Everywhere you turn there’s a new startup accelerator launching. Time will tell if that’s a good thing or not, but as I said in my last post, if you’re not doing it [an accelerator] to make money, you’re doing it wrong. The feedback was mixed, which doesn’t surprise me, since there are many accelerators operating at the moment that don’t have a primary mandate of returning dollars on investment. They’re designed for other primary goals: economic development (which really should be about making money), building a startup ecosystem (again, really driven by making money), job creation (bad idea), etc. But ignore that for a moment, and let’s assume you’re hell bent on starting an accelerator no matter what. Great! So what should it look like?

10 Things to Think About When Starting a Startup Accelerator

I’ve thought about this a lot and talked to a bunch of people running accelerators or thinking about it. I certainly don’t have a monopoly on the truth, and furthermore, the jury is out on the entire accelerator model; none of them have been around long enough to really know what’s going to happen and how to create long-term, sustainable value.

First and foremost I’ll say this – all accelerators are not the same. It’s important to realize that they’re all different, tackling different stages of a startup (some are pre-idea even), in different locations, with different amounts of money and different goals. We can bundle accelerators into an industry class, but you have to go into the details to really understand them, the nuances, and build a model that works for you and your circumstances.

1. Who runs the accelerator? Like startups, I’d recommend at least two partners (founders). You want experienced entrepreneurs that have “been there, done that” in all aspects of a startup’s life. Managers and administrators need not apply. There’s a lot of management and administration (in running an accelerator), but you need hands-on, experienced entrepreneurs that understand what it’s like to run a company.
2. How much help will you provide? The word “mentorship” has a pretty broad definition. Same with “hands-on”. What does that actually mean? It’s not obvious, but I can tell you that it takes more work than you think. At Year One Labs we coined the phrase “co-create” which meant we wanted to genuinely help in the creation of the startups, and work with them every single day. That puts us squarely in the shoes of the entrepreneurs, but on the other hand we’re also investors. So what side of the table are we sitting on? Probably right square in the middle.

   In places where the startup ecosystem isn’t super strong, you’re likely going to attract a lot of first-time entrepreneurs that are going to make the same first-timer mistakes that everyone else makes. They won’t have the network, experience or safety net of support to guide them and de-risk things. Expect to be doing a lot of work in these situations.

3. Are there mentors and what do they do? Most accelerators have a strong batch of mentors. How they actually help isn’t as clear. You need to define clear expectations for mentors and make sure you have buy-in from them. Putting names on a list for the sake of it might help attract a few founders, but it’s not sustainable. Mentors need to be active participants and add serious value add to the equation.
4. How long will the program run? The popular accelerators are 3 months. I still think that’s too short to really build anything of significance if you’re starting from an idea or a rough prototype. If you’re starting with more than that and joining an accelerator for the network it’s a different story. But you have to decide how long you want to run the program for. Year One Labs is 12 months, but as I pointed out in a previous post (Lessons Learned Launching a Lean Startup Accelerator), I think that’s too long.

   More important than the question of length of the program is this: What’s the point? What’s the end goal of the accelerator? What stage should startups be at when they get out? How baked do they need to be in order to be on solid footing and raise follow-on capital? I’d argue that in places other than Silicon Valley and New York, startups need more baking. They’re at a disadvantage in terms of building early adopter traction and density of investors for follow-on — so they need more traction and more de-risking before investors will pay attention.

5. How much equity will you take? The popular accelerators are in the 5-12% range. We took more at Year One Labs, but we also offered more money and more time. So you can follow the “accepted model” of the popular accelerators, but I’m not sure it makes sense. This might be an entirely separate post, but the economics around accelerators are not crystal clear. It’s not obvious that you can make a lot of money running an accelerator. I’d say the risk increases as you move into locations that don’t already have strong startup ecosystems (more on that below). I’m working out an alternative model (it’s not new, just brainstorming in my head at the moment) and will share it in a future post.
6. What will you invest in? Most accelerators are fairly broadly targeted. Year One Labs was “web and mobile.” Not much in the way of specialization. But I think specialization is worth exploring. We’re seeing accelerators emerge in verticals – RockHealth is an example. In non-tech startup hubs attempting to launch a bunch of consumer Internet startups is going to be crazy tough. But a lot of places around the world do have specialties; things they’re known for. In Montreal it might be gaming. In another city it could be something else. Creating a vertically-focused accelerator has a number of advantages. You get leverage across the startups in terms of people you attract and the mentors. You can attract more focused investors and align investors much earlier who might be interested in a batch of companies. If one startup is failing, it’s more likely that the founders can jump to another one that’s working because they’re in similar areas of interest. Leverage is a good thing. I expect we’ll see a lot more specialization in accelerators coming soon.
7. What stage will you invest at? As I said above, all accelerators are different. There’s lots of overlap, but plenty of opportunity for differentiation as well. YCombinator and TechStars are definitely mixing things up, investing early but later as well, even after companies have raised a seed or Series A round. If you’re not in a startup tech hub, you might find it challenging to even find enough fully baked teams and fully baked ideas (let along prototypes or products with some amount of traction). That might force you to make bets on individuals, pair people together, have to brainstorm ideas inside the accelerator, etc. It’s worth thinking about this and what your expectations are for the stage of startups joining.
8. How many startups will you do? You need to know how many companies you bring into your accelerator. If you go small, you can work more closely with each one. But your odds of financial success are lowered. Go for volume and your odds of there being a breakout or two increase, but your time commitment to each one drops. There’s also the question of deal flow and how many you can actually bring in that are high enough quality. You don’t want to compromise to fill a quota.
9. Who will provide the follow-on investment? In many places outside the Valley, NY and Boston, there aren’t a lot of investors. And most investors like investing close to home. That’s changing (a bit) but I’d say it’ll always be true. It just makes sense to keep your investments close to home. So what do you do if your accelerator is not in the Valley, NY or Boston? Do you prepare every company to be exported? Do you build strong networks with distant investors in the hopes that that’s good enough? And what about the local investors? If they don’t invest, what kind of signal does that send to investors looking in from afar?

   One possibility is to have an accelerator with a sidecar fund that’s able to cherry pick the deals it wants to add more capital into. Signalling (for those the sidecar doesn’t invest in) will be a problem, but at some point everyone has to make a bet. You can’t expect every accelerated startup to raise capital.

   The risk that we fund a ton of small startups off a cliff is real. Eric Ries writes a fairly foreboding but honest post, Winter is Coming, where he says:

   “I expect that a shocking number of the current crop of incubators, accelerators, and other startup-support programs will suddenly disappear. In summer, it’s all-too-easy to have your program look like a success, because there is an endless supply of talented people becoming first-time entrepreneurs and an endless supply of investment dollars chasing them when they graduate. It’s hard to know, in summer, which of these programs actually add value and which are glorified admissions officers. Winter will tell. If you depend on one of these program for support, be ready.”

   Bottom line, accelerators need to think a great deal about where follow-on capital will come from, and how to manufacture the follow-on funding as much as they can.

10. Who are the acquirers? In non-startup hubs this is a big issue. Most cities / regions don’t have a lot of big acquirers (especially of consumer Internet companies), and many acquisitions are a result of previous closeness between the startup and acquirer. Big acquirers do look all over the place, but they look mostly in their backyards. Talent acquisitions are more likely when the acquirer is closer. Knowing the major acquirers in an area, and having good relationships with them, can have a big impact. It might create more clarity around a specialty for the entire accelerator. A good accelerator has to have a bead on the entire food chain for a startup — from the talent to mentors to investors to acquirers. If the accelerator hasn’t build that food chain or doesn’t have a strong network across that food chain it’s going to be much harder to drive results.

I don’t believe we can take the most popular accelerator models and automatically copy them in other places. The variables are different. This has a lot to do with physical location and the health (or lack thereof) of the startup ecosystem in specific places. Copying blindly is usually a recipe for disaster. Where I think there’s opportunity is for accelerators to differentiate across different variables (just like we recommend startups should do): business model, strategy, target market, etc. Take a Lean Canvas as an accelerator and fill it out. What’s it look like?

Accelerators need business models (and they need to make money!) to be successful and sustainable, and those business models have to take into account all the critical variables described above.