14 Dec Making Sure Your Metrics are Actionable
Building an effective dashboard means keeping it simple. You don’t want to spend a few minutes trying to figure out what the vast array of charts and graphs mean. So you need to limit the KPIs and metrics that you show to only those that show the health of your business and how the work you do affects it. That means finding actionable metrics.
Actionable metrics are those that reliably change depending on specific and repeatable steps that you take. These measure how effective you are and directly contribute to your bottom line.
The opposite is vanity metrics. Vanity metrics often sound good on paper, but may not actually contribute to the health of your business or change when you take action. Things like website visitors, newsletter subscribers, even total number of customers can mask what’s going on with your company. Use these numbers in your marketing, sure. They sound impressive. But you might not want to base major decisions on them.
Here’s a few things you can do to identify actionable metrics.
Break Down Each Metric
For a KPI to give you information that you can act on, the more specific the data it gives, the better. Your top-level metric – revenue, for example – will tell you how your business is doing. But it won’t tell you why your business is performing that way. For that, you’ll need to break down that KPI into specifics.
Suppose you run a hotel. Your revenue is going to be linked to your occupancy rate. You notice that your occupancy rate overall has fallen about 10% from the previous year. With that information, what can you do? Not much. But when you break it down to more specific details, you get actionable intelligence. Check the class of room – did single bed occupancy change more than doubles? Is the age or national origin of your guests different? Are stays getting shorter or longer?
Now you’ve got data you can affect, because by breaking down the factors involved in your KPIs, you can now target your marketing, change your room offerings, or offer discounts on multiple day bookings. With your business, you’ll need to do the same. Monitor the big picture, sure, but you’ll need to be able to drill down to the details in order to act on that picture.
Know Your Goals
As a manager, you’ve got data streams coming at you from every direction. But you’ve got to make sure all that data corresponds to your business goals. As Eric Reis says, make sure your metrics measure the macro effect of an action, not the micro effect. This can be metrics like how many people downloaded an eBook with a signup form instead of how many visitors the page had. Those visits aren’t the end goal of your email, so why track them?
Dave McClure came up with a simple set of macro metrics he calls metrics for pirates:
- Acquisition – How’d your customer find you?
- Activation – Did they have a good experience?
- Retention – Did they stick around or come back after finding you?
- Revenue – Did they buy your product or service?
- Referral – Did they tell their friends?
It’s for pirates because the acronym is AARRR. If you can tie a KPI to each of these items, you’ll have a pretty good picture of your sales funnel. Let’s take a look at how that might work.
Suppose you have a website that sells fancy electric widgets. Your acquisition metric might be the number of visitors. Activation may be based around bounce rate, newsletter signups, or brochure downloads. Retention could be page counts or return visits. Revenue is easy – sales. Finally, referrals could be social media shares or reviews posted.
The specific metrics that reflect your business will differ from the example above. And depending on how you measure each metric, they might not be telling you information you can act on.
Group Your Numbers by Cohort
Cumulative numbers won’t be very useful. They’ll always go up and won’t reveal whether something you did had much effect. Aggregate numbers – the product of multiple individual metrics – can deceive you as well. Your sales funnel could take a long time, and an aggregate metric – customers who converted to sales – won’t show the effect of your actions taken within the time frame of that sales process.
In order to measure how your actions affect user behavior, divide the users into some sort of cohort. Time-based cohorts are the simplest to use. Your business actions take place within a time stream, so the effect of those actions – customer behavior – should, too. Whether you organize them by day, week, month, or year is up to you, but each top-level metric should be time-linked. That way, you can tell if that ad you put on the side of a bus actually sold more cars, or if it just increased the number of people who came into the dealership and wasted a salesperson’s time.
Another useful cohort involves user behavior, like users who viewed a specific product or talked to a specific salesperson. You can directly set up user behavior cohorts with A/B testing. You can have salespeople try different pitch styles on potential customers during any given day and see which gets the better response. The same idea works for a lot of either/or changes: email subject line, length between lead acquisition and first call, even little things like fonts and product colors.
A cohort can be any shared quality of a group of your customers and potential customers. Age, gender, hair color, whatever. Using cohorts and breakdowns, you can refine the data that you have into detailed snapshots that show not just the symptoms of business issues, but their root causes.
Data can make your business decisions easier. But only when that data provides a clear response to actions that you take. Finding which metrics and KPIs are actionable may take a little research, but it’ll improve the health of your business.