Trovz.com taught me that it’s important to get some core users early. But how do you get them, and how do you make use of them? First things first: what does a core group of early users do for you?
First, users will tell you what’s wrong with your site. (Or if they don’t, they’ll just leave, and that should tell you something.) Without real users, you’re guessing what your eventual users will want. In my case, I was building the site for myself, so I assumed that the eventual users would want what I wanted. But even if that was the case, you can only work on so many things at a time — how to prioritize? Having users you can poll sure would have been helpful.
Second, the notion of compound interest can apply to users. If your users mention the site to others, some of them will be interested. Of course, you can help your users do this by making it easy to put stuff in their Facebook activity stream, or other ways. Those new users will mention the site to others. Nothing new here, just the basic viral marketing stuff. The key point is that, like saving for retirement, the earlier you start, the sooner the base can grow, and the more compounding can help you.
Third, users validate your idea. If you have trouble getting people to use it, that should set off warning bells.
So what will I do differently next time around? Again, nothing earth-shattering, but here it is:
Define the core of the application
What is the minimal functionality that makes it useful? With Trovz, I got caught up with some things early on that weren’t core. For instance, we had the idea that people might want to make different profiles visible to different communities. If someone from a work community wanted to borrow something, they should probably get your work email or phone. But for one of your neighbors, your home phone makes more sense. Nice idea, but not core. But because we did some work on it relatively early, we had to maintain a number of database tables that weren’t really that helpful.
Make the core work right
You don’t need all the bells and whistles, but what you build needs to work without glaring bugs, and efficiently enough for small membership. When I find a site that looks interesting, I’m willing to forgive certain things not working perfectly, but if bugs prevent me from doing what I want to do, I won’t use it.
Make the user experience good
I talked about this a bit in a previous post. Your site doesn’t have to be a work of art, but it should be usable and at least not unattractive. Between Trovz.com (where I went through multiple basic layouts) and work at my full-time job (where we have people whose job is specifically to draw up the views before they are implemented), I’ve seen the value of getting to work from wireframes. Pay attention to the user experience from the beginning, and at least sketch it out before you start coding.
Start targeted marketing
While you’re working on building a solid core, think about who you’ll try to get to use it early on. You probably know some friends and family who will be interested. One benefit is that they will be patient with your development — your friends and family want you to succeed, and that’s a good thing. But you should go beyond them and look for some total strangers. If you can get some of them interested and using your system, you’re doing something right.
For Trovz, I collected a list of groups that I thought would benefit. The groups included local community theaters, Boy Scout troops, and “cul de sac communes” that I heard about across the country in L.A. We never did get the core to the point where we wanted to contact them, but my intention had been to 1) find at least one individual to whom I had some connection (hello, LinkedIn); 2) tell that person how the site could help the group specifically; 3) point out that the site was in progress, and mention that early users would influence how the site got built. The idea of this last point is that by being involved early, a group can shape the site to make it more helpful to them.
Have the database record when each user was last seen
Consider these two metrics: registered users and users active in the last n months. Which one is more informative? You can get to the latter by updating a last_seen” field on your user table during login. I actually started off with this feature, then lost it while re-architecting. That was pretty silly, because I’d already seen how useful it was.
Actively solicit feedback
Maintain some kind of regular contact. I was bad about this. Even though we did set up a small group of beta testers, I only contacted them sporadically and only ever heard back from a few of them. Next time, I’ll think I’ll pick some regular schedule for contact. Not so often as to be annoying, but often enough to let them know I’m still on it and, importantly, that I’m making use of the feedback I’ve been given so far.
While working on Trovz.com, I didn’t really do the defining step, and I had a lot of trouble with the user experience part, which led me to neglect marketing. And even though I talked to lots of people who thought the idea was great, I failed badly in converting them into actual, regular users.
Lesson learned: get early users who can validate your idea and guide your development plans