By Chris Wu
Impact Engine recently had our annual portfolio summit where many of our portfolio company CEOs gathered together to share their successes, learn from each other’s experiences, and discuss key challenges. One of the highlights of the event was when Mike Evans and Josh Evnin from Fixer, one of our most recent investments, led the group through an extremely useful exercise on developing and evaluating key performance indicators (KPIs).
Here’s a recap of the exercise:
Performance Takes Priority over Measurement
Mike began the session by setting performance indicators (PI) in the proper context. He made it clear that tracking and measuring business performance should be subordinate to making sure the business is actually performing well. The primary goal of every business is to achieve success (regardless of what constitutes success for each individual company). It’s far more important to ensure that the company is performing well than it is to always have an exact value to report. Reporting on performance metrics has limited benefits if your business has little to no sales revenue; your time would be better spent on building up the business before worrying about measurement.
Every Measurement Needs a Reference Point
If a company’s odds of success shoot up dramatically by achieving the target for a particular PI, or if the PI is necessary in order to have a proper handle on the business, then it is actually a Key Performance Indicator (KPI). Generally, it’s recommended to track no more than 10 KPIs at a time, any more than that can be too cumbersome and less effective.
A KPI value is meaningless unless you can evaluate it against a reference point, a foil, in order to determine whether that value reflects positively or negatively on the performance of the business. Josh and Mike went on to describe four main types of references:
Gut Feel — Early-stage companies often lack the requisite amount of historical data necessary in order to make meaningful archival comparisons. In situations like these, gut feel can serve as a valuable foil.
Comparison to Historical — While past performance is not necessarily an indication of future results, when done properly, comparing KPIs to historical values can be a helpful tool to understanding the health of the business.
Comparison to Ideal — Some KPIs have optimal values that represent an upper bound; the goal would be to maximize their value (i.e. # of clients, MRR). Others KPIs have optimal values that represent lower bound numbers; the goal would be to minimize their value (i.e. downtime, defect rates). Also, there are certain KPIs that should ideally shoot for a local maximum, in other words a sweet spot between the lower and upper bound values (i.e. sales pipeline conversion rate)
Comparison to Projections — Modeling helps provide clarity about the key levers of the business, but it can require a significant commitment of time and/or resources in order to generate that level of insight. If you’re able to incorporate elements from the other three reference types (gut, historical, and ideal) into its design, then the model becomes a much more powerful and effective predictive tool.
KPIs values should be reviewed and compared to its benchmark reference point on a regular basis. Every delta from the benchmark serves as an indicator of which aspects of the business management should focus on, and the company must act to resolve the underlying cause of the deviation in order to improve the performance of the business. Furthermore, the company needs to put in place a framework that prioritizes or ranks the different deltas in terms of importance.
Figure 1: Cycle of Model Refinement
KPIs Aren’t Theoretical, They’re Actual
As startup companies set about determining which KPIs make the most sense for their business, they should bear in mind that KPIs aren’t some theoretical construct, rather they are grounded in real numbers that are directly related to how the business is actually performing. KPIs are empirical, which means that they have inherent flaws. The act of measuring and recording them will typically introduce a certain degree of inaccuracy and error, which needs to be taken into account during post-analysis.
The Dualities of Metrics
Next, Mike and Josh broke down some of the distinguishing factors of KPIs:
Maximal/Optimal — There is a clear distinction between optimal and maximal KPIs. For certain metrics achieving a maximal level is ideal (i.e. # customers — no upper bound), whereas for some KPIs reaching a local optimum is sufficient (i.e. optimal # clinicians on staff per day — bell curve/sweet spot).
Leading/Lagging — Weight loss regimens are a classic example of leading and lagging indicators. In order to reach your weight loss goal, you must measure and record both the amount of calories consumed and the amount of calories burned. These are leading indicators. Stepping on the scale to track change in weight over time represents the lagging indicator.
Influenced/Not Influenced by Macroeconomics — Is the KPI sensitive to system level changes in the aggregate economy (i.e. net profit margin, net burn rate)?
High/Low context — High context KPIs have an implied meaning that are context-specific (i.e. operating costs, market share). They typically require further explanation in order to be understood properly, whereas low context KPIs are clearly and explicitly spelled out, and can stand on their own (i.e. cash & cash equivalents).
Mapping KPIs Graphically
Plotting all the company KPIs graphically (Relevance versus KPI-Type) can be a useful exercise and lead to meaningful internal team discussions.
Figure 2: Mapping KPIs
Placement along the Y-axis will clearly illustrate which metrics are closely tied to the health of the business versus other metrics which are less relevant. The metrics that lie close to the top are true KPIs, while the metrics that fall closer to the X-axis are essentially just PI’s.
Placement along the X-axis refers back to the different reference or foil types described earlier. At first, the company will most likely have a broad mix of KPIs across the different foil types. Over time, you may start to see more of the company KPIs bunch towards the upper right corner. Modeling involves a thoughtful combination of the other 3 categories, and as the business matures its ability to build more accurate predictive models also improves.
KPIs are like vital signs for the current health of your company, they can validate past success and lend support to claims of future growth, or they can serve as an early warning system when the business is in trouble. Every startup is aiming for success, however they choose to define it. Having the right set of KPIs and foils in place will allow you to keep your finger on the pulse of your startup.
Want more from Impact Engine? Click here to sign up for our monthly newsletter!
