Here's the thing with KPIs--they can be anything you want. And too often, they end up being just that. However, performance should be judged based on the value that is created. Unfortunately, most people don't realize how many assumptions they are accepting when they determine value. "Of course having more Facebook fans is important." "Of course impressions have an impact on brand affinity." "Of course we need to be running ads in every keyword group that we can think of." These are all assumptions and metrics that track them are representing assumed value.
The truth is, the assumptions that underlie many metrics are usually most valuable to the people who are peddling them rather than the businesses that are paying for them. Are your social ads low on leads? Well, just point out the number likes that those ads are getting. When you add this assumed benefit, of course the ads are providing much more value then you realize! In this way, true value can be rhetorically obfuscated by assumed value.
There is a good argument to be made that, as a vendor, your job is to keep the client happy, no matter what. And there is no harm in that. However, this does not account for that fact that vendors are usually chosen based on their area of expertise. They are appreciated because they bring a knowledgeable perspective to the table that helps inform and improve the marketing effort. Unfortunately, this is not always the case. The actual people working the account may have no more actual experience than the client, even if the agency website and account managers tout their credentials. They will naturally skew toward the self-congratulating story. They want to be providing value so they find ways to show they are providing value. They lean on their assumptions and find KPIs that represent perceived value as opposed to actual business value.
Questioning your assumptions is always a good idea.
It's very easy to show perceived value if you accept the assumptions. However, it is much more important to show actual value based on reliable metrics that are tied to reality.
So, how do you get to actual value? Well, that is where it is easy to fall through the cracks (and there are a lot of cracks). Here are some things to think about:
1. What is the mission of the marketing department and what can be measured? Marketing teams' core mission is always something like "build brand awareness", "increase favorability", "grow the audience", "increase market share", "generate more leads", or "generate more demand". Things like sales and leads are pretty straight forward to measure. But what about marketing teams that are tasked with the more nebulous ideas like "awareness" and "favorability" and "education"? These can mean almost anything and people tend to chose the most favorable metrics they can. Take some extra time and make sure you are getting the KPIs that tie as closely as possible to the specific mission of the marketing organization. Look for ways to understand the impact your efforts are having, not just the response. Use metrics like Net Promoter Score, Lifetime Value, Lead Quality Score and Customer Sentiment Trends to gain insights that are more meaningful. Don't stop at simple response data. Measure deeper.
2. It is good to be skeptical when it comes to determining and measuring value. Ask "why?" a lot. For instance, why is content consumption (as measured by video plays or time on page) valuable? We assume that when someone watches a product related video, then they are interested in the product and have learned something by the end. This fulfills the "education" objective many marketing teams are tasked with. But has the viewer really gotten an education? Did it really sink in? Did they really understand what they just watched? Can they name the most important features of your product? How do you know? You'll need to find a way to know. There are a lot of assumptions going on that need to be validated. The best KPIs are the ones that show the validation of those assumptions, not just track the assumptions themselves. Surveys and customer feedback can provide good insights for this.
3. Handle KPI evolution with care. Sometimes new data becomes available or a better understanding of how to measure is developed. That's fine, but evolve your KPIs with caution. Also, I'd recommend that you continue tracking the old KPIs for a span of time so there is an overlap. This will help prove the value of the new metrics. If the old and new metrics tell very different stories, then it is a good time to regroup and make sure you're evolving in the right direction and not simply leaning on new assumptions. Get all stakeholders to agree on the new KPIs.
4. Watch out for KPI dilution. If a new voice comes to the table and wants to start mixing in new metrics that "of course" add value, then raise the yellow flag. It almost goes without saying that they are trying to control the story by shifting focus to the metrics they like best rather than the metrics that represent true, agreed upon performance. Unless the marketing mission changes or better data becomes available (see #3 above), the KPIs shouldn't change.
Overall, the world of marketing is rich with metrics and it is rich with practitioners of various levels of ability and judgement. In this environment it is easy for assumed value and KPI dilution to become a real problem for marketing teams. Because of this, Gartner's recommendation to hire analytics people "with the seniority and experience" is especially poignant. It is too easy to assume your way into a no-value situation. Keep in mind the four items above and you'll be on your way to measuring true value.