.png)
Three years ago, I launched Shaper with a vision of co-founding companies across industries that matched with pattern recognition I’d built through the LiveRamp and Datavant journeys.
Our progress has been faster than I ever would have imagined, and this was a giant year for us. Across the Shaper ecosystem, one company went from 7-figures to 9-figures of revenue this year, one company won a $25-million contract as its first major deal, and one company had a successful exit that yielded a nearly 100x return for Shaper. Across the companies, we’ve raised >$150 mm. and hired 300+ employees. And I think we’re just getting started!
In each of the last two years, I’ve written a set of reflections on what I learned (see the posts from 2024 and 2025). Here are 10 lessons from Year 3.
I’ve watched several early-stage Shaper companies ramp at meaningfully different speeds.
The companies that made the most progress finding product-market fit maximized their surface area to learn from the market: publicly building, iterating quickly, getting in front of customers, and trying to sell something from day 1.
Early-stage companies can spend a lot of time debating product, pricing, network effects, or long-term strategy. Those questions matter, but early on it is usually impossible to answer them in the abstract. Motion creates information.
There are lots of tactics to create surface area and motion, and it often involves decisions that feel “short-termist.” Go mid-market for faster sales cycles, even if they aren’t your ideal long-term customer. Do some things that aren’t automated, even if there won’t be a way to scale it up. Sell through channels if they get you to more customers, even if it might one day create channel conflict. Build in public and get hype, even at the cost of competitors taking your ideas. All of these tactics increase your speed of getting market feedback, and get your business into fast-iteration mode sooner.
At an extreme, within 2.5 years, Protege became a 9-figure revenue business — which is the fastest a company I’ve been involved in has ever made it to this stage. The key was creating surface area.
There is a lot of groupthink against services in the startup world. Some of it is reasonable: services businesses can be hard to scale and hard to defend. But I think a lot of this thinking is outdated, and the bias against services creates interesting opportunities to simply do the thing customers actually want.
Fractional AI was the clearest example. Most VC-backed companies entering the forward-deployed engineering world were searching for the productized angle. Customers, however, wanted exceptional engineers to help them solve important AI problems. We focused on what customers wanted, and we believed there was a $100 billion opportunity to get this right. Fractional AI still had to think carefully about scalability, margins, and defensibility, but it was fine to build a services-focused business with eyes wide open about how to build around those, and we ultimately got a home-run by staying true to this vision of the world.
Investors can be the worst culprits in perpetuating groupthink. When you fundraise, you’ll get 50 different opinions on your business, and more often than not, rejections. It is always worth understanding investor feedback and understanding the signals and insights that come from it, but founders should remember that most VCs are quite often wrong, especially when you have a novel point of view.
Personally, I started to formulate this view with the LiveRamp and Datavant journeys. They were both home-runs for the VCs that invested, but were not especially popular businesses with investors. Both were category-creation plays, and 99% of VCs looked at a nonexistent market and concluded “small TAM,” rather than seeing the opportunities I believed existed to create a winner-take-most, network-effect-intensive new market.
I gained more conviction watching several fundraises this year: businesses that fit the zeitgeist often had an easy time, while some companies with phenomenal underlying business models found it surprisingly hard to raise.
Fun fact: six months before Fractional AI was acquired, we struggled to raise a round. Most VCs wanted to push us toward productization and didn’t believe in a services vision. Listening to the pushback, pressure-testing it, and then deciding where we still had conviction was valuable. VC feedback is a useful signal on the business model, but that’s it.
Most founders are focused on all their first hires being either “builders” (engineers, product) or “sellers.” Within the first ~5 hires at most Shaper companies, we try to place a generalist Head of Business Operations who owns some mixture of recruiting, people, operating cadence, and making sure the trains run on time. I’ve come to see this as one of the highest-leverage early hires, and have never seen the hire happen too early.
Join Medium for free to get updates from this writer.
Subscribe
Remember me for faster sign in
I’ve now watched several Shaper companies go from startup to ~100 people in roughly two years. That is a hard ramp, and companies will generally break if they don’t think methodically about recruiting, culture, onboarding, and basic process before they need it. A superstar in this seat figures out how to build enough infrastructure so the company can keep moving quickly as it grows.
There are always rational reasons not to overdeliver: the customer is annoying, the deal was underpriced, the company has too much on its plate, or the product is harder than expected. It is appropriate to debate all this when deciding whether to take a deal or sell a particular product. Once you’ve sold it, though, I think the best CEOs take an uncompromising view: you have to overdeliver.
This becomes cultural. If quality is continuously traded off against other priorities, every team develops an explanation for why it is okay that a particular customer is unhappy or another deployment is late, and eventually none of your customers actually love you. The best companies do the opposite: going above and beyond becomes part of the culture, and the team rallies around delivery even when it is hard or not perfectly scoped.
We are in another period where many startups can raise very large rounds relatively easily. Even in this environment, having a low burn rate (and the optionality to flip toward profitability or use the balance sheet) is incredibly valuable.
Low dilution shows up in returns: “skipping” a round dramatically changes founder returns from companies (and many of the wealthiest entrepreneurs, eg. Bezos, Gates, Zuckerberg, etc., were able to raise relatively little in their early days and effectively skip rounds).
But beyond the financial logic, there’s a much bigger strategic reason: operating discipline is hard to install late in a company’s culture, and disciplined companies run more effectively. Lean companies prioritize better, avoid simply throwing people at problems, and often move faster because they have fewer handoffs and owners. Treat headcount growth as unfortunate, even when capital is unlimited.
Over and over again, I’ve seen that the worst trap entrepreneurs get into is “slight/medium” success — not a hockey-stick journey, and not a clear failure that is easy to walk away from. This can lead to entrepreneurs spending years of their life stuck, with a massive opportunity cost. This is compounded by guilt (letting down investors, employees, customers) and an intertwined sense of identity. The standard motivational stories (“we bootstrapped for 12 years, then found success!”) make it even more difficult to get to no.
Earlier this year, the CEO of Zenith Health decided to shut down her business. This was a very hard, and very correct decision: she is a phenomenally talented entrepreneur with a huge opportunity cost, and — while there were some signals of progress — the timeline to scale and the amount of early traction just didn’t justify more investment. So we navigated a shutdown, and helped the team land as gracefully as possible.
In this situation, my job is to listen, and help the CEO with their reflections. But making it an acceptable option to move on is an important part of my job.
Last year, I wrote that ~80% of my job is recruiting. This is still the case.
Over and over, the highest-leverage thing I can do for a company isn’t a stroke of strategic genius; it is getting the right person into the right company, in the right seat, at the right moment. This year, we’ve increasingly turned Shaper itself into a recruiting machine — helping companies build great early teams and maintain very high talent density as they scale. That is usually more valuable than anything I can contribute in a meeting. [By the way — We’re hiring!]
Founders are often asked, “Who might acquire this?” or “What is the exit strategy?” I don’t think either is a particularly relevant question to spend much time on. While it is good to know the market, and it can be useful to build relationships over time with big companies that are looking at the space, that’s not what drives M&A.
Both Datavant and Fractional AI reinforced that the timing of an exit, and even the identity of the buyer, can be a total surprise. What the two had in common is that we were building really good businesses, thinking long term, and not in a hurry to sell. That also creates the best negotiation leverage that leads to the best outcomes: if you don’t need to sell, a buyer has to pay a price that convinces you to change course, and you can genuinely walk away if the deal doesn’t make sense.
One of my pet peeves is people trying to put a clean label on Shaper. We aren’t exactly a fund, a family office, a venture studio, or an operating company. We sit somewhere between all four, and in different situations we lean in different directions.
I increasingly think that refusing the label has created alpha. Labels come with rules and conventional playbooks; staying in the gray lets us choose the right structure for a particular company and remain patient across a wide range of outcomes. Fractional AI is a good example: we were comfortable with different financing paths, levels of involvement, and exit scenarios, and that flexibility helped us make better long-term decisions.
* * *
Overall — what an incredible third year. Thank you to everyone across the Shaper ecosystem who made it possible. It still feels like we are at the beginning, and I’m even more excited for the year ahead.
19
