Lesson: Consumer Growth with Jim Scheinman

Step #6 Testing: When to kill a feature

I would say that if you're an early stage consumer startup you have to be very aggressive on your customer growth strategy and so yes you're going to test out things that are not sustainable, again never crossing over into the gray area where it's going to be offensive or illegal to do anything. But you need to be aggressive to figure out how you're going to get traction and do you have the right product market fit? You should be testing daily if not weekly, weekly if not daily, on your different strategies, but you may end up with things that are completely a waste of time and that's fine. You kill it and you move on.

There really isn't a magic number to this. I think that as an early stage startup you have to be really diligent in coming out with new products and features on a regular basis. When we were building Bebo we would come out with a new minor feature every week and a new major feature every other week and it was constantly iterating on the product.

I'd say sometimes we might have been overly aggressive on iterating on the product, but I'd rather be overly aggressive and move quickly than be too risk averse and move slowly. There was a very interesting lesson learned. We had a very talented Engineer and really an early growth hacker at Bebo who came up with a product feature called "Share the Love." It was really unique at the time, where you would get a heart to give out every day on Bebo, once a day. If you didn't give it out it would accumulate and we thought this thing would be great, people would love this. It was almost a precursor to the "like" that you see on Instagram and other companies, and it didn't take off.

For whatever reason in the beginning it just didn't take off and it was on the slate to be killed and move on. For whatever reason we either got side tracked to do something else but we didn't kill it that week and it went on for the next week and it took off. So we're not sure if it hit an inflection point. You never really know sometimes exactly why it takes off, but you can see that it definitely hit that point and it became one of the most popular features driving a lot of viral growth on this site.

We're very lucky to learn a lesson that sometimes you might be a little bit too quick to kill something. I'd say that if it's not working within a month it's probably not going to work, especially if you have a decent amount of customers and engaged customers you can tell if something's going to work pretty quickly.