A word on incubators

An incubator - much like a producer - combines ideas, people, and money to create a new company. I won’t cover accelerators - which like YC - take an existing team and idea and help them move faster.

As much as there is a incubator market, I don’t think it matters too much. Our incubator, Boulton & Watt, doesn’t compete against other incubators. Our companies compete in their own markets.

Below, I highlight some of the tradeoffs, review some common criticism, and then review a long (but not complete) list of incubators where there’s something to learn from. Multiple different models will be successful but they have clear trade-offs. I’ve thought about this model since 2015 when I worked at Sidewalk Labs, where we mostly failed but had a few successes in Cityblock and Replica.

What to think of all this? I’d pull apart a few axes:

• Idea maze or clear execution? Some ideas require hundreds of tiny tweaks without much data - I’d argue these are generally bad ideas for incubators or require a committed co-founder for much longer. The classic social idea fits in this mold - it would have been hard to start Facebook or Snap in an incubator. (Tinder, incubated within IAC, is a clear exception.) Other ideas have strategic execution risk, but the customer/product/technical loops are much cleaner.

• Milestones to partner with CEO. Do you partner with the CEO at inception (I like to call it the “deck and a check” stage) or once more milestones (revenue, team, funding) have been accomplished? The former model is more ‘scalable’ but can run into adverse selection of team.

• Role of incubator. Is the incubator a financial sponsor, a ‘producer’ pulling things together, or doing the dirty work?

• What kind of risk? Technical (SHV, Flagship), rate of growth/margin (Thrive, Juxtapose), execution (B&W, Fractal)?

Common criticisms of incubators

• Does the incubator own too much? Is the founding team/CEO properly incentivized?

  • Answer: IMO depends on the stage. If you give less than true founder level economics at ‘deck and check’ stage, you set up risk for founder motivation and future financings.

• Will the founder of incubator just go work on the ‘best’ one? What does the ‘institution’ of the incubator look-like?

  • Answer: Why doesn’t a venture investor go full-time on their best investment? To some degree, this is a question of what the founder wants to do (and their ability to hire ‘better than them’ talent to lead their founded companies)

• The biggest outcomes (Google, Apple, Amazon, Facebook) were not incubated.

  • Answer: Sure - but Genetech, Snowflake, Moderna, Palo Alto Networks all were. If you aim to start a trillion dollar company reliably, best of luck to you.

Incubator examples

Fractal. Focused exclusively on vertical SaaS, at peak was starting around 10 companies a month, has now paused new company creation. Extensively researched hundreds of industries, paired vetted research with $1 million and (generally) less experienced CEO + CTO.

Thrive. Generally works on capital intensive traditional industries where a light tech layer + consumer orientation can redefine it. Examples: Oscar, Nava Benefits, Cadre. Thrive does concept development in-house, pairs with an experienced founder/CEO at roughly slide deck stage and fairly standard terms but with high Thrive ownership.

Greylock. As far as I can tell, a Greylock incubation means the idea emerges in conversation between Greylock investor + founder, and the company starts with funding and office space from Greylock on day zero. Examples: Workday, Palo Alto Networks, Sumo Logic, Tellapart

HVF. Max Levchin’s incubator which started Affirm, a period tracking app and other companies. Essentially dissolved when Max went to Affirm full-time. This is a common ‘pitfall’ you’ll hear described - will the founder just go work on the ‘best’ one?

Flagship Pioneering. The incubator that started Moderna. They aim for ‘moonshot’ biotech ideas, with extensive gating along risk milestones, resulting in multiple public companies. Best article I’ve found on them, I liked this quote from the founder:

“If you do something in parallel, you are forced to get to the essence of it,” he says. “I think of it as chess. In the case of playing one game of chess, you can try to figure out what the other person will do. If you play parallel chess, you have got to get to the essence. You cannot procrastinate over the particular move. That mindset of parallel entrepreneurship to me was a provocation.”

Boulton & Watt. We work on one company at time with an experienced team, taking each company through series A, and then partnering with an experienced CEO. Low volume and high success rates, focused on vertical SaaS. We’ve brought Moxie to millions in annual revenue and Series A in less than a year, with more to prove.

Alleycorp. Kevin Ryan’s incubator. Kevin founded MongoDB, Businessinsider and Gilt Groupe in 2007 (!). Since then the formalization has taken a more scaled and flexible approach - pairing Kevin’s and sourced ideas with EIRs and capital from their fund.

Sutter Hill Ventures. They start one or so companies per year with extreme technical risk. SHV bets on secular trends where the market is obvious if the product works - usually harnessing an existing demand made 10x cheaper or faster. They pair technology expert founders with a partner from SHV who acts CEO until the company reaches a commercialization phase. Kevin’s article is canonical piece on Sutter Hill. Examples: Snowflake, Pure Storage, Observe

Redesign Health. Focused on healthcare exclusively. I love this model because you can build compounding networks of healthcare customers and executives in a huge market. That said, they’ve also built a few direct to consumer healthcare companies. They partner at the early validation/idea stage with CEOs. I’m not sure how successful they’ve been.

Juxtapose. Largely gone after private-equity style businesses in fragmented markets with low NPS, like dental, vitamins, and property management. Build initial concept in house, seed fund it, and then go after a been there, done that CEO. It’s an interesting model because while the initial capital + risk is venture equity, from the outside subsequent rounds are PE risk/returns. Examples: Tend, Care/of, Great Jones.

Accretive. No longer active (although founder is now chairman of Juxtapose). Their model was to partner deeply with a large corporate (generally in healthcare) sign the deal, and then fund the company.

---

Missing an interesting incubator I should have mentioned? Got a fact wrong? Email me at sam@boultonwatt.com

Thanks to Coyne, Dan, Jaclyn and Jess for giving this a read.