Not all KPIs are the same – discover how calculation logic influences your decisions and what really matters.<\/p>\n
The path to informed, data-based decisions begins with three essential steps:<\/p>\n
You know the KPIs to measure these goals<\/p>\n
How are the metrics collected for this?<\/p>\n
& Understanding data<\/p>\n
What should be achieved – and how can this be measured? This question is at the beginning of every good analysis. Ideally, you formulate your goal according to theSMART method<\/strong>. You then derive suitable KPIs from this.<\/p>\n Examples of goals and associated KPIs:<\/p>\n How can a single KPI – e.g. B. theSwipe rate<\/strong>in the online shop – deliver completely different values? Quite simply: it dependsdefinition<\/strong>to. Which metrics are used? How exactly are they collected? And what is the formula?What does swipe even mean?<\/strong>Swipe describes the wiping movement, e.g. B. on the smartphone. In online shops, this means browsing through product images – an important interaction that can provide information about interest in a product.<\/p>\n Example calculation variants:<\/strong><\/p>\n Which variant makes sense depends on the goal and the general conditions: If the swipe function is not available on all pages, it may make sense to use “only page views with swipe function” for the calculation. This means that only relevant page views are taken into account.<\/p>\n If the number of images per product differs significantly, the metric can be distorted – unique swipes can help here.<\/p>\n If “pages on which a swipe was carried out” is used as a metric for the calculation, the number is always >=1. This is important and must be taken into account when interpreting the number.Important<\/strong>: The KPI is not wrong –but it must be understood and interpreted in the correct context<\/strong>. Comparing different calculations without disclosing this quickly leads to incorrect conclusions.<\/p>\n The following examples will show you how different KPIs can be calculated – and why it is so important to know exactly which metrics are collected and how. Here are a few food for thought:<\/p>\n Calculation:<\/strong>Clicks\/Impressionsor<\/strong>Clicks \/ Unique Impressions<\/p>\n But when does an impression actually count? Is it enough for 1% of the banner to be visible – or does it have to be 50% or even 100%? Details like this are crucial when comparing CTRs. So make surehow impressions were measured<\/strong>and whether the same fundamentals apply.<\/p>\n Calculation:<\/strong>Sales\/advertising expensesor<\/strong>(Sales – cost of goods) \/ advertising expenses<\/p>\n A common stumbling block:Gross or net sales?<\/strong>While platforms like Meta often calculate with gross sales, controlling often uses net values. When comparing ROAS, make sure youon the same database<\/strong>stay.<\/p>\n Calculation:<\/strong>Conversions \/ Visitorsor<\/strong>Conversions\/Sessions<\/p>\n And what actually counts as a conversion? Shopping? A registration? An add-to-cart? You should define this clearly – and communicate it openly. This is the only way to ensure thateveryone in the team thinks the same thing<\/strong>when they talk about conversion rate.<\/p>\n Key Takeaways:<\/strong><\/p>\n\n
One KPI – many calculations: That’s how relative “the” number is<\/h2>\n
Example swipe rate – what do you have to pay attention to?<\/h3>\n
\n
\n
\nAlthough variant 1 is less precise, it is often available more quickly and is easier to collect – useful for standardized comparisons (e.g. country comparisons).<\/p>\nMore examples<\/h3>\n
CTR (Click Trough Rate)<\/h4>\n
ROAS (Return on Ad Spends)<\/h4>\n
CR (Conversion Rate)<\/h4>\n
What you can learn from all of this<\/h2>\n