TL;DR: Startup accelerators face a fundamental challenge to their value proposition: they don’t “own” their networks, and therefore struggle to continue extracting fees for accessing them. Classic disintermediation. Their responses to that challenge take a number of forms, and generally involve either dropping their price or attempts at controlling ecosystem players; the latter of which is misaligned with the interests of entrepreneurs and startups.

Related reading:

• Relationships and Power in Startup Ecosystems
• “Top Startups” Lists and Accelerators

As I’ve written before in the above posts, Startup Accelerators became “a thing” in ecosystems because they were a reasonably optimal method for solving the “noise” problem faced both by startups and investors; a problem which became more visible as the cost of starting a company went down. With far more people “starting up,” early-stage investors needed someone to help them filter out duds. The solution, referred generally as “sorting,” is similar to the value prop offered by elite universities to employers needing talented labor, and students needing credible ways to signal their talent.

By creating credible brands (signals) for quality entrepreneurs, accelerators reduced the search costs for early-stage investors who, instead of needing to filter through lots of duds themselves, had a concentrated place to build their pipeline. That value proposition attracted investors, advisors, great employee hires, etc., and over time successful “alumni,” which magnifies the value proposition to entrepreneurs who, in exchange for equity, got a fast-track to building their network and raising capital.

For some time, you had a virtuous cycle with clear “network effects.” Attract great entrepreneurs, which then attracts investors and other key people, which then attracts more great entrepreneurs, and so on and so forth; just like a classic network effect for a software platform. During this period, accelerators can build significant leverage over their ecosystems as gatekeepers to talented entrepreneurs, and use that leverage to push the market in directions the accelerator wants.

The “Network” Can’t Be Controlled

But accelerators face a distinct problem that doesn’t get talked about a lot publicly, but local market players absolutely know is there: they can’t lock in (air quotes) “their” network. It’s not proprietary. The “networks” of startup accelerators are really just compilations of individual peoples’ networks; not at all like a “network” of a tech platform for which the tech “owner” can sustainably charge access fees. Those people in the accelerator’s “network” aren’t employees of the accelerator, nor are they paid out of its returns, and so they aren’t aligned in propping up the network’s “access fee.” Inevitably, people find it worth their while to simply bypass the accelerator and makes themselves accessible to founders directly, after having built their own personal brands with a few iterations with the accelerators’ initial cohorts. If a team needs X, Y, and Z, and I know X, Y, and Z and can help them get access with my own branding/signal, why should they have to pay this 3rd-party a fee to access those people?

So after a few years of an accelerator having filtered and aggregated a network, helping great people find great founders, and great founders find great people, the network takes on a life of its own. Suddenly with a little hustle and networking, it’s not nearly as hard as it was 5 years ago to simply navigate the “network” without ever needing to pay the gatekeeper. I’ve seen this play out in a number of startup ecosystems across the country, where accelerators faced an initial golden age when they were seen as prime “sorters” of an opaque ecosystem willing to pay for the sorting, and then suddenly the quality of entrepreneurs they can get to pay their “fee” starts to take a clear downward turn. Top entrepreneurs are, by definition, fantastic hustlers. They aren’t going to pay you for something once they’ve realized they can do it themselves with a little effort, or that someone else is offering similar “access” at a lower “fee.”

Once top entrepreneurs realize that they can bypass the accelerator and access its “network” directly, and word gets around, the value proposition of the accelerator can begin to unwind. Suddenly the accelerator cohorts start to fill not with the most highly skilled entrepreneurs (those hustle it out on their own now), but with lower quality entrepreneurs less capable of making things happen “in the wild” and therefore more needy of the accelerator’s high-touch, high-priced assistance. As the quality of the accelerator’s average entrepreneur goes down, the leverage over key people on the other side of the “market” – investors, advisors, etc. – goes down, and fewer of them show up to the accelerator; which then reduces the value prop for entrepreneurs, and you get the exact reverse of the original virtuous cycle.

Seeing this dynamic play out, accelerators have three ways of responding, and I’ve seen them in different markets.

Drop the Price

The first is to simply acknowledge that the accelerator cannot maintain the original value proposition they had before the ecosystem/network had matured, and drop their price accordingly. With less significant of a signal, and less leverage over the market, the high 6-8% fee can’t be sustained, so build something leaner that can be offered at a 1-2% level perhaps. I’ve seen these “leaner” accelerators enjoy some success. Some accelerators started out with the expectation that they were going to dominate a startup ecosystem with high “access” fees, and then over time got humbled when the market delivered a reality check.

Employ the Network

Another option is to convert the accelerator into a kind of “startup studio,” where the main pieces of the network are actually employees paid by the accelerator, or at least with deeper economic ties to the accelerators’ performance; reducing their incentive to leak out of the network. The key challenge here is whether the accelerator really has the cachet/leverage, and resources, to employ those people; or whether A-players find it far better to simply stay outside and keep their pipelines more open.

Another way to “employ” certain network players doesn’t require actually employing them, but simply maintaining some economic control over them. For example, a prominent accelerator might use referral relationships with certain law firms as a way to keep those firms from questioning the accelerators’ behavior, even if it’s clearly at times not in the best interest of the startups the firms represent. That strategy is straight out of the playbook of VCs. See: When VCs “own” your startup’s lawyers and Relationships and Power in Startup Ecosystems. Offering or restricting “access” to potential investments, clients, employers, etc. has always been a currency used by startup power players to keep other market participants loyal and “well-behaved.”

Try to Lock Down the Network

This is where things start to get interesting. So I’m an accelerator enjoying success, but I can clearly see that over time my ability to keep extracting gatekeeping fees over my “network” is weakened by my inability to maintain control over the investors, founders, advisors, etc. within it. Possible solutions:

• Lock Down Demo Day – Maintain tighter control over who gets access to demo day and, importantly, “discourage” founders from raising financing outside of demo day.
• Lock Down Financing Structures – Maintain tighter control over how financings within the network occur, by “soft mandating” that they follow templates created and controlled by the accelerator.
• Lock Down Network Communication with Technology – Create proprietary message boards, mail lists, and other media platforms for communicating within and navigating the network, to “incentivize” networking in ways that give the accelerator visibility and control.

Of course, none of this will ever be communicated openly as mechanisms for the accelerator to maintain power over an ecosystem/network, including founders. They’ll be spun as ways to provide efficiency and value for founders and other people. But as with much spin, there is a point at which it fails to pass the laugh test.

Listen in the market (what gets said privately rarely mirrors what is said publicly), and it becomes clear that the more aggressive accelerators have for some time been building local resistance; irritating investors who resent having a “big brother” dictating how to do biz dev and deals, irritating founders who don’t want to pay a gatekeeping fee for accessing specific ecosystem resources, and irritating other market players who don’t want a rent-seeker standing in-between them and potential business.

When an accelerator “discourages” a startup team from fundraising outside of demo day, it’s going to offer some paternalistic platitude about how having a controlled process helps “protect” the entrepreneurs, but what it’s really about is ensuring the accelerator has (i) leverage over the investor community via ability to deny and control access to its founders, and (ii) leverage over founders by controlling the venue in which they fundraise; which sustains the power of the accelerator to charge high gatekeeping fees.

Once I’ve publicly announced my cohort, the sorting is done and the signal is out. Investors don’t need me (the accelerator) anymore, and in many cases nor do the founders whose main purpose of joining the accelerator was to get “branded” to make getting meetings with investors easier. That threatens the power of the accelerator, which wants to charge not just for sorting/signaling, but for access to a network. “Locking down” outside fundraising, with some clever spin as to why it’s good for startups, is the response.

If an accelerator builds proprietary communication channels for alumni to utilize, maybe that’s to be helpful. Or maybe it’s a way of preventing the network from doing exactly what networks do organically, which is resist gatekeeping and build multiple nodes/channels to prevent a single point of entry through which a rent-seeker can extract access fees. Accelerator’s don’t hold monopolies on brands/signals that startups can leverage to get funding, and therefore other people (like angels, seed funds, and respected founders) within a “network” who can connect founders to money/other resources (offer cheaper “signaling”) are, in a sense, competitors whom the accelerator has a strong incentive to control. Maintaining control / visibility over communication channels is a way for accelerators to prevent leaner competition.

Accelerators are Service Providers, Like Everyone Else

The general conclusion from all of the above should not be that startup accelerators are bad or good; on an individual level many are of course full of great people. Instead, it should be that accelerators are profit-driven service providers and political actors, just like everyone else. They want to charge a higher price, and will do what they can to maintain their power to charge that higher price. Other market players will attempt to build alternatives, and drop that price, and the accelerators will respond by trying to compete with, block, or control those other market players. It’s just like VC, Law, and any other industry that caters to startups.

When transparent meritocracy and markets start to challenge a player’s ability to charge high fees, they often turn to politics; using backdoor relationships to build loyalties and amplify supportive messaging. Accelerators who maintain tight referral and economic relationships with specific funds, firms, and other market players do so in order to ensure there’s a loyal base of people out there toeing the party line, even as opposing voices in the ecosystem start to emerge.

For entrepreneurs, the message is simply to understand where their interests are aligned, and where they’re misaligned, with the interests of accelerators. Branding and signaling are useful. To the extent they are useful to you, use them, at the appropriate price. But by no means allow them to dictate how or when to fundraise, or how to navigate the network. It’s in startups’ interest (and that of ecosystems generally) to stay flexible and keep their options open, even if accelerators would prefer having a tight grip. The golden era of accelerators is almost certainly over, as startup ecosystems and networks have begun to mature, offering multiple accessible paths to networking and investment. But they will still have a place and function for a pocket of the ecosystem that needs them.

To the extent accelerators use politics and leverage to lock down ecosystem resources that founders could otherwise access on their own just fine, or demand that startups and investors do things in a specific way favored by the accelerator, they are no longer transparent market players; they’re rent-seeking gatekeepers. If there’s anyone that startup entrepreneurs love painting a bullseye on, it’s gatekeepers.

After-note: see Why Startup Accelerators Compete with Smart Money for some observations on how early-stage VCs are eroding the value proposition of accelerators further by bundling new roles/services alongside their investments, and moving up-stream.