How To Fix a Crappy Conversion Rate

Discuss on HackerNews

The Problem

We’re an early stage startup searching for product market fit. We built our MVP and released it to the public in September of 2012, and people started finding our site and sharing it with others via email and hackernews. 

It seemed that people were really interested in Kera and we’ve been getting a fair amount of signups. All of our advisors and investors pressured us to keep driving those numbers, and we thought we were on the road to early retirement. After all, 50% month over month signup growth for a B2B SaaS company is pretty awesome. 

The point is we weren’t converting many of our signups into customers. Unfortunately for us, we fell victim to a vanity metric. The worst part is, we chalked it up to deficiencies in our product instead of looking for more effective answers. 

The Cause

After a lot of reflection, the big reason our conversion rate sucked was that too many different types of customers were signing up for Kera. There were the tire-kickers who just wanted to take a peek at our UI and UX. There were the late adopters who took three months to make a buying decision. Most of all, there were a lot of signups who just hadn’t felt enough of the pain that we solve. 

After all, if you expect people to pay for an MVP with a ton of missing features - you need to find the people who hate their current alternative with a passion. 

The Results

As a result, our attention was spread very thin. Learning from too many customer types meant that our product roadmap informed by inconsistent feedback. We started building features that only a small portion of our customer base really needed. 

Worst of all, we were providing crappy service to too many people - oftentimes people that never intended to become customers in the first place. What a waste. 

Our Solution 

The first thing we did was start gathering criteria about our successful customers. What were the characteristics of the companies that adopted our software quickly? We looked for patterns and started building a persona of our ideal customer. 

The next step is asking our website visitors to self-select. We’re using a google form to gather information about the companies that want to use our software. This isn’t because we want to sell to enterprises, and we don’t think it takes the “touchless conversion” off the table. 

Expected Results

We’re expecting our signups to take a big hit, perhaps as much as 80%. Our hope is that loss will be made up of tire-kickers, late adopters, and companies who haven’t felt enough pain.  In other words, companies that never intended to become customers in the first place. 

And because our signups will decrease, we also expect to see a boost in our conversion rate from trial to paid. And since we’ll be working with ideal customers, I expect a clearer picture of what features we need to add and remove from Kera in order to keep improving our metrics. 

And finally, I’m expecting to create a small army of delighted, paying customers. 

Have any of you tackled this issue in the past? What’s worked? What hasn’t? 

- @maxcameron