Data health & quality

A common question with analytics tools is why the numbers in Shahin may not exactly match other tools such as Google Analytics.

In most cases the gap is not an error — it comes from differences in definitions, timing or processing. The main causes follow.

The first and most common cause is that platforms define metrics differently.

The numbers may look similar, but since “what is measured” differs, expecting a 100% match is unrealistic.

Data only makes sense within a date range. Gaps can come from projects using different time zones, or from seconds-to-minutes differences in when events are recorded and processed.

Even with identical date ranges, different day boundaries or processing delays can cause small differences.

Analytics tools depend on the script running successfully in the user’s browser. Network problems, dropped requests or browser restrictions can cause data to reach one platform but not the other (or vice versa).

These gaps are usually scattered and unpredictable.

Some metrics — unique users in particular — cannot be computed with 100% precision. Like other leading tools, Shahin uses estimation algorithms such as HyperLogLog for them.

These algorithms offer:

• lower resource usage,
• high scalability,
• roughly 90–95% accuracy.

Small differences in such metrics are therefore natural and expected.