Information Technology Dark Side

Like a lot of Software-as-a-Service products, TroopTrack gives potential users a chance to try it out for free for thirty days. After 30 days, if the customer likes, they buy a subscription.

TroopTrack conversion rates vary from month to month, but at a minimum, 10% of my trial customers convert to paying customers. I’ve been trying to figure out ways to improve that number over the years and some of my experiments have yielded fruit. For example, if I am able to reach a trial customer by phone early in the trial they are dramatically more likely to convert than those I never talk to. The conversion rate (anecdotally) is upwards of 70%. Basically, almost everyone I talk to ends up buying. They don’t forget the one product they tried where the owner actually called them.

I’m very interested in figuring out how to increase the likelihood a customer will convert, but I’m also very interested in predicting whether they will. To this end, I’ve been gradually developing a model for measuring their engagement.

It all started with login counts
A couple of years ago I started displaying how many times the person who signed up for an account has logged in. This was immediately insightful – at that time 90% of my trial customers were logging in once and then never logging in again. Of the 10% who were left, most of them bought a subscription.

This intrigued me and I started calling customers, as I mentioned above. I was really trying to figure out why they never logged back in after signing up, but I got distracted by the effect calling customers had on their purchasing decision and stopped asking about why they hadn’t logged back in. My logic then was simple – calling them had created an incentive to look more thoughtfully at TroopTrack so why bother figuring out why they’d given up the first time?

Engagement is more than just how many times the account owner logs in
Since that discovery way back when I’ve watched login counts quite carefully, if not formally. If the person who signed up for TroopTrack (i.e. the account owner) logs in a second time, I have a dramatically better shot at signing them up. I began to wonder what else I could look at to figure out how engaged my customers are during their trial period. So I added another metric – the total number of users in a troop and the number of times they had logged in.

New observations emerge
The user count helped me understand something new about my customer’s behavior. Most trial accounts that don’t convert only create one other user – a scout. It makes sense to do this – they want to see what info we track for a scout right away so they go there first. That seems to be the general first milestone: Add a Scout.

The next milestone I’ve observed on the path to purchase is similar: the account owner adds another adult user and invites them to check out TroopTrack. If this person logs in more than once then things look good. How good? No clue yet.

There’s one more milestone that I can currently see that is important: loading users en masse. At some point in the trial a troop adds more scouts and adults and invites them all to check out TroopTrack. At this point they have pretty much already made a decision to purchase and they are just waiting for the trial to expire and for me to ask them to pay up.

The Engagement Metric
This morning I decided to add a new metric to my admin console. I call it the activity score. There are a lot of things a troop can do during their trial period, and I’ve decided to give them points for each of them. The points aren’t weighted (yet) but they may be later on. Basically, I assign an activity point every time a troop member does something I care about, like log in, add a user, record an award, upload a photo, plan an event, etc. I aggregate these points in my admin console and I plan to keep a record of the number of points a troop has on the last day of their trial period as well as the outcome of their trial (purchased, didn’t purchased). It will be interesting to see what I learn from this.

I’m also thinking of adding a features used / not used list for each trial account. That will help me know what to talk about when I give a troop leader a call and offer to help. If a troop isn’t using event management, I can talk with them about that. They might be having trouble figuring out the features they aren’t using and I can help them out with it. I hope that will help.

PayPal sends me more emails than anyone else except pivotal tracker, and that’s awesome
Right now I get an email from PayPal almost every day. Yeah, you got it – someone sends me money almost every single day. It’s pretty cool. It’s not a lot of money, from $5 – $200, with the average being significantly closer to $5 than the latter.

I’m not entirely sure how relevant that little aside is to my rambling about engagement, but it’s important to me. After four years of plugging away on TroopTrack, I still believe, and the evidence that I should believe gets a little better every day. Just sayin’.

Stumble it!