KeyPerformanceIndicator
KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs They are everywhere KPIs KPIs KPIs KPIs KPIs KPIs KPIs in tech KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs
KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs and venture capital, KPIs KPIs KPIs KPIsKPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs KPIs but what do they really KPIs KPIs KPIs KPIs KPIs KPIs KPIsKPIs KPIs KPIs KPIs KPIs KPIs KPIs measure? KPIs KPIs
What’s in your KPIs?
If somebody asks you about the key performance indicators (KPIs) of your business, what do think about? Revenue? Signups? Visitors? Downloads? Sure, all of these are important for your business. But if you really want to have the complete picture on how your startup is ‘performing’, shouldn’t these numbers include an indicator for actual usage and therefore value your users are deriving from your product? I believe so! After all, the first goal of your startup should be to create value for your customers. And everything else should* follow.
In a recent rant about vanity metrics, Ev Williams pointed out that the key metric at Medium is Total Time Reading and he makes a great case for why that is a better estimate for value they create than page views, unique visits and other vanity metrics. I believe that this kind of key usage indicator (KUI) should be included in every set of KPIs.
*This is not to say that you should focus only on usage and neglect monetisation, although that can be a valid path. It’s just to stress that especially early on you should optimize for customer happiness and not for extracting value (i.e sales)
How to find your KUI
Your KUI is easy to find it if you think about the atomic user journey, which is a step up from what Fred Wilson calls the atomic unit. Which use case describes how the majority of users derive value from your product?
Granted, this question is easier to answer for early stage consumer startups than a fully-blown SaaS solution for instance. But from my experience, there is often a 80/20 distribution of feature usage in more mature products as well. So if you are later down the road, focus on that 80%. Let’s look at a few examples for KUIs from verticals we are most interested in at PNC.
Mobile
SaaS
- Contactually: Follow-ups sent
- Typeform: Surveys created
- Algolia: Successful searches
Marketplaces / Networks
We could look at many more companies, but I think you get the idea. ☺
Completing the picture
Once you have defined your KUI, you can start tracking it over time, alongside the rest of your KPIs. If you do it on both — an aggregated and (active) account/user level — you should be able to optimise around it and hopefully increase it over time.
Of course there are always some short-comings when looking at aggregated numbers and averages. So if you want to have a complete picture, you should always complement your KPIs (incl. KUI) with a cohort analysis and other relevant data sets for your particular business model.
It’s also important to keep in mind that not all users are equal. As Angela of Version One has stressed before, there are different levels of engagement for social networks for instance. Still, I do believe that the KUI can be a great foundation for that kind of user segmentation and a proper understanding of your product’s activation path. But that is something for a future post… or Josh Elman.
This is still thinking in progress and I realise that the idea of KUIs applies more to startups looking for product / market fit, than those that have reached initial scale. However, I do believe that paying more attention to usage related indicators can provide additional insight across all stages.
