Word-of-mouth is the tried and true way to spread the word about your business, news, or product updates. For businesses, allowing your customers to tell their friends about how awesome your product or service is can be a great way to increase your brand recognition and attract new customers to come in and check out what you’re doing. As Basecamp wrote back in September, the web-based project management system has grown increasingly in popularity because customers have been able to tell their friends and colleagues about it and bring them over to the service.
Curebit, an alum of the Y Combinator winter class of 2011, launched at demo day back in March as a way for online stores to increase revenue through referrals by turning existing customers into marketers. Curebit wants to optimize referral systems for eCommerce platforms, and today they’re launching a new product to help do that more effectively. It’s called the “Social Referral Optimizer” and Curebit Co-founder and CEO Allan Grant tells us that his product is essentially like Google Web Optimizer for referral systems: It solves the hard problem of getting high conversion rates from referral systems.
Curebit supports the type of split referral incentivization that has worked so well for companies like Dropbox. For those unfamiliar, this split referral system is when you recommend Dropbox to a friend, and when they sign up, both you and your friend get some kind of reward, be it 5GB of storage for free or a shiny nickel.
According to Grant, Dropbox worked at optimizing their referral system for months before it began to have any real sort of effect on customer acquisition and conversion, so Curebit wants to take this optimization (i.e. pain in the butt) out of the process for any site — even yours.
What the Curebit team came to understand as they tested different form of referral optimization is that conversion depends a great deal on the details of the offer (the language the offer is presented in, the type of incentive, etc.), so their Social Referral Optimizer allows sites to automatically break an offer into its various permutations and test them across Curebit’s partner sites to see which has the highest conversion rate.
The startup’s optimizer enables sites to vary the amount of the discount, the offer text, the message one uses to share it with friends (whether that be via Facebook or Twitter), the landing page, as well as the graphic design of each page. But really the coolest part is the cross-site optimization: For sites that don’t have enough volume to get those statistically significant results, users are able to take advantage of Curebit’s software, which learns as it tests from other sites across the Web (about 700 of these are already available for testing).
Interestingly, from the data the startup has collected so far, they’ve learned that the conversion rate depends not so much on the amount of the discount that one offers friends for their referrals, but more on the text one uses — how the entire offer is expressed. And so far, the results have been encouraging. With Curebit’s optimizations, users are able to get 30 to 60 percent of of the customers that buy to share exclusive offers with their friends, resulting in a direct, measurable life in sales of up to 15 percent.
The startup is also launching an additional two new features today, including “Facebook Sponsored Stories” integration, which is designed to optimized shared messages, turning them into high-conversion social ads, as well as “Social Influencer Tracking”, in which Curebit identifies customers that have shared an offer (and are super influencers) so that merchants can personally thank them.
For example, when Gina Bianchini, co-founder of Ning, shared one of Curebit’s offers, the DODOcase founders got an email about it right away and were able to reach out to say thanks immediately thereafter. Next time the Ning co-founder buys something that uses a Curebit referral, the DODOcase guys will know right away, even if she doesn’t share. Pretty cool.
For more on Curebit, check out the video below: