Earn the fruits of your labour using KPIs
KPIs are the proof if your effort has been worth it and how your product or feature could be improved.
Everybody wants to earn the fruits of their hard work on a project, but only a few of us take the correct path. Using key performance indicators is a powerful way to define how successful the product or feature is after release, and we can continuously improve it.
KPIs are the proof if your effort has been worth it and how it could be improved.
Even when you have done many user tests during the entire developing process of your product (low-fi state, visual design state, live state) and everything seemed to work well, it is essential to check out how your product performs in real life and if it brings the value you have expected.
Continuously measure the success after release
We do not only design things, test them, improve them and release them. Product development is not a linear process. We should check the success of our product as often as possible (especially after the first release) and continuously improve it. This is what we call iterative product development. It doesn’t happen very often in the world outside, but I think we are on the right track. Big players like Google and Facebook do it very well, and we can learn from them.
I have seen many products that failed because of the lack of measurement and improvement. The situation is always the same, for example somebody thought that they were finished with the release of the product because of the positive feedback they received from the user testing.
But the truth is, we are never finished with our products/features if we want it to survive in real life and become better than the competition. You don’t have to decide that it’s finished but you can either put it on the back burner or decide to completely shut it down.
That is why you shouldn’t trust your own positive feelings and always define KPI to measure success and avoid the huge number of drop-offs as early as possible.
What are KPIs in detail
KPI or Key Performance Indicator is a metric that we define to measure our product’s success continuously. It relates to a specific value that we want to achieve and can be easily proved with data. The core team should define them at the beginning of the project.
As you can see, the value of your metric is significant. We can’t measure the success without it.
Some developers define one KPI, but in my opinion, it’s more helpful to define the main one and a few sub-KPIs, because this way you get more reliable information on whether your product/feature affects other products/features or not and how it contributes (positive or negative) to their value.
It is very important to find out how your product/feature performs in general and how everything relates to each other. When you use more than one metric, you can be more specific in your result and find out that your customers, e.g. are interested in your product, but they also generate a considerable number of drop-offs.
You can use quantitative and qualitative methods to prove if your KPIs are fulfilled or not.
What isn’t KPI?
- KPI is not the goal of your project, but it can relate to it
If the goal of your project is “Creating a new user area where the customers can leave their personal details to be more flexible by filling out some forms, to avoid repeating this process.” the main KPI could be “increasing the % of new registered user”. - KPI is not the definition of done for your project
The definition of done is the commitment in your team of what should be delivered to mark the first project iteration as done.
When should I consider the KPIs?
You should always define them at the beginning of the developing process. To be precise, in the product discovery phase.
You can define a KPI for an already existing product/feature you want to improve, but you can also define a KPI for a completely new one.
How to define your KPIs:
As we said above, a KPI includes a metric and a value. The metric could be any possible item that can be measured in value e.g. :
Let us imagine you have a product that has a considerable amount of drop-offs. That is why it must be improved. You probably have the some data about the current usage, and you should use it as a basis.
e.g.
“The current drop-off of the particular product is 2%””
So the metric should be drop-off and the value lower than 2%
Your main KPI could look like this:
“Minimizing the drop-off rate from 2% to 1,5%”
You must set a period of time for your measurement. A day, a week or a month. Usually, 2% is a huge amount when you have a large amount of traffic on your website. So probably the value could be too high and almost impossible to be achieved for such a short time. You must consider the traffic of your product.
How to check your KPIs
As I mentioned, the best way to check your KPI is to use a good medium of data. I always combine quantitative with qualitative data and use not only analytics, but recordings.
The analytics data are beneficial, but they simply don’t show us where the issues are. You must be careful because analytics is about the visits and less about the visitors indeed.
If you want to find out more about the real visitors and their behaviour, you should consider recordings. They are beneficial when you want to know the real struggles of the customers.
I recommend you consider them both. Numbers can tell the amount and real visitors tell you where the problems are.
Conclusion
KPIs support the iterative way to work on a product, so you have a never-ending story, and this is good so, because there is always something that you can improve.
But you should also be careful with your measurement medium and the value, especially because analytics numbers can be very tricky. To avoid misunderstandings ensure yourself with recordings, user test or BI data.
It is very important to use KPIs as early as possible in your product, because it helps you to consider them in your concept. Additionally, to that, you start more purposefully with your project and get more solid insights if it works well outside in the real world, or there is something you should do better.
