Narrow down investigations
What to look out for
Real User Monitoring (RUM) will vary a lot depending on devices, countries, type of visitors, the network they have on their device, their battery level, location, languages of the website etc.
Expectations of visitors will also vary a lot depending on devices they use, where they are in the journey and the kind of actions they take on a website.
Within a session, users might face pages with good or bad performance. The more sessions they have, the more they will potentially see pages with poor web performance. This can make quantification at session level a bit imprecise.
Best practices
As a result it’s important to narrow down investigations as much as possible on specific segments and pages.
It is also recommended to make sure to exclude bots and spiders because they have crawling-only behavior coupled with an inexistent conversion rate which completely distorts the correlation between web performance and UX improvement.
Use goals carefully
What to look out for
Contentsquare allows impact quantification on business goals so it is important to understand correlations between behaviors and goal achievements; It is not a measurement of causation. Please keep this in mind when building action plans based on results.
Best practices
Focus as much as possible on the goals of pages you analyze. Avoid using goals potentially achieved 3 or 4 pages after the page is analyzed.
Ask precise questions
Best practices
When completing an analysis, try to answer precise questions. For example: Do users who faced a LCP >3sec on the Homepage bounce more than those who don't.
Use comparisons as much as possible, examples:
- Analyzing based on before and after a big slowdown of web performance
- Before and after some improvements are made,
- Users who face bad performance due to old devices vs others