This article is an abstract from a talk held on 12th of November at the World-Usability conference in Berlin.
The last month I often asked myself, do we do enough enough to evaluate our ideas in front of "real users"? Do we talk enough to those people who are actually using blogfoster or do we implement our own assumptions?
User-Research in Startups
The honest answer was frustrating: We don't do anything (or really less) of what you can describe as user-research from a scientific point of view. Nevertheless, our product works for a lot of users day by day and we get some great feedback. So at some point, we did the right thing (what doesn´t mean that we can do better) and I tried to figure out what and how. I took a look on how we communicate with users besides the regular conversations via email and it turned out, that there is a lot, what you can describe as user-research.
Looking for User-Research? Look twice!
For example, I am disciplining myself to have one call with a user a week. It´s not that hard, because we still are a small team and so I often answer the phone by myself. It´s important just to not only answer the question of the user (which often show you the real challenges users have with your product), but also to ask specific question on how they are dealing with your product day by day. Its also often really interesting to ask how they discovered your service and how they are using it on a daily basis. They are often very keen on sharing their view on your product.
I talked to other founders and product guys and surprisingly they made a quite similiar experience, like I did. When you have graduated from university (Slide X), you are full of motivation to build up an awesome, design and user-research workflow. But the reality smashes these resolutions quite fast. User research needs a lot of time and human (sometimes even financial) resources to prepare, operate and evaluate. Things which are mostly known as sparse thing in startups.
I also experienced in several projects that user research is always a thing which brings you out of the process, if not 100% integrated in the workflow. And as mentioned above, startups often lack to integrate it properly. So I asked myself, if there isn´t a better way to do user research. Not like a thing which you do from time to time with much effort, more like a steady flow which brings you results on a daily basis.
There are several ways and tools to collect user-feedback on a regular basis. Here are two of them, which we use at blogfoster:
Surveys are a great way to collect user-feedback. Most times users are really keen on giving feedback, especially if its happening anonymously. You can trigger feedback via questions you ask in the survey (eg.: What is your most favorite part ouf our product) but we also got great results when asking for feedback and optimizations by the end of the survey. The trick behind those surveys are, that they don´t take so much time and the user is asked for participation in various situations. So for example, linking them in email signatures or in the application. Also take the chance and present results "back" to your users. Most times, they are also interested in results and are more willing to take part in further surveys.
Support requests are a great source of user feedback - if you manage them in the right way. Often they provide a lot of information how your users are using your product without asking them. The important part is, that the request not just gets solved, but that you write a short summary about the concern of the user (for example like in the screenshot below). More important: Tag requests with keywords like "dashboard" or "settings" when the request relates to one of these views. That makes it possible to find them again if you have questions concerning how users are dealing with a special view.
Data & Analytics
Measure every action that an user performs in your application. There are great tools for doing so, like Mixpanel, Heap, Google Analytics, Go squared, Kissmetrics etc. Track every click, every interaction the user performs, but also take care about the "actions in between": How long are users staying on a view? How long does it take for them to fulfill a task? Are they going those steps you want them to do or are they following different paths? The nice thing about data analytics is, that it does user research without asking - user research by numbers.
User Research in Advance: Have Answers Before Questions Arise!
Those three actions are just a few ones but can help you to achieve this goal: You have answers ready when questions towards user behavior are coming up. You can answer questions out of tickets, support requests and surveys, like: What is the most painful feature for our users on this view? What wishes do our users have regarding the signup process? Are they using feature XYZ and if yes, how frequently and how long? This sets you in the position to have user research in advance. You don´t have to reach out to users when you have questions regarding the use of your product. You just just "ask" the collected data and feedback.
Of course: Not every question can be answered with that procedure. But nevertheless, if you have specific questions towards users, you can pick those, which you can select as heavy (or non-heavy) users of this feature. You don´t have to start at 0. Its also hard to predict success or failure of new features from historical data and feedback. However this approach saves you a lot of time and ressources which is crucial for startups which move fast and can´t afford longterm user-research in the beginning.
Two more things
There are two more things which we want to share.
Use your hackers. Hackers are power users which are mostly tech-savy and handle your product in ways, which are different to other users. Talk to those users first. It appears that they often have the most valuable feedback for you and have ideas, which can drive innovation. These people are most lucrative for user research.
Share user feedback in the company: We made good expierences with running two Slack-Channels, where everyone in the company has access. The first channel shows every support request automatically and unfiltered - you get a broad overview of the actual mood in your community. The second channel is for pasting and highlighting feedback which is worth sharing and which came along different channels like emails, social and talks. Its also important to present results from the surveys in front of your team to spread a broad understanding on how your users see your product. Always keep in mind: Your product is nothing without users. Your product is not what you think it is - it´s what your user think about it.