Audience Exclusions And Wasted Spend

An account that has been optimised for a year is usually still paying to reach people it should never have paid for.

Targeting gets attention because it feels like strategy. Exclusions get ignored because they feel like admin. The return on the second is frequently higher, because removing waste improves the signal the bidder learns from as well as the cost.

The exclusions that pay for themselves immediately

Existing customers, on prospecting campaigns. An acquisition campaign serving ads to people who already bought is paying twice and polluting the conversion signal. This is the most common and most expensive omission in accounts of every size.

Current leads in the pipeline. Somebody who enquired last week does not need a cold ad. They need the follow-up sequence.

Recent converters, for a window that matches the actual repurchase cycle rather than a default.

Job seekers and competitors, who click, cost money and never buy. Both are identifiable by behaviour and by search term.

Your own team and contractors, who inflate engagement metrics on small audiences.

On search, the equivalent is the search terms report

Every account has a list of queries it is paying for that nobody would have chosen.

Free, cheap, DIY, jobs, salary, course, how to do it yourself, and the competitor brand terms that convert at a tenth of the rate. Each one is a negative keyword and each one takes thirty seconds to add.

The report has to be read on a schedule rather than when something looks wrong, because the terms change as the match types broaden on their own.

Placement exclusions on display and video

A display or video campaign left on default placements will serve across a long tail of apps, games and low-quality sites that generate impressions, clicks from mis-taps, and no outcomes.

The fix is a placement report read monthly, with the obvious waste excluded. On any account spending meaningfully on display, that single pass is usually a double-digit percentage of the budget.

Mobile app inventory is the standard first exclusion for a business selling anything considered.

Geography is where the quiet waste sits

An account targeting a metro area will, on default settings, also serve to people outside it who show interest in it.

For a business that physically delivers a service, that is pure waste. The setting that controls it is not the obvious one, and the default favours reach over relevance.

Same problem in reverse for a business with a service area that does not match a city boundary. Radius targeting around a point serves a circle, and a circle rarely matches where a business actually works.

Why exclusions improve performance beyond the saving

This is the part that gets missed.

Automated bidding learns from conversions. Feed it conversions from people who were always going to buy, and it optimises toward finding people who were always going to buy, which it cannot do at scale.

Remove existing customers from the signal and the bidder starts learning what an actual new customer looks like. The improvement shows up as better acquisition, not just lower spend.

The maintenance rhythm

Customer lists uploaded and refreshed, because a list from last year is excluding people who are no longer customers and failing to exclude the ones who are.

Search terms reviewed on a fixed interval. Placements reviewed monthly on anything running display. Exclusion lists audited quarterly for rules that no longer apply, because an exclusion that made sense two years ago can be blocking a segment that now converts.

The test to run first

Pull the last ninety days and split conversions by whether the person was already a customer.

If a meaningful share of attributed conversions came from existing customers, the account's reported return is overstated and the acquisition performance is worse than the dashboard says. That number is usually the most useful thing an audit produces.