Just about two weeks ago we launched the a redesign of the account chart in Basecamp. This is where people can upgrade or downgrade their accounts.
The goal was to increase overall upgrade conversions and encourage people who are on Basic plans or lower to upgrade to the Plus plan (our lowest priced full-feature plan).
Results
I’m glad to report our design hunches appear to have paid off. We’re only about two weeks in, so we don’t have a ton of data yet, but we can compare the 14 days since the upgrade with the 90 days prior to the upgrade.
- Average upgrades/day: up 13%
- Average Plus upgrades/day: up 33%
- Average $ value increase per upgrade: up 8%
We’re thrilled with these numbers. We’ve moved the new design to the Highrise account chart as well. We’ll watch and see if we see the same improvements with Highrise as we have with Basecamp.
Justin Jackson
on 30 Jun 09I’ve followed this “Basecamp account screen redesign” series with interest; neat to see everything from the design process straight through to how these decisions created a better experience for the user, and benefitted the bottom line.
Tim
on 30 Jun 09I wonder if the upgrades were a result of the new design or simply because of the blog post.
Eric Anderson
on 30 Jun 09It’s well known people like novelty. After 90 days of the new design if your conversion rates are a success, you’ll actually know
Carl
on 30 Jun 09What is the % upgrade increase compared to the overall new accounts increase?
Overall, the economy is improving and maybe thats making more projects necessary (meaning: people just need more projects?)
Travis L
on 30 Jun 09@Eric Anderson
That said, for most people it’s likely that the upgrade screen is going to be a novelty to almost anyone who visits. I’d imagine most users who visit that page only do so once per year, at most. So, to the users who will see the page, it’s always going to be novel.
However, I would love to see 37 signals revisit this again in a few months.
JF
on 30 Jun 09Overall, the economy is improving and maybe thats making more projects necessary
I don’t think it’s the economy. The numbers from 14 days before and 14 days after show roughly the same growth as 90 days before vs 14 days after.
Killian
on 30 Jun 09Love these recent types of posts- a little design backed up with a little metrics. Please keep em coming….
Drew McKinney
on 30 Jun 09There are very few studies exploring the ROI of design including a soup-to-nuts view of the project, and metrics to support the change. Thanks!
Jeremy
on 01 Jul 09Seems to me that it’s a bit of a jump to say the design is the major factor. Did you run an A/B test to measure the improvement?
rick
on 01 Jul 09That’s ridiculous. How fun is it to tweak things and then see the results?
Thanks for sharing the info.
Brian Armstrong
on 01 Jul 09A real A/B test would have been better here, there are too many variables to call a percentage like that….although its a lot more work so I understand why you may have wanted to skip it. Probably safe to say it was an improvement overall.
Adam
on 01 Jul 09Many companies who host online services push their UI designers to justify their design decisions with A/B analysis. Jakob Nielsen recommends this in the glowing blog post you wrote about him. But it looks like you went for the “opinionated software” approach instead.
I like what you’ve done and I believe that a few insights from a handful of talented people is worth far more than all the A/B analysis in the world. But opinionated design creates political problems.
Jason: how do you justify opinionated visions to skeptical stakeholders? Whose opinion wins? Or maybe put into context, what if one of your employees believes strongly in something but you disagree. Do you settle these disputes through procedure and analysis or a battle of wills?
mikemike
on 01 Jul 09@Tim
Hahahah, that makes sense. I wouldn’t have thought it, but that’s probably where the success came from.
Harlan Lewis
on 03 Jul 09@Tim
That’s exactly what I was thinking. I don’t doubt it’s better, but I’d be pretty surprised if the new rate of conversions holds at the level it’s currently at.
This discussion is closed.