Accelerators · Venture capital

Can a "full stack Accelerator" work better than existing VC models for early stage?

Eric Wold

April 7th, 2015

Been reading a lot about how broken the VC model is for early stage funding, and how the majority of Incubators & Accelerators don't do enough, or invest enough.

Some founders have access to capital (friends & family, etc), others don't.  Some founders are great at starting an organization.  I don't think there is any real correlation between those two things.

Why can't we (in theory) merge a VC + DevShop + Accelerator into a single organization and use performance review, drip funding, and real working mentors to align interests?

My attempt to define the problem more clearly:

If you are an Angel, why or why not would you prefer to write a check the old way, or invest via an intermediary that holds $$ and releases them over time?

If you are an Entrepreneur, why or why not would you consider joining such a group?

Mike Jones @ Science in L.A. seems to be doing this very well.  I'm hoping more groups will adopt the model or explain why it's flawed.
As a founder, you’re always in fundraising mode (whether active or passive). In this course, we’ll teach you how to successfully raise follow-on capital, establish a valuation for your company, build an investor pipeline for your next round, and more...

Jessica Alter Entrepreneur & Advisor

April 7th, 2015

There is a real trade-off between the amount of equity/control you give up or retain. Science is the perfect example of this - most of those CEOs are holding a lot less of the company than you would expect. The second reason is that because when you invest in the early stage th idea is 100% going to change and so you're investing in the people or person and if they are just brought on after the fact that's a very different thing. it's also different when you don't have as much at risk - you're getting a salary, you're not worrying about someone else's salary, etc.

John Seiffer Business Advisor to growing companies

April 7th, 2015

@Eric - regarding your comment 
>>I think it's possible to figure out a model that makes 3-8x "base hits" more common, and de-emphasizes "crap shooting" for unicorns.<<

I say this from the perspective of an angel investor, and having run an angel group in CT for 2 years. I think this has to do with the investment vehicle (equity in a non-liquid market) more than anything else. When investors buy equity that can't give a return until the company has an exit, then their time frames have to be pretty long 2-10 years. And because the markets are non-liquid they only get a return on their money when there's an exit. That means that even if the company is profitable, they are most likely to get zero return. From that follows the idea that the few successes in the portfolio have to be home runs to cover all the losses. Therefore they can only afford to consider companies that have the potential to be home runs.

If you could change the vehicle of investment, you could change that calculation. So debt is a vehicle that typically returns money sooner, and has fewer events with a zero pay-out. However the upside is much more limited. This means it's not going to work where the downside has no protection therefore it's not useful to lend money to startups. 

A vehicle that bridges the gap between the two is one that is not very common. It's a revenue based pay out (sometimes called royalty based). Rather than purchasing equity - which could be considered a claim against a future exit - a revenue based payout is a claim against future revenue. So the company pays some percentage (usually 5-10%) of sales off the top until the investor gets a suitable return. This means if the company has sales but never has an exit, there's something in it for the investor. 

Eric Wold

April 7th, 2015

Wow, great points from everybody. Really high-quality challenges which might lead to material refinements. FYI, I consider this an "open source" concept because of how many voices have already contributed to aspects of this. If anybody can benefit, feel free to explore on your own or reach out to collaborate. I'm happy to share more details of the experimental model than would be reasonable here (too long).

@ Jessica: Nice point! Ok, so balance between "grabbing" all the equity you can vs making sure the founder has enough to be motivated is critical - that rings true, and I agree. FYI, Mike @ Science told me he sometimes he "resets" a cap table before spinning the company out on its own at A round time if what you are describing has happened. That's amazingly generous (ok not totally altruistic, he does it to make a project fundable), but to me, it is evidence of a model weakness. For such a model to really be viable we can't count on people like Mike that will reset a cap table to restore more share to the founder.

@ John: Well said, I agree strongly that focus on validating the problem and a solution customers are willing to pay for are critical and should precede any real product dev or hiring. I think it's possible to figure out a model that makes 3-8x "base hits" more common, and de-emphasizes "crap shooting" for unicorns. However, I like the idea of a crack dev team coding to disprove portfolio models early as part of the market fit challenge. Some people have an instinct for teasing assumptions out of a product hypothesis, and then designing tests to try and disprove them.

@ Lane: Great point on team. What if interests are truly aligned and the startup is allowed to take the staff developer(s) that worked on their product? The goal here would not be hold the startup back, it would be to defray hiring costs until a bit later than normal. Think about the stair-step way hiring bodies ramps up. Now imagine a startup in the very early stages being able to ramp up between 0 - 3 developers in 0.10 increments of a body, while still getting top notch work. I can say from experience that most first-time technical co-founders write spaghetti code when there is weak oversight. As a coder and former product manager I'm confident the average project started by a mature onshore dev group will be better than "random 1-2x startup coder". Normally a dev house has a profit motive to cut costs (time spent on your work). There won't be the same motive to cut corners in a dev team that works across the startups in a shared portfolio. If anything, I can imagine a strong argument going the other direction that extra care needs to be taken to ensure they aren't too "comfortable" and producing at less than maximum velocity. So performance/velocity must be monitored in a transparent way and even incentivized. Motivation might be addressed by the entire team earning equity in a blended employee compensation pool across the portfolio.

@ Eoin: Yeah, that's the kind of broadside I was hoping for! I'll post a few thoughts after I digest implications.

John Seiffer Business Advisor to growing companies

April 7th, 2015

I read a post a while back that said one major VC firm was trying this - instead of hiring just analyst, they were hiring people to support their portfolio companies with things like recruiting and marketing. But I never heard much after that - perhaps they are doing it in stealth mode, or perhaps it didn't work out. 

Having said that, I think the big problem is startups are not methodical enough in discovering a repeatable, scalable business model. Any hiring done before that - and much product development done prior to that - is wasted. Too many investors don't understand how to evaluate these things either so they invest where they shouldn't - or perhaps when they shouldn't is a better way to say it. 

Till an entrepreneur can fill in the numbers on the business model formula LTV-CA*N and has data to back up the customer's description of the problem they're willing to pay to solve, then investing in them is a real crap shoot. I will admit there are situation where the potential is so huge, the crap shoot is worth it, but they are not as prevalent as people think. 

This kind of learning would work well with your "drip" idea but you'd need people who really understood it to be good judges of whether the right milestones were achieved. 

Benjamin Olding Co-founder, Board Member at Jana

April 8th, 2015

I may be missing the subtlety of what you are suggesting, but this reminds me of two experiments from the dot com era: IdeaLab and Reactivity.  We called them incubators at the time, but that word has changed a lot in meaning since then.

They attracted talented individuals who were more risk averse than founders uninterested in incubators.  I have no idea what the roi was - my guess is they had both a lower failure rate & lower overall returns.  In other words, likely they net a lower roi, but with less risk compared to "average". I'm hand-waving - I have no data.

I tend to think it's the people involved that matter more than the funding structure, so if I was an investor, I'd be more concerned about attracting the people I wanted to invest in than worrying about the structure I put around them (I.e. even if your structure is 100% better than anything else, if it biases you away from individuals most likely to create the big wins, you shouldn't expect the approach will result in big wins).

IdeaLab is still around, though not quite the same.  Reactivity spun its last company out in early 2000s with its own name, a clever way to silently kill itself.  I don't think it's an indication of a failure of this approach necessarily; they were undercapitalized.

If you're a mediocre vc or angel struggling to attract quality early stage founders, finding highly talented but more risk adverse potential founders to fund sounds like a pretty good pivot.  If you're a high-performing vc or angel, I doubt messing with the finance strategy is at the top of your to-do list - especially when it's the big wins that contribute the lion's share of your roi stats.

Eric Wold

April 8th, 2015


Well said. Hopefully we can figure out a structure that doesn't suck as bad as the older models because, as a 3x founder, I couldn't agree more with this:

<<All that said - I'd love to see something like this as a service model, where the incubator/studio is working for the founders, rather than founders reporting to an incubator or a co-CEO. I want to run my own company and build my own team, but like all founders, we often need extra hands but have little to no cash to rent them.>>

I learned a lot in the enterprise SaaS model about aligning interests and keeping customers by honestly meeting their needs. Not by locking them in with contract terms. My ideal scenario is where a founder is offered a basket of services and chooses which ones they will consume, based on the holes in her/his own team, and how fast they want to get milestones done... and they can fire us at any time and any vesting would stop at the same time as service delivery.

My ERP system, just before I left, was achieving an annualized retention rate of 99.2% and projected life per account at >120 months and rising. With NO contracts. When I say I hope to fully align interests perfectly, that's not BS.

Thank you for the honest feedback!

Ken Carpenter Software Engineer at Arista Networks

April 12th, 2015

This sounds like a great idea to me.  I had discussed a similar 'incubator' type of concept with a friend here in Vancouver.

On the tech side, there are so many startups recreating the same things over and over.  Websites, dashboards, databases, reporting, mobile apps, security practices, backup & recovery procedures, etc.

Obviously each startup has unique needs, but I can definitely see creating these things once and reusing a lot of the work from one startup to the next.  The legalities of this sharing would need to be figured out, but that shouldn't be hard.

Harder might be getting an existing tech co-founder to accept your choice of software stack.  For startups with no tech co-founder or a cooperative one though, this could save huge amounts of time and money, and give the startup a battle-tested framework to build from.

Mark Sylvester CFO at Invenshure, LLC

April 10th, 2015

Very interesting conversation - thank you Eric for beginning it.  The model that has been described above, and specifically by Eric in the original post, is very similar to the model that we (Invenshure) utilize -  

We build companies based on novel and disruptive technology.  Invenshure's staff (operations, finance, engineers, scientists, etc...) "roll up our sleeves" from the inception of the company, to the ultimate disposition.  By the time we would consider exiting the company, the operational infrastructure would have been built, product line refined, the market has been scoped, the management and support team is in place, and the overall strategy of the company is being executed.  Think of us as a group of serial entrepreneurs with different strengths, all aligned in a unified vision to build multiple different high tech companies at the same time - with the backing of our fund(s).   

We began this model in 2011 and continue to refine/expand it.  


April 7th, 2015

Don't think it's a terrible idea but to play devil's advocate:

1. "If you are an Angel, why or why not would you prefer to write a check the old way, or invest via an intermediary that holds $$ and releases them over time?"

Usually, the funding process to any fund from their LPs involves capital calls as funds are needed. It's unusual for a fund to be holding the $'s. If anyone, it's usually the co that sits on the $'s. 

Why? Because drip-funding means you can't just focus on the business -- you're always worried about the next little feed. Drips are good when someone is on life-support; if you want to be healthy, you build up some reserves for the journey ahead. 

Also - drip-funding creates some dependencies that most founders would rather avoid. Competition in some form of funding marketplace is almost always a much better deal for the founders.

2. Of the folks doing stuff like this, they're quite good at testing lean businesses rapidly -- i.e. they are constantly building "small" products and doing market-testing before expanding. This version of lean is one way to build a company but it's not the only way and implicitly defines the types of products that can be built.   

3. VC + DevShop + Accelerator -- that's a lot of people who are going to be working on failed companies. If they are staff at the organization, they're not going to have the big equity motivation that founding teams have so they're more expensive and (arguably) less motivated/likely to succeed. 

4. What core competencies do the founders bring? Why wouldn't this organization just incubate its own ideas a la Idealab?

5. + Jessica's and Lee's respective points.

Eric Wold

April 7th, 2015

@John: Another great point. I'm hoping two factors might have an impact on that liquidity issue.
1) Evergreen perpetual structure removes some of the artificial time-based pressure to liquidate the whole portfolio by maturity. It also encourages constant reinvestment and growth throughout fund life.
2) Also, hoping to see more growth in the secondary market for quality early growth stocks that are pre-IPO such as: and 

Doing research around the possibility of lightly using secondary liquidity to keep a fund cash-flow neutral so we can always be recycling into more new startups.

Perhaps what you are talking about explains why the last $20m raised by Science was via debt instruments?
Does the Evergreen model also permit early returns to be used in creating such liquidity internally within the fund?