation on Why AI Adoption in Marketing Is Hard,” digs into why most teams stall.

If you run marketing with 2 people and a stack of AI agents, you’ll leave knowing how to automate the right work and where a human still does it better.

Harry Dry, Marketing Examples

• X
• LinkedIn

Harry Dry built Marketing Examples into a widely read marketing newsletter by being relentlessly simple and he teaches copywriting better than almost anyone.

Mastermind: “Copywriting Cardio: Learn to Write by Actually Writing”

It’s a hands-on writing workshop, nine rounds of actually writing, not slides about writing. He’s also giving a day two talk on making your work land: “The Half Hour You Learn How to Write a Story: The Opposite of Storytelling”.

If you write your own landing pages and homepage copy (and at the early stage, you do), make time for both the workshop and the talk.

Why Ahrefs Evolve matters for founders

Search is shifting faster than most playbooks can keep up, and Evolve has gathered the people rewriting them. Across 2 days in San Diego, these speakers will map where AI search is heading and what founders should do about it.

What founders publish now becomes the source AI answers quote and the data future models learn from. That’s a rare opening: the founders who move now get cited before the category gets crowded and become the default answer in their space.

About the Author

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post Speakers Tech Founders Can’t Miss at Ahrefs Evolve 2026 appeared first on BYVI.

Tally asks its new users one simple question during onboarding: where did you find us?

Marie Martens and Filip Minev founded Tally in Ghent, Belgium in 2020. For years, it grew mostly through word of mouth and the “Made with Tally” badge on forms people filled out. Then, in the spring of 2025, more and more new users started answering that onboarding question with “ChatGPT.”

By June 2025, the team reported that ChatGPT had become their “#1 referral source,” with over 2,000 new users signing up via AI tools in the previous months (and that was only the ones they could track). At the time, Tally had 8 people, 4 of them part-time.

In April 2026, at $5M ARR, they wrote that their viral loop “is no longer our biggest acquisition channel. AI-powered search is.”

By September, Tally had crossed $6M ARR and 2.5 million users with a team of 10, and 43% of new users were discovering it through ChatGPT, Claude, Gemini, Perplexity and other AI assistants.

That’s a tiny team winning attention in the place where buyers now make decisions. And the way they did it is good news for every founder who can’t afford a content team.

The tiny-team advantage in AI search

Early-stage founders tend to assume AI search belongs to big brands with big budgets. But the data points somewhere more encouraging.

In 2025, Ahrefs studied 75,000 brands to see what correlates with how often a brand shows up in Google’s AI Overviews. The strongest signal was branded web mentions, meaning how often people talk about you across the web, with or without a link. Backlinks, the thing founders used to pay agencies for, came in far lower at 0.218.

There’s a second reason to start now. In 2025, University of Toronto researchers asked 4 AI assistants the same questions and compared their answers. For well-known brands, the assistants recommended the same names 76% to 81% of the time. For niche players, that dropped to 71% to 76%.

For a startup, that gap is a huge opening: AI’s picture of smaller players is less settled, so what you publish and what people say about you this year can still shape it. (The study is a preprint that hasn’t been peer reviewed, so treat the numbers as a rough guide.)

Being talked about is something you earn by being useful to people, one conversation at a time, and that’s work a founder can do better than any agency.

Founders often ask me for the trick to showing up in AI answers. I always give them the same one:

Become the most helpful brand on the internet for your ICP.

Google’s own guidance lands in the same place. Its guide to AI search boils everything down to one question for site owners: “Is this content that my visitors would find satisfying?”

Be where your users ask

Tally’s team spent years showing up in the forums, Reddit threads and blogs where people asked about forms and survey tools.

In their words, they “invested for years in showing up in those places by sharing what we learn, answering questions, and being human.”

When ChatGPT’s spring 2025 update turned web browsing on by default, all of that history became source material. In Tally’s view, those years of real recommendations are why AI tools started sending people their way.

Their April 2026 update says it best:

“Both of these channels, word of mouth and AI search, are powered by the same thing: our community. Happy users sharing the word, recommending us on Reddit, supporting us online, and now even organizing meetups.”

The mix keeps widening, too: Tally says ChatGPT, Claude, Gemini and AI Overviews “continue to grow as referral sources.”

The lesson for founders is simple. The conversations you join today become the answers AI gives tomorrow. Google is clear that faking this doesn’t work: its guide says that seeking inauthentic mentions across the web “isn’t as helpful as it might seem.”

So skip the paid Reddit posts and the review swaps. Answer real questions, as yourself, under your own name.

Publish what only you know

The second thing small teams have is data nobody else has seen.

Tally has shared its numbers in public almost from day 1. Every milestone goes on its blog, from a revenue table that starts at $0 in September 2020 to its upgrade rate (about 1% of badge signups eventually pay). In 2021, the founders told Failory that building in public helped them get noticed by other founders.

Google’s guide tells site owners to avoid anything a generative AI model could easily produce. If ChatGPT could write your post, it has no reason to send anyone to read yours.

Start with what you already have: onboarding data, churn reasons and the questions customers ask on every demo call.

Make your product easy for AI to read and use

Getting recommended by AI is the first win. Tally also made its product easy for AI tools to read and use.

Its site has an llms.txt file, a plain-text map written for AI tools. It describes Tally in 2 sentences, then points to the help center, the API docs, pricing and integration guides. Its developer docs go further: they’re available as a single plain-text file, so an AI agent can read the whole API reference in one go.

Then Tally let AI assistants use the product directly. Its free MCP server, along with connectors for Claude and ChatGPT, lets people create forms, edit them and pull in responses without opening Tally.

In August 2026, Tally crossed 1 million AI tool calls in a single month across Tally AI and its MCP server.

One caveat before you spend a sprint on this. Google’s guide says Google Search ignores llms.txt and similar files, so they won’t move your Google rankings. Treat them as a bonus for the other AI tools and agents that read them.

How to show up in ChatGPT answers: 5 steps for startups

You don’t need a content team or a big budget for any of these. You need about 2 hours a week and the knowledge already in your head.

Then check what AI actually says about you. Once a month, ask ChatGPT, Claude, Gemini and Perplexity the questions your buyers ask, using a private chat mode with memory switched off, so your own past questions don’t skew the answers.

Become the answer AI gives your buyers

Big companies have budgets, content teams and, often, agencies. Early-stage startups have something more useful in AI search: founders who talk to customers every day and know things nobody else has written down.

In 2020, Tally was 2 founders in Ghent with a free form builder and no investors. Today, 43% of its new users arrive because an AI assistant recommended it.

None of that took a big team or a big budget. It took years of answering community questions, sharing numbers only Tally had and making the product easy for AI to read and use. You can start on all of it this week with the knowledge already in your head.

And if you keep showing up, AI search won’t be able to leave your product out.

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post How Tally Turned AI Search Into Its Biggest Growth Channel appeared first on BYVI.

Michael Wang’s piano students have homework they can’t do alone. Every RCM practical piano exam includes ear tests, and practicing for them usually takes 2 people. “Somebody else will have to play it for you,” Michael explained to a room of builders at Shopify, “and you have to identify by ear.”

Most kids practicing at home don’t have a second musician in the house, so those drills wait for the next paid lesson. Michael is the teacher waiting on the other side of it. He’s also a computer science student, and he’s building the app that lets his students drill alone.

Michael pitched on September 13 at Builder Sundays, the free weekly sessions where Shopify opens its offices in Toronto, Montréal and Ottawa to solo builders and small teams working on early ideas.

His product: A piano skill trainer that helps students learn complex musical concepts in an organized, simple way.

The piano homework nobody can do alone

Start with the part of music lessons that parents rarely see.

In Canada, the best-known graded path for piano is the Royal Conservatory of Music (RCM). Beyond performing pieces, every practical exam tests the ear: the examiner plays an interval, a chord or a short melody, and the student has to name it, clap it back or play it back. In the RCM piano syllabus, ear tests are worth 10 of the exam’s 100 marks, with sight reading worth another 10. Michael put the “learn it through reps” share of the exam at 10 to 15%, which lines up.

Ear skills improve almost entirely through repetition. The catch is who plays the notes. As PianoTV’s RCM guide points out, practicing ear tests at home usually means you need a teacher or a friend to play the examples for you. Most kids don’t have a musician on call on a weeknight, so the drilling happens during paid lesson time.

Michael sees the cost of that from the teacher’s bench.

“I’m wasting a lot of time with students in class,” he said, on skills “they could be working on at home.”

His fix is surprisingly simple. A student picks a level (or builds a custom set), the app plays a sound, the student names the interval, and a stats screen shows how they did. 2 modules are live. The student can run the whole drill alone, which doubles as a pitch to parents: nobody else has to sit at the piano.

He didn’t find this problem through a market map. He found it at the piano bench, lesson after lesson.

Why teaching gives Michael an unfair head start

Michael’s customer list came straight from his own week: music academies like the one he teaches at, school music teachers who could send drills home as homework, students cramming for exams, and parents, whom he called a “very good paying customer.” He’s watched those parents up close. “I see how parents really care about their children’s musical education,” he said, calling it “a place where people are willing to spend money.”

The exam body itself proves him right. The RCM sells its own Online Ear Training & Sight Reading tool for $9.99 a month, and its product page lists more than 5,000 exercises per tool plus a progress tracker.

Families already pay for this category, which saves Michael the hardest early question: will anyone pay at all?

He named that competition honestly in his pitch: similar tools exist at $9.99, he said, but they’re harder to use, and the older ones don’t work well for younger kids. That’s a real opening, and a narrow one. The RCM tool already tracks progress, so “easier” has to be something a young student and a busy teacher both feel in the first 5 minutes. The strongest version of his edge is the teacher’s workflow: assign the drill, see who did it, and walk into the next lesson knowing where each kid is stuck.

The monetization question: free app or paid subscription?

Michael came with 2 business models and asked the room for feedback, which is the right instinct at week 3.

Option 1 was a free product funded by business deals or by selling user data, which he compared to Duolingo. Option 2 was a proven subscription at around $9.99 a month. He’d already done the tradeoffs himself: free is enticing but earns little until you’re huge, subscriptions work but people cancel when they don’t see value. He also flagged his own gap on the data route: “I don’t have a background in cyber security.” Knowing where you’re weak at week 3 puts you ahead of most founders at year 3.

2 facts make his choice easier.

First, Duolingo makes most of its money from subscriptions. In Q2 2026, Duolingo reported $258 million in subscription revenue out of $298.5 million total, with 12.7 million paid subscribers. The free app brings people in, and the subscription pays the bills. Duolingo also launched a piano-based music course in 2023, focused on reading notes and playing songs, which is worth watching even though it doesn’t target RCM ear tests.

Second, Michael’s users are kids, and children’s data is some of the most tightly regulated data there is. In 2023 the U.S. Federal Trade Commission took action against Edmodo, an education platform, for collecting children’s personal data without parental consent and using it for advertising. The order carried a $6 million penalty (suspended because Edmodo couldn’t pay) and required the company to delete models built on that data. For a solo founder, one mistake there can end the company.

What founders can take from Michael’s first 3 weeks

Michael is early, and he said so. But his pitch still holds 4 lessons any early-stage founder can use.

The next test is whether students who drill with the app walk into their ear tests better prepared. A student who needs the reps, a parent who’s happy to pay and a teacher who wants class time back add up to a real business. Michael is 3 weeks in, and he’s only getting started.

About the Author

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post How Michael Wang Turned His Piano Teaching Into a Startup appeared first on BYVI.

LinkedIn tells you who’s connected. a16z’s new Cosign wants to tell you who would actually vouch for you, and in startups, that kind of trust is the most valuable currency there is. I don’t think it replaces LinkedIn, but it could fill a gap LinkedIn was never built for, and founders who show up early will feel the benefits first.

The sample profile on Cosign’s homepage belongs to Sam Altman, with 42 cosigns for AI leadership. Cosign couldn’t have picked a better poster child. Back in 2009, years before OpenAI, Paul Graham put a 23-year-old Altman in his Five Founders essay, on the same list as Steve Jobs and Google’s Larry Page and Sergey Brin. That single public vouch did more for Altman’s reputation than any résumé line could, and the team behind Cosign points to it directly in their launch essay.

Most vouches never get that stage. They happen in backchannels, over dinner, in a DM that says “you have to meet her.” a16z partner Olivia Moore summed up why that matters:

“The most valuable data in all of Silicon Valley is who has conviction in whom.”

Olivia Moore, Partner at Andreessen Horowitz

On September 25, a16z launched Cosign to put that data on a profile page. The obvious question is whether it becomes the next LinkedIn for tech.

My take: it won’t dethrone a 1.3 billion member giant, but it has a real shot at becoming the place startup people check before they hire, invest or join.

What is Cosign? a16z’s new professional network

What Cosign actually does

Cosign calls itself a curated professional network for the people and companies that drive the startup ecosystem. The mechanics are simple. You build a profile, and people who worked with you attach cosigns to specific strengths (“company building,” “product vision”), each with context about how they know you.

The announcement draws the line against LinkedIn: a LinkedIn connection tells you two people clicked “accept,” while Cosign tries to show who worked side by side or who shaped someone’s career as a mentor or advisor.

It goes further than people. You can cosign companies and build public lists, like investor and writer Packy McCormick’s companies he’d tell a new grad to join. Cosign then surfaces open roles at those companies automatically.

The team compares it to investing without the money. When you cosign a company, you’re publicly betting your reputation on it, and if it takes off, everyone can see you spotted it first.

Add community questions (“Who’s a great fintech angel?”), an Industry Watch feed of talent moves, fundraises and product launches, and a jobs board, and you get something that looks a lot like the private tools VCs have kept for themselves for years. That’s exactly how a16z describes it.

The job LinkedIn never nailed

LinkedIn tried to solve the “is this person actually good?” problem once. In September 2012 it launched skill endorsements as a way to give kudos with one click. It worked a little too well. By 2016, LinkedIn had racked up 10 billion endorsements, including for talents like “chewing gum” and “punching.” And recruiters noticed this pattern early. One told Forbes in 2012 that endorsements carry “no skin in the game.”

That’s the gap. A one-click endorsement costs nothing, so it tells you nothing. A cosign attached to your name, with context about how you worked together, puts your own reputation on the line.

Erik Torenberg, the a16z general partner who recruited the Cosign team, named the second gap in his launch post: strong endorsements happen on X every day and then vanish into the feed. Cosign wants to make them permanent.

Excited to introduce @Cosign: the curated professional network for the startup community: https://t.co/yAnRgqRifu

Our goal is to create the following:

– A comprehensive startup directory of investors, companies, and operators, including what they worked on and who they worked…

— Erik Torenberg (@eriktorenberg) September 25, 2026

Why LinkedIn is so hard to beat, and why its rivals didn’t make it

Cosign isn’t the first startup to take on LinkedIn, and the last 2 serious attempts both shut down in early 2025.

Polywork let people show off their skills and side projects, not just their job titles. It raised more than $40 million from backers including Stripe’s Collison brothers and went viral in 2021 and 2022. Then growth stalled, and it shut down on January 31, 2025. Founder Peter Johnston still believed in the idea. He said he was “hopeful that one day, another team takes a crack at this problem again.”

Read.cv, a clean, design-friendly LinkedIn alternative, went the same way. Perplexity acquired it in January 2025 and wound the product down.

Today I’m excited to share that @read_cv is joining the team at @perplexity_ai in their mission to make the world's knowledge more accessible to everyone. This is incredibly bittersweet for us, as the start of this new chapter will mark the end of our time with @read_cv.

It has… pic.twitter.com/6CUinOEGsi

— andy chung (@_andychung) January 17, 2025

Both were good products, and both hit the same wall: people keep their profile where recruiters look, and recruiters look on LinkedIn. That habit is hard to break. LinkedIn has 1.3 billion members, and its revenue grew 12% in the first 3 months of 2026, an extra $521 million.

Why Cosign has a better shot than Polywork and Read.cv

Cosign starts with 3 advantages Polywork and Read.cv didn’t have.

There’s also a nice bit of history here. LinkedIn itself started as an insider network. It launched on May 5, 2003, and ended its first month with 4,500 members, mostly the co-founders’ own contacts. Starting narrow with a tight crowd is how LinkedIn began too.

Cosign’s biggest bet: private, mutual matching

The part of the launch I’d watch closest isn’t live yet. Torenberg describes an intent network where you privately signal “I’d invest in them,” “I’d hire them” or “I’d work with them,” and eventually even “I’d acquire this company.” When interest exists on both sides, Cosign makes the match.

It works a bit like a dating app for deals. Say you’re a founder and you’d love a certain angel investor in your next round. Today, you send a cold email and hope for a reply. On Cosign, you’d privately mark “I’d want them to invest.” If that angel has also marked “I’d invest in them,” Cosign introduces you. If they haven’t, nothing happens, and nobody ever sees that you asked. The same works for hiring: an engineer quietly open to a new job and a founder who’d love to hire them can find each other without either one making the awkward first move.

LinkedIn’s closest version is the #OpenToWork badge, which 40 million people used in a single month in 2024. It’s one-sided, though, and career coaches warned Fortune it can make candidates look “desperate.” Cosign’s version would work in both directions and stay hidden until there’s a match.

If a16z gets this right, Cosign stops being a place you visit to look people up. It becomes the place where deals, hires and first checks start, and that’s a much harder habit for any rival to copy.

My verdict for founders

Cosign won’t be where your mom updates her job title, and it doesn’t need to be. The real opportunity is more focused: becoming the reference check for startups.

When an investor wonders if your co-founder is the real deal, or a senior engineer wonders if your startup is worth the pay cut, Cosign wants to be the first tab they open. That job is worth a lot, and LinkedIn has never owned it.

Cosign still has to earn it, and 2 things could get in the way.

First, the vouches have to mean something. If cosigns mostly go to people who are already famous, or people start trading them (“I’ll cosign you if you cosign me”), they’ll become as meaningless as LinkedIn’s one-click endorsements.

Then there’s the question of neutrality: a16z has to stay fair to everyone. Cosign is built by one of the biggest VC firms in the world, so founders backed by rival firms need to trust it won’t favor a16z’s own companies.

I’m getting in early myself. I plan to cosign the people who shaped my career, with real specifics about how they helped, and build a list of the founders I’d back without a second thought. If you’re a founder, operator or investor in the startup world, it’s worth doing the same while the network is still forming.

The best people you’ve worked with are probably people most of the internet has never heard of. Cosign is a bet that saying so out loud, with your name attached, can change careers. Believe in someone early, and put it in writing. Your reputation grows every time you’re right.

About the Author

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post Will Cosign Replace LinkedIn for Tech? Why a16z’s Vouch Network Has a Real Shot appeared first on BYVI.

Ryan Cohen pitched Chewy more than 100 times before anyone said yes. “Nobody wanted to give me money,” he later told the Wall Street Journal. The pitch was an online pet store that would out-service Amazon, and Silicon Valley kept passing (nobody wanted to fund the next Pets.com).

Then, in late 2013, a Boston firm called Volition Capital wrote a $15 million check. What convinced them was how often Chewy’s customers came back. Less than 4 years later, PetSmart bought Chewy for $3.35 billion, the largest e-commerce acquisition on record at the time.

That same firm just closed a $950 million fund to do more of exactly that.

Volition Fund VI at a glance

How Volition Capital’s $950 million fund works

On September 14, Volition announced the final close of Fund VI at its hard cap of $950 million. The fund was oversubscribed, it’s the firm’s largest ever, and it pushes Volition past $2.6 billion in assets under management. Since 2010, the firm has backed more than 60 companies, with exits including Chewy, Assent Compliance, Connatix and Rounds.

We're excited to announce the closing of Volition Fund VI with $950 million in limited partner capital commitments, bringing our total assets under management to over $2.6 billion.

With Fund VI, we'll continue to focus on investing in high-growth, founder-owned technology… pic.twitter.com/rshxpu7Pjp

— Volition Capital (@volitioncapital) September 14, 2026

For scale, the previous fund closed at $675 million in early 2023, so Fund VI is roughly 40% bigger.

Co-founder Larry Cheng gave The Information the inside view. Per Dealroom’s summary of that interview, typical checks run $25 to $50 million, Volition takes a board seat, and the sweet spot is companies doing $5 to $50 million in annual revenue that raised little or no institutional money. Cheng said Chewy returned the whole of Fund II, and the sale of Rounds returned about half of Fund IV.

Why Volition is doubling down on bootstrapped founders

The timing is the interesting bit. The Information framed it against a VC rush to back the hottest startups before they’ve made any revenue at all. Meanwhile, limited partners (the pension funds and endowments that invest in VC funds) spent this spring and summer nervous about AI disrupting software. Volition raised a record fund into that anxiety anyway.

Cheng’s read on AI is the part founders should underline. He told The Information that AI is making products cheaper to launch, and that he’s seeing more companies hit roughly $10 million in annual recurring revenue in their first year with fewer than 10 employees, a pace he said he’d never seen before. His take in the Fund VI announcement:

“We have backed founders with aspirations for greatness, and we believe truly remarkable businesses are being built today in the new era of AI.”

Larry Cheng, Founding Partner at Volition Capital

So let’s connect those dots. If a tiny team can reach real revenue without a seed round, the pool of “bootstrapped companies with traction” gets much bigger.

That’s Volition’s hunting ground, and it explains why its partners are leaning into AI rather than running from it.

The promise: big outcomes, less founder risk

For all the AI talk, Volition’s promise to founders hasn’t changed since the firm was founded in 2010. Managing partner Roger Hurwitz said it best in the Fund VI announcement:

“Our mission to help founders achieve their dreams without risking them hasn’t changed. That north star has guided us through every market and technology cycle, and we believe it will continue setting us apart.”

Roger Hurwitz, Founding Partner at Volition Capital

“Without risking them” is the part worth underlining. It’s a promise the firm has been testing since it raised its first fund in the aftermath of the financial crisis, and the release spells out what it means in practice:

For a bootstrapped founder, that’s the whole trade: you’ve already taken the early risk with your own time and money, and the right growth partner should help you scale it without giving up the company you built.

How to get on Volition Capital’s radar

Volition publishes its investment criteria openly, which is rarer than it should be. Read them as a checklist for becoming fundable without fundraising first.

A recent example of the model in action: Volition led a $40 million minority investment in Keynes, a connected TV ad company, earlier this year. Minority means Volition doesn’t take the company over.

When Volition isn’t the right fit

Growth equity is a great deal for the right company and a poor fit for others. A $25 to $50 million check comes with a board seat and a partner who expects a large return, often through a sale (Chewy and Rounds were both exits). If you’d rather run a profitable company forever and pay yourself dividends, you may never need this money, and that’s a perfectly good outcome.

It’s also a small club. Volition calls itself “extremely selective,” and a fund this size still backs a limited number of companies.

The good news: the work that makes you fundable is the same work that keeps you independent. Paying customers, strong retention and healthy margins give you options, whether you take a $25 million check or never raise a dollar.

Lidia Vijga

Lidia Vijga is the Editor in Chief at BYVI, founder-driven media that spotlights early-stage startups at the moment most press skips, when coverage can actually open doors. She’s also a content engineer who builds in-house growth engines and trains AI agents that help companies scale content and win brand visibility across traditional and AI search. Previously, she co-founded DeckLinks, which gave her firsthand experience building and scaling tech products.

The post Volition Raises $950 Million for Bootstrapped Founders Who Skipped VC appeared first on BYVI.