When Freemium Ain’t Free: Crowdbooster Offers a Painful Case Study
Crowdbooster is a Palo Alto, CA-based startup that solves one of the nettlesome new problems posed by the social media explosion—helping users figure out how many people they’re actual reaching when they share information on Twitter or Facebook. A lot of people like the service: within just two years of graduating from Y Combinator, Silicon Valley’s famed startup accelerator, the company’s real-time social media dashboards hit the 100,000-user mark.
And the company didn’t have to drop a dime on traditional marketing. Its customer base grew purely through word of mouth. “There was no effort spent on advertising or marketing of any kind, except just doing a good job for our users,” says co-founder Ricky Yean (pictured above right).
Of course, it helped that the tool was completely free—anyone could sign up for Crowdbooster and start tracking how many of their tweets got retweeted, and how many of their Facebook updates got liked. (I did it myself.)
But by the middle of 2012, Yean says, it was clear that something wasn’t working. The original business plan at Crowdbooster had been to develop a paid version of the dashboard that offered deeper analyses and fancier visualizations, and then convert a certain fraction of the free users to this paid plan—the classic “freemium” strategy. But the startup’s four employees found they were putting so much energy into creating and fixing features for their hordes of free users that the planned premium features just weren’t coming together.
So Crowdbooster decided this month to take a drastic step. It’s getting rid of its free product, and asking everyone to switch to one of three new paid plans, ranging in price from $9 per month to $99 per month. Non-paying users will be cut off after February 28. Yean says he knows that means the company will lose the bulk of its current users—but it had become clear that the freemium idea was a flop.
“It’s a decision we should have made much earlier,” Yean says. “But what we’re confident about is that through this whole experience, we have built a relatively strong brand.” The company has already attracted a lot of users at marketing, branding, and advertising agencies; now, Yean says, the company has the opportunity to ask them to pay for Crowdbooster, just as they do for other monitoring tools.
I’ve written before about the rationale behind freemium pricing plans, which remain popular among Software-as-a-Service companies. Indeed. I’ve even run into companies going in the opposite direction from Crowdbooster, creating a new free tier of service where none existed before. One of the big reasons to offer a product for free, after all, is that it helps get users hooked on your product, so they’ll be more willing to upgrade at some point. As an added bonus, it prevents competitors from undercutting you on price.
But what seems clear from Crowdbooster’s experience is that the freemium approach doesn’t work for every startup, and comes with hidden pitfalls. “I wouldn’t want to jump to saying that [freemium] is a bad idea for everybody; we have obviously heard case studies of it working,” says Yean. “But I would be careful. It obviously didn’t work for us.”
Back in 2010, when Yean and his cofounders Mark Linsey and David Tran were starting Conversely—the company behind Crowdbooster—the idea was to help publishers, marketers, advertisers, and others see which of their posts were resonating most strongly with Twitter and Facebook users. “You can’t just talk at people anymore—you need to say things they find interesting and engaging,” Yean says. “Our analytics is designed to look at individual tweets and Facebook posts and give you real-time feedback on how those posts are trending, and give you suggestions on how you can do better.”
Crowdbooster’s bubble-chart visualizations make it easy to see how many times a tweet has been retweeted, and how many potential impressions have resulted. It also shows how many likes, comments, and shares a Facebook post has received. Based on the historical performance of your posts, it can suggest optimal times of day for sending out new posts, and it can point you to influential social-media users worth engaging with personally.
“During the day, you may send out 5, 10, or 15 posts, and we make it really easy to see which ones are working,” Yean says. “If something isn’t working, you can adjust right away. That is the big use case, and the reason people choose Crowdbooster over other services.”
Yean says his team chose a price point of zero in order to attract as many users as possible, as fast as possible. “When we started out, we weren’t even sure if this was something the world needed,” he says. “The initial thought was just to get people using it, and see if it would be useful.”
The startup’s product roadmap called for the addition of a premium version with features like deeper analytics tools and team collaboration features for enterprise users. But the zero-price strategy worked almost too well, and soon the company had such an influx of users that it had its hands full just keeping the service running. “We had to continue to work on things that the free users were asking for, so we were never able to focus on the more advanced versions,” Yean says.
Yean’s team spent much of 2012 rebuilding its back-end system to grab the newest data from Twitter and Facebook (before that, the numbers had usually been an hour or two old). Adding the real-time feature pushed the company over the 100,000-user mark. But that was a bit of a Pyrrhic victory, as nobody was paying yet.
Which led to the company’s difficult decision. Just after the New Year, Yean announced that the free service would be discontinued, and replaced with a subscription-based service costing a minimum of $9 per month. (At that level, users can track one Twitter account and one Facebook account; for $39 per month, they can track up to 10 accounts, and for $99 a month, up to 30 accounts.)
Yean is realistic about the likely impact of the change. “I expect to lose most of our users,” he says.
Why not keep some kind of stripped-down, free version of Crowdbooster? “That was definitely an option,” Yean says. “But we don’t think you can do two things—maintain the free version and work on another, more advanced tool—and do it well.”
That doesn’t mean Crowdbooster will never release a free product. It just means they’ve decided to focus for now on earning money. (Obviously, no startup can go forever without bringing in some revenue, but Crowdbooster was in a particularly tight spot, having raised just $500,000 in seed capital.) “Our previous efforts at trying to do two or three things at the same time had not panned out well for us,” Yean says. “So we said, let’s focus on our serious users first, and get good at it, and when we have the opportunity in the future, we will create a free product that casual users are looking for.”
While Crowdbooster will now have to rebuild its user base from scratch, Yean says he doesn’t think of this as a do-over. But it might count as a pivot.
“I wouldn’t see it as starting over, because the product itself is still being used,” he says. “We have all these features we have been meaning to work on, and now we can actually work on them.” The startup will focus first on features that will appeal to small and medium-sized businesses, Yean says.
So, what advice would Yean give to other companies considering the freemium model?
“I would dig deep into the successful case studies to understand whether the freemium model is really working for them,” he says. “A lot of companies use it, and a lot of companies raise money with the user base that free attracts. But if free is really working, then why are they raising money?”
Actually, there’s a good answer to Yean’s question. Sometimes you don’t need to convert very many of your free customers into paid customers in order to have a lucrative business—Evernote (with conversion rates in the low single digits) comes to mind. But to attract enough free users to yield a good crop of paid users, you might still need to spend a lot on marketing. You also need to make sure that your free product is actually useful, and isn’t just a deprecated, feature-poor version of your paid product—and this, too, is expensive.
All of which, in the end, reinforces Yean’s point: freemium is risky. Before trying it again, he says. “I would need to find more case studies of freemium working. And I would err on the side of it not working.”