A programmatic advertising channel strategy should be scored, not argued

What to take away

Which channels belong on the shortlist?

A channel strategy starts with a shortlist, not a spreadsheet. A typical shortlist covers five families: paid search, paid social, retail media, video and audio, and open web display.

Each family behaves differently. Search captures existing demand, while social creates it. Retail media sits closest to the point of purchase.

The shortlist should follow the audience, not the sales pitch. Write down where the audience spends time, then check whether each platform reports that reach credibly.

Keep the shortlist to five or six names. A longer list spreads creative and reporting effort thin, and it usually means nobody has settled the objective yet.

Video and audio carry higher production costs than display, so treat them as separate line items rather than add-ONS.

For the wider planning sequence around this choice, the programmatic advertising strategy and planning guide sets objectives, budgets and measurement in order.

How do you score a channel before committing budget?

Build a rubric with five criteria and fixed weights. The weights below are illustrative and sum to 100, so adjust them to your category once and leave them alone for two quarters.

Stability matters more than precision, because the point is comparing channels against each other.

Score each criterion from 1 to 5, then multiply by its weight and add the totals. The ceiling is 500; the floor is 100.

Treat a total above 400 as ready to scale. A score between 300 and 400 justifies a controlled test with a fixed budget. Below 300, park the channel until the evidence changes.

Two people should score independently, then reconcile any difference above one point. Reconcile by returning to the evidence, not by averaging the two numbers.

Permissions need their own line in the rubric. Electronic marketing and cookies sit under ICO guidance on direct marketing and privacy and electronic communications, which shapes what you may collect and reuse.

The rules have moved on since the Data Protection Act 1998, now largely replaced by the 2018 Act and UK GDPR. Score channels against current law rather than habits formed under the older statute.

Smaller advertisers rarely have a research team. The IAB UK SME toolkit is written for smaller businesses moving into digital and programmatic buying.

Agree the weights before anyone writes a media plan, then record scores in the programmatic advertising planning template so owners, dates and evidence sit in one place.

What separates a defensible score from a preference?

A score is only as good as the evidence behind it.

Ask for a placement report, a fee schedule and a named owner. If an owner cannot supply them, treat the gap as a score of 2 and move on.

A floor rule keeps the argument short. Any criterion scoring 1 caps the channel at pilot spend until someone fixes the underlying problem.

For example, a team paying £400 a month for a channel whose click-through rate it cannot verify should buy a pilot, not a full campaign.

Scores also need a review date. Put it in the plan beside the budget line, so the channel is judged before anyone renews it. Evidence that predates the last quarter should be refreshed before it counts.

Without a date, the rubric quietly becomes decoration.

Common questions

How many channels should a programmatic plan fund?

Start with three funded channels and keep the rest as pilots. Fewer channels mean clearer reporting, less creative strain and a cleaner split of responsibility.

Who should own the scoring?

One person owns the rubric, but at least two people score each channel independently. The owner reconciles gaps above one point and writes down the reason.

How often should the weights change?

Review them each quarter and change them at most twice a year. Frequent changes destroy comparability, which is the main benefit the rubric offers.

What happens when two channels tie?

Fund the one with the lower operational load first, then revisit it at the next review. A tie often means the scores are too coarse, so consider a half-point step.