Some definitions
Before moving on, let me get back to some definitions so that we use the same terminology.
A metric is something quantifiable and measurable. There are two main types. Flow metrics measure something over a period of time (eg. Monthly revenue, Weekly Active Users). Stock metrics are a quantifiable snapshot (eg Cash on Hand, Accounts Receivable, Equity). Business metrics usually fall in several categories (marketing, sales, finance, product, operations, impact, others).
An indicator is a metric that has business relevance—i.e. gives useful business insights on a specific topic. So, if you track gross margin, but that does not give you any business insights, it’s a metric. If it tells you that there is something awkward since your product/service costs have increased despite having the same number of customers, then it’s an indicator.
A KPI (key performance indicator) is an important indicator for measuring the performance of the company. This means that a KPI needs context first. What metrics measure the performance of the business? Which ones are key for the performance (not all are important, so not all are key). A general rule of the thumb is to keep the number of your KPIs as low as possible. All others are indicators.
I’ll be using metrics in the rest of this article.
The four metric types
These are four types of metrics that are/should be tracked by any business (startups included):
Vanity metrics
Lagging metrics
Leading metrics
Coincident metrics
Vanity metrics
Also known as PR metrics, or bullshit metrics. I’ll start with these ones because they’re the most overused and the most useless of all metrics. Website visits. New signups. Headcount. Jobs created. You track them because everyone does, or because they look good, or because the accelerator you’ve been through asks you to report them. But if you can’t tell me in the next 15 seconds what decisions they drive in your business or team, they’re just shiny stuff, good for bragging only.
One area where Vanity metrics are immediately visible and standing out is social media. Both organic and paid efforts are measured against the effect they have on the channel itself rather than the business. Page Likes, Post Likes, Followers … they give no real insight into how well your business is performing unless you are in the business of “influencers”.
Don’t get me wrong. If you track these metrics to brag or as a PR stunt, please continue to do so if they bring you joy.
Performance metrics
Also known as lagging metrics. They are most frequently tracked by sane startup founders and managers.
They are easy to measure. Revenue, MRR, COGS, Gross margin, Monthly Active Users, Churn, Billings, ACV, and so on. I bet you have a spreadsheet up your sleeve with them, and that you look at it with hope and awe, or with desperation (let’s hope it’s temporary). They are useful, no doubt. They help you identify patterns and understand the long-term consequences of your actions and plans. They are great for tracking outcomes evaluating the success of a strategy that was implemented before.
But they are not enough. They are hard to improve because they depend on a series of success factors. Revenue depends on a lot of factors—content marketing, branding, product value, sales teams, customer decision-making units or procurement, onboarding process, customer satisfaction, churn, and luck (first of all). We may fall prey to our cognitive bias and think that one action can improve these metrics, but they are only useful for retrospectives. They won’t tell the future. They don’t help you make decisions.
You can stare at your revenue charts as much as you want to. It’s not going to tell you about what you need to do tomorrow.
Decision metrics
Most frequently known as leading metrics. They are the ones that measure action and outputs, rather than outcomes. The number of social media posts per day. New customer sign-ups this week. Sales calls. New leads added to CRM. Bugs fixed. Customer tickets closed. They are relevant here and now and help you make decisions. Should we push on getting new leads or closing qualified leads this week? Should we postpone the product release? Which of X options will increase social media engagement next week? They are easy to improve. Sometimes hard to measure, which can become a headache (so don’t track them if their value is less than the effort to track them). They are irrelevant in the long run and it’s actually hard to say if they are directly connected to the success of the lagging metrics. But without setting the right ones, without understanding the decisions you need to make, without tracking them, without using them to make decisions, it will be hard to increase the performance of your business.
One way to make these metrics more relevant is by adding “annotations” to the charts. If anyone looks back, and sees the development over time of some of these metrics, without the context, there are few decisions to be made. They are good for soothing and comfort when you are heading up, and for triggering an alarm when going down, but context is key here. An annotation log, history log helps the viewer go back to a story of what happened in a certain period, and understand the background of growth/decline. This is useful for decision making.
Coincident metrics
Also known as benchmarking metrics. They are not your business’s metrics, they are related to the external ecosystem—because no business grows in a vacuum. These help you understand market forces, overall or category economic dynamics, and competition. Their most important use is in comparing performance or goal setting. Your performance can look amazing compared to past performance, and lousy when you compare it with your competitors. It can help you gauge your lagging metrics and get extremely useful insights.
My favorite example is the Lipstick Metric, by Leonard Lauder, the founder of Estée Lauder Cosmetics. He discovered that an increase of small luxury items such as lipstick indicates an oncoming recession. He’s right, look at the data today!
(Special note here for SaaS startups: Nathan Latka’s website is a goldmine of benchmarking data for any subscription business. It helped us understand growth, valuation, and much more).
Because the Internet loves step-based recipes to success, here’s my simple seven-step recipe to set the right metrics. (I’m kidding, it’s not easy at all, but what the heck, why did you become an entrepreneur anyway?)
The hardest question: How does success look like for your startup? Close your eyes and picture it. Describe how a day in a successful business looks like in three years. Don’t use any numbers or quantifiable things. Ask your co-founders to close their eyes, picture it, and describe it. Now you’re in big trouble because things don’t look the same. Spend some time deciding on a common picture that you can all agree with.
What can you measure about that future? What are the indicators that measure how successful or unsuccessful the future is? I know this already sounds like a spiritual retreat with mindfulness and aural stuff, but please bear with me. One thing I have learned over time is that each business has ONE key metric which can tell you if you are doing good, or you’re in trouble. The problem is that you need to recycle and check a lot of metrics before finding it. A great book I once read that showed this exercise in action, and its value to the business was The Knack by Norm Brodsky and Bo Burlingham. Let’s say that you have decided that Monthly Recurring Revenue is that KPIs.
These KPIs are influenced by specific success factors. What are the success factors that influence the KPIs? Try to apply some systems thinking theory and not simplistic causality. These success factors should be qualitative, not quantitative. For example, for MRR, success factors could be marketing reach, conversion to MQLs, sales performance, product quality, customer support resolutions, customer satisfaction, referrals, and maybe more.
What actions improve these success factors? Let’s choose marketing reach as a success factor that may drive MRR. What actions do you believe will increase marketing reach? Maybe increasing your social media presence, or writing more qualified content on your blog, or increasing marketing spend for your distribution pipeline? By the way, if you’re looking for some useful and easy to automate marketing tools, you can find some here.
How do you measure the actions? How can you quantify their output? They won’t tell you much about performance but will tell you about progress. Performance can’t be predicted, but you can test whether the actions lead to performance. These tests are quantitative, very short term (daily or weekly), and help you make decisions and stir the ship. By tracking, for example, the number of posts on social media each week, you can understand in a few weeks how they influence your lagging indicators. If you don’t track them, it’s just going to be luck. Or bad luck. These leading metrics (which you should discuss each week, by the way), will help your team understand your and their decision.
Where do you get this data from, to be able to track the leading metrics? As I’ve mentioned, this can be a time-consuming exercise, so re-evaluate whether you have this data or not. If the effort is too big, maybe you’re measuring the wrong output? Or maybe not?
Go put this to work in your team. Very important insight (sorry I left the logic to the end: Your management board should work with lagging metrics. Your team should work with leading metrics. Because of lag and complexity, setting a target for a team member to a lagging metric, such as “$100,000 sales this quarter” is a secret recipe for failure. The more pragmatic and immediate metric “100 demo calls this quarter” offers a more straightforward path. One where the team member has control over it, can be accountable and make short term decisions on it.
That’s it. Simple, no?
We are doing something wrong (or when metrics cross with micromanagement)
Now... we have been in a lot of projects until now, we had been in a lot of roles and experiences. Some of you know metrics are like when we are in the middle of arguing with someone and they (or us) start pushing data one after another trying to make our point valid, just spouting numbers and studies (some of them just theories), but them make hour point valid, or make us think to be valid. Well, metrics give us the same, if we need to ensure something the work, evaluate the results or or improve the process, we need to analyze the data, and that can only be done through the metrics. And those metrics can not be done just to demonstrate our points or to show what we want, but to improve the process, product and ourselves.
Comments