A Baseline Built on Bought Followers Was Never Clean

A case study reports a before and an after, and almost all the scepticism goes to the after. The before number looks like a timestamp, a neutral place where the story starts. It is not neutral. Somebody built it, and there is more than one way to build it.

The ‘before’ number is not just a timestamp, it is a manufacturing history

A baseline is everything that happened to the account before that day, and two of those things leave a number larger than the audience it represents. The first is buying. Purchased followers arrive as a fulfilled order, not a growth process, and the result has a characteristic shape: a block of accounts added across a short window, hours to a few days, rather than a trickle over months. That is a described mechanism, not a measured statistic: the shape of delivery rather than growth.

The second is the follow-to-enter giveaway, which is public, legitimate, and contaminates a baseline anyway. A prize is offered, entry requires a follow plus a tag or a share, and people follow. What it recruits is people who wanted the prize, for whom the follow is the price of a lottery ticket, not interest in the account.

A genuine viral moment also produces a sudden jump, which is where careless scepticism goes wrong in the other direction. The distinguishing feature is not how steep the line is, it is what moves alongside it. When a post travels, followers are a side effect of attention, so comments and shares rise in the same window. A follower jump with nothing else moving is a different event.

A hypothetical baseline chart with the jump built in

Everything below is invented and describes no real account, agency or case study.

Picture a hypothetical account sitting near 2,000 followers for months, drifting by tens. Over four days it climbs to a hypothetical 14,000. Then it flattens for three weeks, at which point a campaign begins. The campaign ends with the account at a hypothetical 40,000, and the case study reports it that way: grew to 40,000 followers.

Run the arithmetic twice. From the pre-jump 2,000, that is a twentyfold increase, roughly 1,900 percent growth. From the post-jump 14,000, the same endpoint is under threefold, roughly 186 percent. Same campaign, same ending number, two unrecognisably different claims.

The point is not that one figure is dishonest. It is that a choice was made and the document does not say a choice existed. Again: invented numbers, not anyone’s reported result.

Why the reader usually cannot see the jump directly

Case studies publish totals, not histories: a before figure, an after figure, a date range. The pre-campaign months, where contamination would show, are almost never included, so the shape of the baseline is invisible unless you look elsewhere.

Where any pre-campaign history is visible, the test is shape rather than size. Organic growth draws a slope. It may be shallow or steep, it may bend, but it accumulates. Bought and contest-driven growth draws a step: a near vertical rise, then a flat line, because the mechanism stopped.

Engagement is the cross-check that survives when the history does not. Followers acquired by purchase or by prize were not recruited by the content, so they have little reason to interact with it later. A count that jumped without a matching jump in likes, comments and shares is a checkable inconsistency.

What to check when a ‘before’ number looks off

None of this needs account access. All four can be asked of the published document.

• Was there a documented promotion, giveaway or contest in the weeks before the reported start? Public giveaways leave traces.
• Does the case study show a longer pre-campaign period, or only a single before and after snapshot? A snapshot cannot show a step.
• Does engagement move in the same direction and at the same time as the follower count, or does the follower line move alone?
• Is the before number unusually round, or unusually large for an account of that stated age and category? Neither settles anything, but both are reasons to ask for the chart.

A contaminated baseline breaks every number calculated from it

This piece is about what the before number is made of. That is a different question from when the before point itself is a choice, which is about where on the timeline it is taken. Composition and timing move the same figure independently.

Composition matters because the before number is a denominator. Percentage growth and multiplier claims divide by it. Engagement rate divides by a follower count carrying the same contamination, where an inflated denominator pushes the reported rate down. Every derived figure inherits the error and none of them announce it, which is why a number is only as good as what backs it up.

Read enough of them and the shapes start repeating

Read enough results claims and a step in a flat line starts looking like a disclosure left out. Browse the case study library, or submit your own if your agency has published work you stand behind. If this is new, start with how to read a case study without being misled.

FAQ

Is a sudden jump in follower count always a sign of bought followers?

No. A genuinely viral post produces a sudden jump too, and treating every vertical line as fraud will make you wrong often. The check is whether anything else moved: a real spike carries a matching rise in comments and shares, because the followers came from attention. A manufactured one shows the follower line moving alone.

Can I check an account’s follower history myself if I only have the published case study?

Usually you are limited to what the document chose to disclose. If the account is public, a longer history is sometimes visible on the profile or preserved in an archive, which is worth a look. Otherwise the shape test and the engagement cross-check are the tools you actually have.

Does a giveaway-driven follower spike ever turn into a real audience?

Sometimes, but not by default, and no figure is claimed here for how often. Those followers were recruited by the prize rather than the content, so nothing gives them a reason to stay engaged once the contest ends. Look for engagement holding up afterwards.

Sources

• The ‘Before’ Picture in a Case Study Is Also a Choice, fetched 2026-09-07.
• A case study number is only as good as what backs it up, fetched 2026-09-07.
• How to Read a Marketing Case Study Without Being Misled, fetched 2026-09-07.