Log inRequest access

Spend spreading beyond the brief

Warningregion_spread5 min read

In shortWhen there is no record of what the targeting resolved to, the honest question is narrower: a brief naming two locations whose spend lands across five regions has a radius wider than the brief — which says something is wrong without being able to say which region is the mistake.

This finding exists because the better one cannot always run. Checking delivery against the regions targeting actually resolved to requires that those names were captured when the campaign launched. For a campaign built before that was recorded, or geo that resolved only to country level, there is nothing to compare against.

Rather than say nothing, it asks the weaker question that is still sound: the brief named two places. Why is the money in five?

What it can and cannot prove

It can establish that the radius is reaching further than the brief does. If somebody wrote down two neighbourhoods and spend is material across five regions, the targeting is wider than the intent, whatever the reason.

It cannot tell you which region is the wrong one. A brief naming two suburbs might legitimately deliver across three districts that surround them. The count is a signal that the shape is off, not a verdict on any particular row.

A weaker check that says what it cannot prove is worth more than a strong one that guesses.

Why small regions are ignored

Delivery always scatters. A campaign will show a trickle of spend in places nobody intended simply because people travel and signals are imperfect. Counting those would make every campaign look broken.

So a region has to carry at least a twentieth of the spend to count at all, and it has to have been given enough budget to expect a couple of outcomes before its performance means anything. That second floor is the same reasoning that stops a campaign with three hundred rupees of spend and no leads being called a failure: absence of evidence needs enough budget behind it first.

What to do with it

  • Treat it as a prompt to open the campaign, not as an answer. Read the region list against what was actually briefed.
  • If the extra regions are the ring around the intended ones, that may be the radius doing its job.
  • If they are somewhere nobody would drive from, tighten the radius rather than excluding one region at a time.
  • Where it matters, record what the targeting resolved to at launch. Then the exact check becomes available and this one stops being needed.

When Admetriq raises it

The actual rule, so you can hold this page against what you see rather than take its word for it.

  • Counts the regions carrying material spend and compares that against the number of locations the brief named. Fires when delivery is spread across more regions than were asked for.
  • Only runs when the stronger check cannot: if the campaign recorded which regions its targeting resolved to, that comparison is exact and this one is skipped.
  • Material means at least 5% of spend (MATERIAL_SHARE = 0.05), and a region needs enough budget behind it to have expected around two outcomes before its numbers count (MIN_EXPECTED_OUTCOMES_PER_SEGMENT = 2.0). Without that floor every small region on every campaign reads as a failure.
  • It also requires the brief to have named locations at all. No locations named is a legitimate reason not to fire, rather than a reason to guess.

Most often mistaken for

Delivering outside the target arearegion_out_of_scope

The stronger version of the same worry. Where the campaign stored the regions its targeting resolved to, spend can be checked against that list directly and the finding can name exactly which region was never covered. This one can only say the spread is wider than the brief, not which part of it is the mistake — so treat it as a prompt to look, not as a verdict.

All findings · Longer writing