David Ogilvy, the father of modern advertising, once said:
“Most people use analytics the way a drunk uses a lamppost, for support rather than illumination.”
His point is that a lot of companies use data after-the-fact. They have an idea, then look for data to confirm it. They twist whatever metrics they need to verify their own opinions.
But data should be used the other way – to illuminate the state of your startup and show you how to create change. So rather than throwing a bunch of metrics on a dashboard you think are useful, it’s your job to find the ones that create change.“Most people use analytics the way a drunk uses a lamppost, for support rather than illumination.” - David Ogilvy Click To Tweet
In this article, we’re going to talk about the limitations of KPIs and find the metrics that matter.
- Ask Questions Before Measuring
- KPIs vs. Growth Levers
- Work Backwards to Find Growth Levers
- The Two Sides to Growth Levers
- Lean on Your Metrics
Ask Questions Before Measuring
A lot of startups make the mistake of measuring before they know what kind of information they need. Sure, having lots of data is helpful because you never know when it will be useful, but you can’t focus on everything.
Plus, every startup (and teams within startups or bigger organizations) need to track different metrics. The company’s KPI may be monthly recurring revenue, but the marketing team might use leads as their primary metric. Customer success cares about the churn rate and your support team measures how fast they close tickets.
This means you have to ask lots of questions to figure out what’s important to your company. For instance, do the number of likes on your Facebook page really matter? No, probably not (unless your model requires a big reach on Facebook), so there’s no point giving it a moment of your attention.
You probably have a good handle on the metrics that illuminate the state of your business; the numbers you compare week-over-week or month-over-month to gauge your progress.
But while those numbers give you a picture of your startup now, they don’t tell you how to affect change.
KPIs vs. Growth Levers
Key performance indicators (KPIs) are after-the-fact metrics. They describe what something is like or what happened. But in most cases, they don’t give you any information as to how to drive them up or down (depending on which way you want the metric to go).
So while KPIs are important because they let you track your startup’s progress, you also need to understand your growth levers: Actionable metrics you can push.
For instance, every software startup wants users. So “total users” or “daily active users” are important metrics to track. They let you plot your startup on a graph to track growth. When you seek funding from VCs, this is one of the first metrics you should report.
Let’s say you have 1,200 daily active users. Obviously you want more. How do you make that number go up?
Unfortunately, the number doesn’t give you any information as to how to affect it. There are a lot of tactics you could use to get more users. Some of them will work for you. Others won’t. In fact, there are even tactics that should work for you but won’t because you can’t do them well or don’t have the right resources.
So basically, “daily active users” only tells you how many you’ve managed to acquire. It doesn’t tell you how to get any more.
KPIs are sometimes called lagging metrics because it takes time for them to reflect changes to your startup. They can hide problems until it’s too late to fix them.
For instance, if Dropbox only tracked a KPI like churn rate, they wouldn’t know if a user was unhappy with the app until he failed to renew his account. Hopefully Dropbox tracks other internal metrics that signal a customer doesn’t enjoy the product so someone from the company can intervene to save the account.
To discover how to affect your KPIs, you have to find the growth lever that influences it.
Work Backwards to Find Growth Levers
To find your growth levers, start with your KPI and work backwards. Keep asking yourself “How do we do that?” until you reach something simple and actionable.
Brian Balfour uses Spotify as a great hypothetical. (He refers to KPIs as output metrics and growth levers as input metrics. The terminology is irrelevant as long as you understand that one type affects the other.)
Naturally, Spotify wants users to spend more time in the app listening to songs. An important KPI for them would be “length of time someone spends listening to music.”
Coming up with ways to get people to listen longer isn’t practical. There are a million things you could do. You would have to test countless tactics until you randomly fell upon the one that worked.
Instead, let’s work backwards to identify the growth lever.
Spotify could rightly assume that in order to increase the amount of time someone listens to music, they would have to a) Get them to spend more time in the app, and/or b) Bring them back into the app more often.
Now we’re getting somewhere. Those are more actionable. But we can go deeper to identify the growth levers.
Next we would look at those two activities – bring users back and increase time per session – to find ways to push those metrics up.
For instance, Spotify might decide they can bring more people back to the app by notifying users about new artists and pushing song recommendations through notifications. To increase the time each user spends in the app, they might curate playlists for users or help them discover new songs.
In the case of our hypothetical example, Spotify’s growth levers would be:
- Curate quality playlists and match them with relevant users.
- Notify users when new artists they might like appear on the app.
- Notify users with song recommendations personalized for them.
- Expose users to artists and songs they would like, but haven’t seen before.
Those are far more actionable than vague goals like “increase daily active users.” They’re tangible and real. Spotify can build strategies around them.
The Two Sides to Growth Levers
Once you find your growth levers, you’ll probably sigh in relief. Growth levers give you direction. They tell you what to do to raise those KPIs and ultimately grow your startup.
But there’s a good chance your growth levers aren’t one-dimensional. They’ll probably create several types of change within your startup.
For example, Facebook surely uses their ad revenue as a KPI. It follows that if they show more ads (their growth lever), they’ll earn more revenue.
But showing more ads might erode the user experience by reducing engagement, time on the site, and click-through rates. If those metrics fall (meaning there’s fewer people clicking on ads), the KPI could fall, even though Facebook pushed their growth lever.
Presumably, Facebook tests this tradeoff to find the right spot that earns them the most revenue. You’ll have to test your growth levers as well to determine how hard you can push them and if there’s anything you have to do to compensate for their symptoms.
Lean on Your Metrics
Once you determine the metrics that create the right kind of change for your startup (your growth levers), lean on them hard. Continue to test them regularly to validate your assumptions while you pour resources into them to stimulate growth.
Instead of obsessing over KPIs that only record what you’ve done, look to your growth levers – the metrics that matter – to find ways to create positive change.