Stop Wasting Time on the Wrong Customer Acquisition Metrics
Metrics have a strange, somewhat inexplicable allure over SaaS providers. It’s as though the number patterns hold a mysterious, elusive cipher to success. Maybe if we eyeball enough graphs, look at the customer usage stats through enough segments and rummage through enough site flows and clickthrough rates, we will be able to unravel that million-dollar metric-molecule that will tell us how to give our users exactly what they want. It’s a question of covering as much data ground as possible, right?
The endless delving into data tunnels is exactly that: endless. The stream of activity stats, customer behavior segments and lists of usage peaks and lows is an addictive distraction. I should disclaim: A lot of those addictive customer acquisition metrics are useful. Micro-conversion goals like element clicks can tell you which calls to action convert. Bounce rates will tell you which pages suffer from bad marketing. A click-flow from a certain page to the knowledgebase will tell a tale of woe, bugs and improvable navigation.
But none of these metrics paint a clear picture of overall customer engagement with your software. Or rather – they ALL do, but it really doesn’t make sense to mash all of them together and try to salvage intelligible, actionable information from them. Sure, you could extract an average of clickthroughs for a certain element, segment which users are below/above it, repeat for a group of elements you have distinguished as key elements and then calculate the level of usage for each user as per the average of those stats, but, well, you wouldn’t have time to do anything else to promote your SaaS. Ever.
The thing about good customer metrics is that they have to reflect one thing, and one thing only (are you writing this down?):
The value the customer derives from your service.
That’s it. Anything else is either a distraction or else vanity metrics. The question that should be guiding SaaS providers in constructing their customer value metrics should simply be: what numbers would tell me if my customers are getting what they came to my service for?
These parameters differ from one service to another: the customer engagement stats for an email marketing service would be calculated monthly, while a CRM provider would measure engagement by the day. You would also need to take into account an onboarding curve that would also vary from service to service. The onboarding curve for Facebook, as it were, is a slow climb (it takes a while to gather enough Friends for the Newsfeed to be of any actual interest) that settles on a steady plateau of constant, hyper-engagement, whereas WordPress may be a steep study that will result in consistent yet in-frequent visits.
For us, being a B2B SaaS, the onboarding period was relatively lengthy and the usage we needed to measure was both for our customers and for our customers’ customers:
- Last Login – when was the last time a customer had logged in?
- Days active in the last week
- Average session time
- In-app messages created
- Was Iridize installed on the customer’s website?
- End-user activity – how many users had used how many in-app messages?
These parameters gave us a pretty clear idea of individual customers’ engagement score. In fact, we started with even more parameters, but gradually narrowed them down to the bare necessities, getting rid of anything that slowed down the calculations. We tested the metrics time and again, every time we lowered the cutoff point – and finally reached the sweet spot where the numbers reflected overall customer mood.
This was a while ago. We have since perfected our calculation methods and don’t you know it, today we actually have an algorithm to do it for us. But these remain the solid six metrics that we rely on to tell us how riveted a customer is to our service. And you should, too.