Different ER or ERR values must first be checked against the formula, and only then should you look for an error. Services may count interactions, views, or reach, use different posts, and take measurements at different times. The same abbreviation does not make the results comparable.
Create a metric passport
Write down four fields: numerator, denominator, sample, and observation period. For a manual calculation, this could be "reactions plus comments / followers; ten regular posts; over 48 hours." Next to it, write out the definition from the second service's report, including excluded posts.
| Metric | What is in the numerator | What to check |
|---|---|---|
| ER in the Boosted calculator | Selected actions | Whether reposts and saves are included |
| Share of views from followers | Post views | Age of the post and date of the follower count |
| Metric from an external report | The service's own definition | The formula, window, and sample composition |
TGStat statistics documentation distinguishes engagement metrics. Do not declare your own division to be a copy of an external service's current algorithm if its window and filters are unknown.
Training check
A post has 1 000 followers, 600 views, and 30 reactions with comments. Actions relative to followers give 3%, while views relative to followers give 60%. These are different ratios, so the difference by itself says nothing about an error or artificial engagement.
In the calculator you can reproduce the first ratio by selecting followers and the required actions. The second requires a separate calculation of the view share; do not fill the reactions field with views just to get the same name.
When it is better to postpone the comparison
If one report measures a single day while another measures accumulated views over a month, keep both with clear labels. Request a comparable export. Telegram warns about the specifics of views; the public counter does not reveal the exact unique composition of the audience. The practical result of the check is an agreed methodology for the next report, not choosing the higher percentage for a media kit.
If a separate public announcement in Telegram has been prepared for this work, you can check the current view terms. The ER calculation does not determine the quality of purchased activity. Purchased activity does not confirm organic interest, recommendations, enquiries or sales.