Some campaigns now live without ever appearing on the competitor's website. A match-day offer opens in an X post, a new product is announced in a YouTube video, a seasonal campaign starts on Instagram. Tracking that watches only the website sees these moves late, sometimes never.
This post covers how to set up competitor social media tracking and the most common measurement mistake: recording a period you could not measure as zero.
Which accounts, how often
List the official accounts for each competitor brand one by one: YouTube channel, Instagram, TikTok and X. If the same brand runs regional or product-specific second accounts, record them separately. Collapsing them into one row hides which audience the brand is talking to.
Pick the frequency to match each platform's rhythm. Several posts a day are normal on X, so you need an interval of a few hours. YouTube, Instagram and TikTok post less often; a few reads a day are usually enough.
What to record
- Post time: the platform's publish time, not the time you saw it.
- Type: video, post, short video, tweet.
- Text and link: does a campaign name or code appear.
- Engagement: views, likes, comments; whichever the platform exposes publicly.
- Read status: could the account be read on this tour.
A day you could not read is not zero posts
Public profiles cannot always be read. The platform may not answer for a while, the account may go private, the page layout may change. You end up with a result that says no posts found on this account, and it slips into the table as 0.
But 0 is a measurement: it says the competitor did not post that day. The right record for a day you could not read is unmeasured. When the two get mixed, the weekly digest says the competitor went quiet and the team decides accordingly. In reality the competitor may have been having its busiest week.
The same logic holds for long gaps. If you could not read an account for two days, the next read may not tell you on which day the visible posts went out, because a profile only shows the latest posts. Marking that window as incomplete is more honest than writing an exact number.
The first read dumps the history
On the day you start tracking an account, the profile shows dozens of older posts. Counting them as new posts inflates the first week's digest. Store the first read as the baseline and count new posts only after it.
With Adversee
Adversee lets you add YouTube, Instagram, TikTok and X accounts to every competitor brand and reads public posts without signing in. An unreadable account shows as unreadable, not as having no posts; in digests an account that could not be measured says unmeasured, not 0.
More on measurement traps in competitor monitoring: The engineering problems behind monitoring competitor sites
Start tracking competitor moves automatically.