We can’t find perfect example KPIs, but by following these 6 steps, we can find some that give us a great starting point to meaningful KPIs.

I don’t believe in shortcuts when they take us around opportunities to get the best result. And most shortcuts in measurement just lead us into the same old problems we’ve been complaining about for decades. Like meaningless measures. That’s why I am most definitely not an advocate of measuring goals by searching lists and libraries and databases for example KPIs. [I prefer this.]

But when we’re stuck for a starting point, or trying to measure a performance domain we’re unfamiliar with, example KPIs can be helpful. We just have to be careful in how we use them.

There are some very real dangers of using off-the-shelf KPIs, but they’re avoidable, if you follow these six steps…

Step 1: Set your domain of performance for the KPI search.

Setting your domain of performance means choosing a category or topic that narrows the scope of example KPIs you will look for. For example:

• video marketing
• digital transformation
• workforce capability
• policy development
• diversity and inclusion

If you fail to do this step, you’ll get overwhelmed with too many KPIs. Imagine how many example KPIs you’d find if you searched for ‘local government KPIs’. If you have more than one domain of performance to find measures for, define each one separately.

Step 2: Search, find and list the example KPIs.

This is a simple step, but it might take some time to trawl through all the available sources. For example, we might find the following example KPIs for the performance domain of video marketing:

• impressions
• reach
• views
• likes
• comments
• consideration lift
• watch time
• view-through rate

You might be tempted to shortlist the example KPIs as you find them, but resist! It’s faster to only capture whatever you find (of course, exclude the obviously stupid ones), and let the rest of our six steps help you with the shortlisting.

Step 3: Make sure each example KPI is clearly defined.

You don’t have a real KPI or performance measure if all you have is a few words. A KPI is a quantitative measure, and the method of quantification has to be clear. More explanation of this is in Assumption 1 and Assumption 2 of this article.

So, make sure each example KPI on your list meets this recipe for a well-written measure. If it doesn’t, ditch it! For example, we want to find details like this [based on this source of video marketing metrics] for each example KPI:

• impressions: the number of times the video is shown to a viewer
• reach: the total number of unique individuals who have viewed the video since it was published
• views: the number of individuals who watched at least 30 seconds of the video
• likes: the number of individuals who clicked the ‘like’ button on the video
• comments: the number of comments individuals have made on the video, excluding the author/publisher replies and excluding spam
• consideration lift: audience perception and whether the campaign has had the desired effect
• watch time: the average time in minutes that viewers have spent watching the video
• view-through rate: the total number of individuals who watched the video until the end, divided by the total number of views (see definition of views above)

From this list of video marketing KPIs, ‘consideration lift’ was ditched because it wasn’t easy to find a clear and quantitative description.

Step 4: Try to articulate the question each example KPI answers.

In step 3 we starting shortlisting the example KPIs based on their clarity as quantitative measures. In step 4, we are shortlisting further, and this time based on whether the example KPI tells us something important.

If you can already see an obvious connection between an example KPI and one of your goals, you can skip this step for that KPI and go to step 5. If not, then try and frame the question that the KPI gives the answer to. For example, our questions for the video marketing KPIs might be:

• impressions: answers the question “Are enough people aware that the video exists?”
• reach: answers the question “Is the video attracting a large viewer audience?”
• views: answers the question “Are people curious enough to at least start watching the video?”
• likes: answers the question “Are people liking the video content?”
• comments: answers the question “Are people engaging with the video content?”
• watch time: answers the question “Are people finding the video content useful or interesting enough?”
• view-through rate: answers the question “Are people finding the video content useful or interesting enough, to watch to the end?”

Another video marketing KPI was ditched: ‘likes’ answered a very similar but not as important question as ‘watch time’ and ‘view-through rate’.

Step 5: Check that this is a relevant question to answer, given your goals.

Now it’s time to align our example KPIs with our goals. No one has time to measure things just because they’re interesting. We make the most impact when we measure what really needs improvement, now.

Map the questions from step 4 to your current goals relating to the domain of performance you set in step 1. Where you don’t see a strong alignment, it
likely means the example KPI just isn’t a priority for you right now.

For example, we might have two goals relating to the performance domain of video marketing:

1. Goal: Our videos reach a large market.
2. Goal: Our videos are useful and interesting to our market.
3. Goal: Our videos lead to sales.

And, using those questions from the previous step, we might map the example KPIs to those goals, like this:

1. Goal: Our videos reach a large market. Measures: Impressions, Reach
2. Goal: Our videos are useful and interesting to our market. Measures: Views, Watch Time, Comments, View-Through Rate
3. Goal: Our videos lead to sales. Measures: none found yet

For the first goal, we found some example KPIs that suited. The second goal might have too many. And the third goal has none.

Step 6: For the remaining example KPIs, keep the strongest.

You might already have shortlisted those example KPIs enough, if you end up with one, two or three KPIs for each goal. But if you have more than that, or you’re just not convinced you found the most meaningful evidence of your goals, there’s another step you can take.

In the PuMP performance measurement methodology, we use a deliberate Measure Design technique that helps us design measures directly for our goals. Firstly, it guides us to design measures where none exist already. This is useful if the example KPIs you found don’t provide comprehensive enough evidence of a goal.

Secondly, PuMP’s Measure Design technique helps us shortlist a set of potential measures based on their strength and feasibility. Strength means it’s very direct and convincing evidence of the goal. Feasibility means it’s realistic to get the data and bring the measure to life.

For our video marketing example, we might ditch the measure ‘Views’ from the second goal, as it’s not as strongly convincing of how useful and interesting our videos are. And we need to do a full Measure Design for our third goal, as none of the example KPIs related to sales conversion from video.

BONUS Step 7: Are any goals missing meaningful measures?

More often than you might realise, freely available example KPIs won’t cover your unique needs. You might have goals that have only one reasonable KPI, but it doesn’t give the full picture. You might have goals that are left with no example KPI suiting them at all.

Good-enough example KPIs might be a case of the glass being half full. But it’s also half empty.

An excellent final check is to make sure you didn’t get blinkered by the example KPIs available. Take a critical look again at the KPIs you ended up with for each goal. Ask yourself a few questions:

• Do I feel excited about each and every one of these KPI?
• Would I feel convinced that the goal was achieved, just by trusting these KPIs?
• Would other stakeholders be convinced that the goal was achieved, just by trusting these KPIs?
• Would I feel more confident and excited if I spent another hour or two and found a few more powerful KPIs?

These six (seven) steps might be a good way to get on the path to better performance measurement, if you still have some resistance to it.

Finding example KPIs, instead of designing them deliberately, might be a case of the glass being half full. But it’s also half empty. [tweet this]