Exclusion Personas: Defining Who a Campaign Is Not For

Most persona documents describe who to chase. The sharper document describes who to exclude: the buyers who convert, consume the budget, and cost more to serve than they will ever return.

Every mature store has a customer it should stop courting. The discount only buyer who returns half of every order. The one time bulk buyer whose support load eats the margin on the only order they will ever place. The segment that clicks everything, buys at promotion depth, and churns before the second purchase window. Positive personas cannot express any of this, because they are built from who the team wants, and wanting is not the filter that saves money. An exclusion persona is built from transaction evidence instead: a named pattern of behaviour, the cost it imposes, the channels where it gets recruited, and the specific suppression or friction that keeps it out without blocking the good customers standing next to it. It is not a blacklist of people. It is a decision, written down, about which demand the business declines to buy.

Build it from cost, not from vibes

Start with the cohorts whose lifetime value goes negative when fully costed, returns, support contacts, discount depth, and fulfilment exceptions all loaded in. Most teams have never run that costing by cohort, and the first run usually surprises them: a high converting promotional segment that looks like the growth engine turns out to be a subsidy. Name two or three exclusion patterns maximum. A long exclusion list is a sign the analysis wandered from cost into taste, and taste based exclusions block future good customers whose behaviour merely resembles the pattern on the surface.

Each exclusion persona needs its recruitment map. Which campaigns, keywords, affiliates, and placements bring this buyer in? Suppression at the source, bidding and placement choices, is cheaper and cleaner than filtering at the door, because a customer you never recruited never needs to be refused, refunded, or supported.

Related reading on this site: PersonaRail and Your Personas Are a Year Old. Your Customers Moved. cover the neighbouring problems in more depth.

Friction belongs where the pattern is certain

Not every exclusion should be a hard block. Where the pattern is certain and the cost is fraud adjacent, exclude outright. Where the pattern is probable and the cost is margin, use friction instead: no discount eligibility on the first order, returns terms applied as written rather than waived, pre purchase information that answers the questions the costly cohort usually asks after buying. Good customers tolerate proportionate friction without noticing it. The pattern you are pricing out notices immediately, which is how you know the friction was aimed correctly.

Review exclusions like you review prices

An exclusion persona written in one year becomes quiet damage in the next if nobody revisits it. The costly cohort may have been a channel problem that got fixed, or the returns cost that made a segment negative may have fallen after a sizing fix. Put exclusions on the same review calendar as pricing: re cost the cohorts, retire exclusions that no longer pay, and check who the exclusion is catching now, not who it was designed to catch then.

Frequently asked questions

What is an exclusion persona?

It is a documented profile of a customer pattern the business deliberately avoids recruiting or serving on standard terms, built from costed evidence such as returns, support load, and discount depth rather than from preference. It guides suppression, bidding, and friction decisions.

How is an exclusion persona different from a negative keyword list?

Negative keywords are one tactic inside one channel. An exclusion persona is the strategic layer above it: the named pattern, its cost, and every channel where it gets recruited, so the same decision is applied consistently across ads, affiliates, email, and on site offers.

Should exclusion personas ever hard block customers?

Hard blocks fit patterns that are certain and costly, such as abuse or fraud adjacent behaviour. For margin based patterns, calibrated friction usually works better: standard terms applied as written, no first order discount, and clearer pre purchase information.

How often should exclusions be reviewed?

On the same cadence as pricing reviews, at least annually. Cohort costs change when channels, products, or policies change, and an exclusion that no longer pays quietly blocks customers the business would now be glad to serve.