Recently there have been a few occasions at Buffer where we’ve hesitated about next steps or thought about spending longer on certain tasks. As a result of my thinking around this, I’ve started to believe that the following question can be one of the most powerful questions for startup founders to say to each other:
"What can we do right now?"
"Right now" means faster validation
The reason I think this question is so powerful is largely based on the core nature of startups. Unfortunately for us startup founders, the key difference between what we do when we’re building a scalable startup and what you would be doing if you’re running a service business is that a large amount of the work we do every day is building things which are not yet validated.
There are three common scenarios in a startup where we’re handling assumptions which need to be validated as soon as possible:
- we haven’t launched yet and so almost nothing is validated
- we’re modifying the experience of a current feature
- we’re adding a new feature we expect will improve one or more metrics
Therefore, a lot of what we spend our time building hasn’t yet been seen by our potential or existing audience. This is a key issue that many founders agree is vital to address:
"Usage is like oxygen for ideas. You can never fully anticipate how an audience is going to react to something youve created until its out there. That means every moment youre working on something without it being in the public its actually dying, deprived of the oxygen of the real world." - Matt Mullenweg, WordPress
Two examples from Buffer
Getting a user on Skype for user testing
We’ve recently been working on an overhaul of the browser extension popup that you see when you use Buffer to add a new Tweet or Facebook update about a page you want to share. Initially, we thought about launching it to everyone without getting feedback. However, there are some new concepts which are quite different compared to the current version, so we decided it is best if we try and validate our assumptions about how clear the new experience is.
To do this validation, we’ve been connecting with a few users through Skype and enabling the new version for them. As they experience the new version, they share their thoughts and we ask questions to specifically get insight into whether they understand the different parts of the experience.
At first, we said to each other: “maybe we could do some user testing next week”. Then, in the same conversation we moved towards “how about this week”. In the end, we took it even further and said “why not today” and then finished by concluding “let’s Tweet right now and have a call with someone in the next few minutes”. Within 20 minutes, we’d done an interview and learned a massive amount. We’ve since done ten more user interviews and learned even more.
Hacking and hard-coding UI changes
Whilst conducting these user interviews, we’ve learned a massive amount very quickly. Some of what we’ve learned has proven that our assumptions were in fact incorrect. Part of the interface we have built for the new browser extension popup is not being perceived as we expected it to, and users are getting very confused about the purpose of that particular section.
As a result of invalidating this assumption, we’ve realised we need to make some changes in order to improve the clarity and help people “get it” faster. As we started to tweak the interface, we quickly agreed that we’d need to do more user interviews to check whether we’ve achieved the clarity we are aiming for.
When we first started to tweak the UI, we were doing it very thoroughly, checking that everything worked perfectly. After a while, since we knew we had to test the UI through user interviews again, we decided we should try and shortcut the process. We asked “what can we do right now” that would help us to learn whether we’ve improved the clarity. What we’re doing now is hard-coding these types of changes and leaving out large parts. Then we jump on another Skype call. If we find something that works, we can built it out further and add polish.
Validated learning is the measure of progress
The key thing that this process of asking the question “what can we do right now?” reminds me of is the way that Eric Ries defines the measure of progress in a lean startup:
"Progress in manufacturing is measured by the production of high quality goods. The unit of progress for Lean Startups is validated learning-a rigorous method for demonstrating progress when one is embedded in the soil of extreme uncertainty." - Eric Ries
When was the last time you asked your co-founder “what can we do right now?”. I’d love to hear from you in the comments!
Photo credit: christopher charles