Different data from Analytics and Metrics - who to trust?

Question

Google Analytics

The data on visits and sources from these two analytical systems really began to differ greatly. Judge for yourself (you can only look at June):

In fact, these metrics are more accurate. Why is that? What about analytics? Whom to believe and how to be?

Answer 1

The discrepancy in the number of users for tagged traffic is less than 1.5%. This is normal for external js analytics. Organically - you need to segment and look at the distribution by sections, where is the difference => why it exists.
Sessions - similarly, you need to look under the hood, you can’t say anything by these numbers.
Failures in G and I are different entities, read the docs. They will never match.
Timing and JS - take it as a 5th approximation to the real median.
In general, these numbers are useless if you want to get answers from where deltas come from. You need a website, you need GA, you need Metrica. Sit down and dig.

Answer 2

for your 5 people a day, you can ignore the spread.
there will always be a difference, since the scripts work differently on different computers of visitors and with different Internet connections

Answer 3

First you need to make sure that the tracking settings are correct and the counters are placed correctly. Only then can one analyze why they diverge.