Kevin William David from Siftery.com interviewed me to talk about Dovetail, how we’re different, and some of our challenges. This post originally appeared on siftery.com and is re-published here with their permission.
Dovetail is where you store, organise, and analyse user research. Think user testing notes, customer feedback, interview transcripts, diary study responses, NPS feedback, and so on. It helps researchers and product managers discover and justify what to build next.
I used to work as a designer on the growth team at Atlassian. While numbers are great at telling you what someone did, to understand why they did it, you have to get into qualitative research. Basically, you need to talk to people.
There are lots of products to help you collect and analyse quantitative data — Google Analytics, Mixpanel, Hotjar, Optimizely, etc, but there aren’t many that do the same for qualitative data like text, images, and video.
I’ve learned that a powerful quote from a user is much more effective at changing people’s minds (and roadmaps) than anonymous statistics. However the tooling just isn’t there to help researchers and product managers identify trends across lots of customer feedback.
I actually answered a Quora question on this one.
On one end of the spectrum there are hardcore ‘Computer-Assisted Qualitative Data Analysis Software’ like Atlas.ti, NVivo, and MAXQDA. These tools are super expensive (thousands of dollars per user), aren’t cloud-based or collaborative, and are horrendously complex and unintuitive since they’ve built up severe feature bloat over 10–15 years.
On the other end of the spectrum, there are highly flexible tools like spreadsheets and sticky notes. The flexibility makes them great for lots of use cases, but you have to bend them to fit your specific one. Dovetail is designed from the ground-up for finding patterns in customer research data, so our terminology and concepts are all built for that.
Basically, we’re trying to fit in the middle somewhere. We want to build something that’s purpose-built for a researcher’s ‘job to be done’, without losing too much of the flexibility that spreadsheets and sticky notes give you.
Our customers tend to fall into three categories: academic, corporate, and design agencies.
Academic researchers primarily use Dovetail to analyze large amounts of textual data — e.g. other research papers, tweets. User researchers and product managers are looking for trends to help prioritize their roadmap, so they use Dovetail to store and analyze customer feedback, user testing notes, NPS feedback, and interview transcripts. And the last bunch, design agencies, do much the same when they’re consulting for companies.
A few of the names you’d recognize would be Cisco, Freshworks, Harvard University, IBM, MacMillan Learning and Pinterest.
Posting on Quora has driven a lot of high quality traffic to our site. Basically, I searched for questions that are relevant to us (e.g. “What are some good tools for analyzing survey responses?”) and posted an answer that started with a disclaimer like “I’m the founder of this product, and it’s built to do exactly what you want…”
The nice thing about Quora is that the questions live forever and they have pretty good SEO. So a fairly consistent acquisition path for us has started with a Google search, then a visit to a Quora question, see my answer, visit our website, and then sign up.
A few big challenges.
I think we underestimated how difficult it would be to decide on a pricing model. The thing with pricing is there’s no perfect option — it’s just a bunch of tradeoffs. Whatever dimension you choose to charge on means you’re effectively increasing friction there.
The feature set of Dovetail 1 was quite different to what we have now, and it was really optimised for once-off projects. However I knew we were eventually going to build features that encouraged recurring use, so we added a subscription pricing model with usage-based tiers. The only problem was that people would cancel after their project had finished. Also, being restricted on usage isn’t great because we want people to use the product more, not less!
With Dovetail 2, we really want to push the collaborative angle since that’s a differentiator for us. So, to completely remove the friction to invite more teammates, we moved to a fixed price point for unlimited usage and users (similar to Basecamp. This makes us a lot cheaper for medium–large teams, but quite expensive for single users.
Another challenge we’ve had has been succinctly communicating our value and working out what the ‘aha!’ moment is. “Qualitative data analysis” is not an easy thing to understand although it’s something people do naturally every day when they look for patterns in things. The challenge is finding out how to communicate that in our marketing, but also how we can show that to people early on during their trial.
The last big challenge we’ve had recently has been getting an “MVP” out which is valuable enough to satisfy changing user expectations. In 2017, people expect web apps to have an amazing visual design, excellent performance and stability, along with table-stakes features like notifications, search, mobile support, collaborative editing, real-time comments, and so on. The product feels bare when you don’t have these. But it’s tough to build all of that from scratch as a two-person, bootstrapped startup.
We don’t have any integrations right now, although we’re currently working on Zapier support which will unlock 750+ apps to import data from. There are tonnes of awesome use cases for Zapier integrations — pulling in survey responses from SurveyMonkey, tweets from Twitter, saving articles on Pocket and importing them into Dovetail for analysis, an email integration so the support team can forward customer feedback, etc.
We did add one-click Google sign up early on which has worked better than I expected. About 25% of our users choose to use Google when they sign up for an account.
We use Google Apps for email and calendar, Mixpanel for analytics, Trello for feature prioritisation and roadmap, Bitbucket for code hosting, CircleCI for builds, Stripe for payments, and of course Dovetail for storing and analysing our user feedback. Apart from Dovetail, out of those my favourite would be Mixpanel. It’s just so much easier to use than Google Analytics, although they’re going through some big product changes right now so a lot of their web app is inconsistent. Looking forward to what they release over the coming months.