Credible — customer reviews and actions as social proofs — hypotheses and questions after helping customers gain 115% conversion rate in 2 weeks

We launched a basic SaaS app Credible a while ago (our journey), with the hypothesis that customer reviews and actions can be strong social proofs and the way they are presented to potential customers matter a lot, because customers would be paying more attention to reviews and actions that are relevant to them. Relevance can be in the forms of many dimensions, such as where and when an action is taken, reviewer’s location, topics of the review, presence of keywords, length of review, and others. Before we launched it, we tested it with an online food ordering website and achieved really impressive results (+115% conversion with initial A/B tests using 50%/50% traffic split with >99% confidence)

Here, we use the term “actions” referring broadly to a range of activities: sales, sign ups, bookings, form submissions, and many others. Based on our knowledge, a large number of existing social proof products are already doing a great job capturing most common types of actions (such as sales) by using REST APIs and integrating with existing platforms.

However, the solutions for capturing and sharing reviews are still lacking. Based on our experience implementing these features, and our past experience building products aggregating and analyzing reviews, we know the technical complexity is magnitudes higher.

So far we have only implemented a fraction of the features we planned for. At this stage, we are in the middle of doing some A/B tests to quantitatively validate the hypotheses, but we would love to hear opinions from people on whether these hypotheses are fundamentally flawed, to the extent that they might not affect the conversions so much, or customers wouldn’t care.

There are many features we have in mind which we haven’t put into this product, for example, built-in analytics, automated A/B testing, automated optimization, inference and visualization of user preferences, auto user segmentations, and many others.

If you have used social proofs in your e-commerce or SaaS products, what do you think of these hypotheses and how strongly would you feel to give it a try / pay for a product like this? Let us know in the comments below. We would love to hear from you.