DATA.
How to create KPIs
Koen Pauwels, Feb 14, 2021
As we discussed last week, ROI is not a silver bullet metric in marketing effectiveness: it requires lots of judgment calls on incremental revenues and costs, focuses on short-term over long-term and on efficiency over effectiveness. In the words of prof Tim Ambler: ROI is dead, now bury it.
So how do we come up with potential KPIs that do cover the long and the short? In book Chapter 7, we outline 3 approaches and the practical check of the Wheel of KPI discovery. Let’s start with the three approaches to KPI creation:
-
Based on your industry experience of the customer journey;
-
Based on best practices / empirical generalizations (general approach);
-
Based on your company goals and industry specifics (targeted approach).
​
First, your industry experience helps you design a sketch of what the typical customer journey or purchase funnel looks like, and your metrics should reflect these stages. The first such sketch I draw with a cereal brand manager on a napkin a quarter century ago:
Profits is what this leading brand was after, and these were influenced by sales and margin. Margin depends on price, which also affects the perceived customer value and the usage occasions consumers would consider the brand for (eg as a late-night TV snack instead of only as a healthy breakfast food). TV advertising also can suggest new usage occasions, and additionally reinforces awareness, some of which leads to Affect and drives purchase intention. Both intention and value drive trial, which translates into sales and profits. We found good data sources to track each of the KPIs in this figure and the manager found it useful for decision making.
But is this sketched customer journey correct? Obviously the last 25 years added much in terms of online metrics and of offline word-of-mouth. Moreover, the sequence of the shown effects Is questionable. In a recent comparison among 178 brands in 18 fast moving consumer good categories, Affect metrics were most likely to react to advertising first, with Cognition and Experience following. This ACE sequence is most prevalent for utilitarian products and less differentiated brands.
Second, the general approach gives us 10 metrics that drive performance according to Tim Ambler’s summary of large US and European studies:
-
Three P&L measures: Top line (Revenues), Marketing Investment, Bottom Line
-
Seven Brand Equity measures: Awareness, Penetration, Consumer Thoughts about the brand, Consumer Feelings, Brand loyalty, Availability, and Relative Price if you can calculate it, i.e. if you can delineate your competition clearly
​
This general approach is favored by managers who want to compare their scores with competitors, as tracked by third party organizations such as GfK, Kantar and YouTube: “Keep your metrics down to a few which can be applied in every company. The metrics message will not come across unless it is simple and can be compared to other firms”
​
In contrast, the tailored approach agrees with the manager statement
​
“No two companies are alike. Each has its own strategy and positioning which should determine the relevant metrics. Just as positioning requires differentiation, so their metrics should differ too”. It calls for the following steps:
-
Start from your specific goals and strategy, i.e. start with your vision;
-
Your Metrics represent what is needed to achieve your goals;
-
If the strategy not clear enough, clarify it or benchmark metrics from the general approach that fits your goals best.
​
This is a demanding process, but worthwhile, especially if environment changes, your performance drops or ambitions increase. For instance, if you need to increase revenues 20% next year, you may realize this translates into the marketing objectives of increasing Awareness by 30%, increasing brand liking by 20% and increasing sales conversion by 10%. The metrics could then be unaided awareness, ad engagement, website perception score, lead conversion rate etc. You and your team can then brainstorm tactics that would move the needle on these KPIs, such as over-the-top (OTT) on Fire TV, a new online video + email campaign, a new TV ad copy, better website user experience, Sponsored Amazon ads.
Finally, each proposed metric satisfy the 4 practical criteria of this KPI wheel:
-
WHAT business question would that metric answer?
-
WHO will use it to make decisions and take action?
-
HOW Important is it compared to other metrics?
-
WHERE will the supporting data come from?
