How to determine KPI's and metrics

During the Outlook a circle gives input to the Circle Lead to determine the KPI's and metrics. This is about "what" KPI's you want to use. If you want to learn more about good metrics and KPI's 📈[11] Metrics & KPI’s is a good read as is .

Determining metrics / KPI's

Metrics are meant to help you learn

The metrics might be seen as something to aim for but they are mostly used to learn from and to measure the results of the experiments you run to improve your metrics.

Tips when creating your metrics / KPI's

  1. Choose a metric that is directly related to your circle and can be influenced (mainly) by the circle the metric is for
  2. Less is more; better a few key metrics than too many metrics
  3. Metrics need to be measurable
  4. Metrics can be linked to actions; when a metric is changing would you need to act? If the answer is 'no' then it's not a good metric
  5. Align your metrics with the purpose
  6. Pick a metric that everybody understands

A metric or a KPI (key performance indicator) is intended to monitor or assess a specific business process. As such, the metric is actually a proxy for the process: The number of conversions on a website says something about a marketing process A success rate of the number of deals says something about the sales process

Since with self-management, there is no longer a manager to set these metrics, the team is now allowed to set them themselves. And you definitely want to do that; it ensures that the team's business process becomes measurable. In doing so, it provides visibility into the impact of improvements being made. But what is a good metric and how do you determine its height?

A good metric has a number of characteristics:

1. Easy to measure

First, the metric is preferably easy to measure. It should also not take much time to collect and measure the data.

2. Easy to understand

Second, the metric should be easy to understand and explain. The metric is often discussed with the team and can also be viewed by other people in the organization. It is nice if it is immediately clear to everyone what is being measured. You can even take this into account when naming the metric. If it is clear and descriptive, it explains itself.

3. It has a direct correlation to the business process

This is a very important one. The metric should teach you something about your business process and actually even the foundation of the business process. Only then can the metric really tell you something about how well the business process is functioning.

An example: you get 100 leads a month from the website. Your marketing team does a great job of that. The marketing team has an influence on the number of leads, which can be their metric. These leads get a quality score from the sales team. They do their best to turn the leads into happy customers. You could then set the number of new customers as a metric. But: the number of leads is something the sales team has no control over, as is the quality of the leads. They do have control over the sales process. The success rate can therefore be a much better KPI for the sales team, than the number of new customers.

4. It shouldn't be data, but information

If I say marketing had 70 conversions last week, that's data. If I say that marketing averages 60 conversions a week, that it was 10 more last week, and that this can be directly linked to an experiment we did, this is information. Ideally, I have historical data that allows me to give context to the different data points. It's totally fine if you can capture that context so that it's transferable to other colleagues or teams.

Percentages often provide more information than hard numbers. An example: a conversion % on a website says something about the quality of the website and about the marketing activities; are we getting the right visitors on the site? A hard figure - x number of visitors on the website - says in this case only something about the end result. You learn less from that!

5. It must be influential for the team that measures it

A metric must be influenceable by the team that measures it. If all my deals come through the website and I measure the number of deals as a metric for a sales department, then a sales team doesn't really have any influence on it. This is because it is the marketing department that can turn the "deal button": they manipulate the number of visitors, the conversions, and thus the number of deals.

However, if I set the success rate as a metric for a sales team, they already have more influence. If I also link the success rate to an individual team member and to the channel over which the deal is handled (mail or phone), and I link it, for example, to the type of customer and lead time, then it becomes really valuable information. Because now, as a team member, I can see who has the best performance on a particular channel and learn from that to become better myself. Or I can choose to specialize in a certain channel because I am simply better at it, which allows me to contribute more to the organization.

In the perfect world, a good metric can only be influenced by the team and not by external factors. Unfortunately, this is not always possible.

6. Supports behavioral change

A good metric supports behavior change - as you can see in the example at point 5. Ideally, the metric also supports the experiments you do with your team.

Some people think that a good metric should be able to benchmark within your own organization or against other organizations. I'm a bit wary of that; look for what you yourself find relevant for your business. "If we can keep competitors focused on us while we stay focused on customers, we'll turn out all right", I think applies to metrics as well.

Note: because a metric can involve behavioral change, the wrong KPIs can also trigger undesired behavior. You can prevent this by putting metrics on the process (are all steps followed) instead of only on the result.

7. A metric helps you learn

Something that often goes wrong in companies is that a metric or target is used as a stick to beat. That is not the purpose of metrics. It should help the people in a team to learn and to improve. A metric is therefore often part of the build, measure, learn cycle discussed earlier. It helps you measure and learn.

The target of a KPI or Metric

The target of set KPI's is not set during the outlook. For that we use forecasting and this is done by the Circle Lead or role that is the KPI owner. Forecasting is part of .

How to do successful forecasting

Forecasting can help you to show if you are on the right track to reach your goals and targets. But also gives an idea of where we are heading.

Easy/pragmatic way of forecasting: Google Spreadsheet might be the easiest way to do forecasting. Take the numbers of the last year and just "drag" them on and Google will do its magic. This will just pull the current trend straight through to the next days/weeks/months. See the video example below.

You might want to make some adjustments to the numbers (seasonal, because you know or expect some experiments will have a high impact on the numbers).

More accurate/time-consuming way of forecasting:

When you actually want to make a prediction of where the current numbers are going. For example, for the growth of the number of customers, you also want to take other factors into account. Like seasonality, events that had a big impact on growth, etc. A prediction model can help with this, there are a lot of prediction models out there. Chances are that #gcc_data can give you a hand with this. Can more accurate forecasting make a big impact or difference, than this way is preferred!

Example video:

Link: Spreadsheet example

The data circle is regularly organising a Better Metrics Workshop which goes into detail on how to create good metrics. You can check the for details. Look for