training wheels copyI’ve noticed a common refrain with investors at accelerator demo days.

“What did you think of the companies?” I ask.

“Eh,” they say with a shrug.

Over the last month I’ve asked around 15 of them the same question, it became difficult to find a VC who’d argue, in general, IN favor of the industry. (None cared to play the role of the jerk who issued a dismissal on the record.) But it was clear that the special accelerator magic of injecting of insta-growth into seed-stage companies has faded in the last year.

For VC’s, these programs act as a filtering or vetting system for early stage companies. If a company has the TechStars or Y-Combinator stamp of approval, it is probably less risky than, say, a random person who emailed you with an idea.

The problem is that now there are just too many programs to get excited about. Investors have grown weary of demo days that place more emphasis on wowing the audience with a snazzy presentation and big potential numbers than on company fundamentals. And it’s not even clear how great the vetting program is — if 100 programs graduated 20 to 30 startups each last year and the average VC only meets with a few hundred companies a year, how much of a filter is it? The category has been been hurt by overproliferation.

Some investors take the argument a step further, saying the jury is out on the whole premise: VCs that say the elite Y Combinator is the only program that matters, but in the next breath complain that it is an “over-fished pond,” with too much investor interest driving up valuations. And even though most YC companies have no problems raising additional capital, the program has only produced two big “winners.” Dropbox and Airbnb make up three quarters of the value of Y Combinator’s $10 billion portfolio. That’s two huge successes out of something like 400 companies. While it’s common for investors to get the majority of their returns from just a few big hits, that ratio is a bit more skewed than most would like. Paul Graham of Y Combinator recently admitted his program had grown too quickly, announcing the next class would be scaled back to fewer than 50 companies after graduating 84 last year.

That said, it would be lazy to issue a sweeping dismissal of accelerators. The “should you or shouldn’t you” question is nuanced and different for each company. But I managed to gather a few takeaways over the course of my conversations with investors, founders and accelerator programs.

A familiar chorus was that accelerators don’t help most companies. They don’t necessarily hurt anyone, but they don’t help. The best accelerators — Y Combinator especially — are up front with this reality. They don’t guarantee success. The program gives companies tools for success, but the actual success is in the hands of each company.

The worst accelerators do guarantee success. This is where the sheer volume of new programs becomes a problem: Accelerators need the good companies more than the good companies need the accelerators.

In order to get the best companies to apply to their program, they need to promote themselves aggressively. They need to sell entrepreneurs on the benefits of their brand and services. More accelerators mean more competition for the good companies. These companies are, after all, giving up 6 to 10 percent of their equity in exchange for three months of services and some cash. The services include a range of things, like office space, discounts on servers and cloud storage, access to mentors, access to PR, access to press, access to investors, access to lawyers, access to recruiters. The thing I’ve heard repeatedly from founders who chose not to join accelerators after being accepted is that they thought the program was over-promising. Many of the services offered can be found through a little networking and email hustle or through a strong co-working community like General Assembly or WeWork Labs. And bonus, they can do it that way without giving up any equity.

The promises eventually start to feel a little predatory, to the point where startups can go in feeling like the accelerator will supercharge their company, no matter how crappy the idea is. Airbnb was a success because Brian Chesky and his team went in and worked like crazy. They had no choice — the company had been on the verge of death before its founders wedged their way into YC out of sheer persistence. Having scraped by with baseball card binders full of credit cards, subsisting on nothing but promotional Cap’n McCain cereal, Chesky and his co-founders redefined the word hustle. They knew merely getting into the program was no gravy train to success, and they essentially “won” the accelerator.

Chesky has since argued that the most valuable thing he got out of Y Combinator was a piece of advice from Paul Graham, who told him to stay in close contact with his customers. That shaped the company’s way of doing business in the crucial early days. Sure, wouldn’t have had access to Graham’s advice had he not joined Y Combinator. But founders in major ecosystems like New York, Boston, LA, Boulder or Seattle still have access to many potential mentors. The key is hustle. Having a good idea helps, too.

So don’t join an accelerator unless you can win it. That requires being at a stage where you can actually benefit from meetings with investors and the press. If there’s a chance you’ll pivot halfway through the program and demo a one-month-old idea to a crowd of investors and media, you’re not going to win. If the program has truly notable services that can’t be obtained on their own (the biggest one is access to mentors), you won’t be able to take advantage of them if you’re still in development.

Going into a program is not unlike a launch at SXSW. It’s nearly impossible to break through the noise. And despite the low odds of breaking out at SXSW, the success of Twitter, Foursquare and GroupMe have tempted plenty of companies into trying to repeat their formula. Even if you do break through the noise, you better have something damn special, as the unremarkable Highlight learned the hard way last year.

It’s that old saying “the harder I work the luckier I get.” Those who tend to outperform when they graduate from incubators likely would have made it without them.