In my career, I have often seen measurements come from an off the shelf software tool or off-the-cuff comment from leadership. It is important to apply rigor to build out KPIs that align incentives in order to build the business and increase profitability. Sometimes industry standard KPIs work well, but sometimes they do not fit your company. The following framework has served me well in defining measures for my teams to work toward.
The issue to resolve: The first task at hand is identifying the problem you are trying to solve or prevent from happening. Starting here allows you to tie your metrics directly back to a business case. As tech leaders, many of us have some type of budget and profit and loss responsibility guiding our moves. Starting with the bottom line and identifying issues will allow for your metrics to directly tie back to a strong business case. In turn, it will be easier to manage up and down with your new set of KPIs.
An example could be visibility into various digital marketing campaigns. This might be prompted by your colleagues on the leadership team not grasping the fact that direct and organic e-commerce leads are not free. Whatever the case may be, it is helpful to start with the issue you are trying to improve.
Indicator: Now that you have the issues identified, what are the indicators that might mean progress towards the goal? One common fault of KPIs is people jump straight from the problem to considering what to measure. This can lead to misaligned incentives. I have seen a company try to improve their customer's online experience by measuring bounce rate. While well intended, they did not realize at the time that bounce rate can easily be gamed and has several causes. Improving bounce rate doesn’t necessarily make for a better experience or bottom line for that matter.
Tying back to the previous example, the leader of this team might simply set the goal of improving the efficiency of his or her digital marketing campaigns. Note, this is still high level and abstract, but it lets you know what will improve your issue.
Metric: People often use the terms "metrics" and "KPI" interchangeably. While colloquially they have become synonymous, using their denotation is helpful when framing what you are trying to measure and improve. A metric refers to a standard measurement. The difference between the two is that KPIs are geared toward end state measurement. It is important to note that KPIs are a subset of metrics. KPIs should reflect the actual performance of a company, meaning it should be actionable data.
For example, a metric to track could be cost per conversion by channel on digital marketing campaigns. This captures what you measure in making your digital marketing campaigns more efficient. While this might sound specific, there is still a lot of ambiguity for a frontline analyst. "Do we include both fixed and variable costs?" and "How often do we measure this?" are two of the most apparent open questions.
KPI: This is the very tactical way to explain what you are tracking. At this point, it should be actionable for someone on your team, and it should be a direct measure of something that will affect the bottom line of the company.
For example, this could be cost per conversion by channel on a digital marketing campaign, month over month factoring in all associated fixed and variable costs including human capital, and general and administrative overhead. If I were to tell someone on my team this KPI they would know exactly what I want to measure and how often I want it.
Source: The final step in defining KPIs is identifying where you want that information from. For anyone who has worked in a large organization that uses multiple data analytics tools, you know these numbers rarely, if ever, are exactly the same. Data already has enough noise without the additional variance you would see by comparing different methods of tracking. With that being said, a good question to ask your team is if their sources are directionally the same.
For example, if Google Analytics is always X% lower or higher than a competitor like MixPanel or Adobe Omniture on total app visitors, they are likely both implemented correctly, they just have different methods for collecting data. As a technical leader, it is helpful to have an informed opinion on which source is most accurate, but be willing to let your team teach you a thing or two. An informed approach is vital for being able to manage up, out and down.
One thing that can make or break even the best plan — assign an owner for every KPI. Like they say in the NFL, when you have two quarterbacks you have none. The same holds true with metrics -- when two people are responsible for one, no one is. Ensuring the success of metrics must include putting them in the hands of a specific, talented person.
Building out metrics that drive bottom line growth requires some forethought, but in the end, your company’s future depends on doing it right.
Rob Versaw, Co-Founder and Head of Product at Green SE3D AR; Head Product Adviser at Blockbox