10 media planning mistakes that quietly wreck campaigns

1. Planning to the budget instead of the objective

The brief says $400K, so the exercise becomes distributing $400K rather than asking what the objective costs. Sometimes the objective needs $250K done well; sometimes it needs $700K and the honest plan says so. The fix: cost the objective first, then reconcile with the budget in the open. If the money cannot buy the outcome, the plan should scope the outcome down explicitly rather than quietly under-deliver.

2. Audiences defined as demographics

'Adults 25-54' is not an audience, it is a census band containing your buyers, their neighbours and everyone in between. Demographic-only definitions waste reach on people who were never in the market and price every channel wrong. The fix: define audiences by motivation and behaviour, then map them to segments a platform can target - the working method is in audience personas in media planning.

3. Channel splits inherited from last year

Precedent smuggles last year's objective, last year's prices and last year's platform mix into this year's plan. The split survives because relitigating it is uncomfortable, not because it is right. The fix: rebuild from the three real inputs - objective, cost to reach the audience per channel, historic response - as covered in how to allocate media budget across channels. Precedent is one data point, not a decision.

4. Ignoring the reach and frequency maths

Plans that never state a reach target or a frequency cap end up buying the same heavy users repeatedly while the intended audience stays unreached. It looks fine in spend and terrible in incremental outcome. The fix: put reach, target frequency and the cap on the plan per channel, and check delivered frequency in-flight rather than in the wash-up.

5. Flat flighting against unflat demand

Even weekly spend across a quarter is the default because it requires no thought. Demand has shape - seasonality, paydays, launch moments, competitor noise - and flat flighting overpays quiet weeks and underweights the ones that matter. The fix: shape the flighting to demand, and let the pacing target follow that shape so alerts measure drift from the plan, not from a straight line.

6. The plan dies at launch

The master mistake. The deck gets approved, campaigns go live, and the plan is never opened again. From that moment nobody is comparing actual to planned, so every drift compounds silently until month-end. The fix: treat the approved allocation as a pacing baseline with a cadence and an owner. Live pacing against plan makes this automatic; the discipline is deciding it will happen before launch. The mechanics are in campaign budget pacing, explained.

7. Measuring what is easy, not what matters

Clicks and CTR get reported because platforms serve them up, not because the objective asked for them. A campaign scored on convenient metrics optimises toward them, and the objective drifts out of view. The fix: set the primary KPI from the objective at planning time, add one guardrail metric that stops it being gamed, and refuse to let dashboard availability rewrite the framework.

8. No experimentation reserve

When 100% of budget goes to proven channels, the account learns nothing, and in two cycles the 'proven' list is stale. The channels that will matter next year never got a funded test this year. The fix: ring-fence roughly 10% for experiments with success criteria written in advance, and protect it from the mid-flight raids that incumbent channels always attempt.

9. Knowledge that lives in one workbook

Every planner has their own spreadsheet, their own conventions, their own logic - and when they leave, the account's institutional memory leaves in their notice period. Agencies feel this hardest: quality varies by author and onboarding takes months. The fix: one shared structure for how plans are built, stored somewhere the team works, not in personal files. It is the core argument for a standardised planning workspace for agencies.

10. Defending the plan with opinion

A split presented without reasoning invites a taste-based debate, and taste-based debates are won by seniority, not evidence. The plan then mutates in the meeting room. The fix: attach the rationale to every weighting at planning time - what the objective demanded, what the audience costs to reach, what history showed. Recommendations with data behind them get challenged once and then survive; the same logic applies in-flight, where AI recommendations ranked by predicted impact give reallocation decisions the same evidentiary footing.

The pattern underneath the ten

Read the list again and it collapses into one sentence: decisions made by default instead of on evidence, and never revisited. The budget defaulted to distribution, the audience to demographics, the split to precedent, the flighting to flat, the measurement to convenience. Each default was individually small. Campaigns are wrecked by the compound interest. The antidote is equally unglamorous - write the reasoning down, make the plan buyable, keep it alive after launch - and it is worth noting that the mechanical parts of that discipline are now largely automatable, starting from the brief itself.