Analytics

Analytics: reading your numbers

Per-account and per-video performance for TikTok, Instagram, and Facebook, and how to query it in chat.

Every post Pamba publishes is tracked: views, likes, comments, and follower movement, per video and per account, across TikTok, Instagram, and Facebook.

An account Pamba ran keeps counting after you archive it, by removing it from analytics or replacing it with a new one: its past views and posts stay in your totals and lists, marked Archived, and nothing new is read for it. An account you track for analytics only, posting from your own phone, leaves your analytics completely when you stop tracking it, history included.

Where to look

  • •The Analytics page: trends over time, KPIs, and per-account breakdowns. Filter by account or avatar to compare presenters.
  • •The agent: ask directly: "which video did best this week?", "how many views did we get this month?", "compare my two accounts." The agent reads the same data and answers with specifics.

Who watches: audience reads

For accounts on a Pamba phone, the phone reads TikTok's and Instagram's own analytics on the account's daily visit: who watched a post (gender, age from 18, top locations, new and returning viewers, followers and non-followers, watch time, and how many were still watching at each second, which shows where viewers drop off), how they engaged with it (its likes, comments, shares and saves, when in the video viewers liked it, the words its comments use most on TikTok, and the share of viewers who liked, commented, shared, saved or reposted it: Instagram's own rates, worked out from the counts for TikTok, which prints none) and who follows the account. A post is first read on the account's next daily visit once it has 200 views, however new it is, so you see how a video held its viewers about a day after posting. A slower post is read three days after posting once it has 100 views. Then it is read again each time its views pass 1,000, 5,000, 10,000 and each doubling after that. A reel removed from an Instagram account is not read again. The account's viewers and followers are read about once a week. TikTok and Instagram show follower details only from 100 followers. Pamba never shows viewers under 18: Instagram's 13-17 age group is left out everywhere.

  • •All Videos, by columns: "Columns" switches the table between Overview (views, likes, shares and the other listing numbers), Retention (how many viewers were still watching 5 and 10 seconds in, halfway and at the end, the skip rate, the share watched and the curve) and Engagement (likes, comments, shares and saves, and the share of viewers who did each), or a mix you pick. Sort by a Retention or Engagement column to rank every post with a read across all your accounts. A post not read yet shows a dash in those columns. TikTok prints no rates or skip rate, so for a TikTok post a rate is the count over the views at the read and the skip rate is the share gone in the first 3 seconds; both carry a *.
  • •A quick look on the Dashboard: in the daily batch and Top videos, a post that is read shows how many viewers were still watching at 0:05. Hover the post for its curve and the other numbers.
  • •A post's details: click any TikTok or Instagram post in Analytics to open its numbers, its audience, when it is read next, and the screenshots the numbers come from in a side panel. Close it and you are where you were; "Open full page" gives the post its own page.
  • •An account's audience: on the Accounts tab, open "Audience" on an account for its viewers and followers, and for a TikTok account its daily health checks with every screenshot.
  • •The agent: ask "who watched my last video?" or "what is the audience of @handle?".

What the numbers mean

  • •Metrics refresh on a regular cadence (roughly hourly for TikTok, and in 15-minute ticks for Instagram/Facebook), so a just-posted video may take a little while to show its first numbers.
  • •Views on TikTok move fastest in the first hours; judge a video after a day, not a minute.

Using analytics to grow

The point of the loop: find what worked and make more of it. Look for outliers, not averages: one video at 10x your median is a signal about hook, topic, or format. Ask the agent to break down what your top videos have in common and to generate new concepts in that direction.

Was this helpful?

Still stuck? Message the team