A media plan arrives as a spreadsheet with channels down the side, months across the top, and a total at the bottom that somebody has to approve. It is a forecast dressed as a budget, and almost everything worth knowing about it is in assumptions that do not appear anywhere on the page.
Here is how to read one properly.
The lines, and what each commits you to
Channel split. How much goes to search, social, video, and anything else. The question to ask is why this split — if the answer is "it is what we usually do", the plan was not built for your business. Search captures demand that exists; social creates it. The right ratio depends on whether people are already looking for what you sell.
Monthly phasing. Flat monthly numbers are a sign nobody thought about seasonality. In India, the festive window is the most expensive and highest-intent stretch of the year and a flat plan either underfunds it or overpays through the quiet months. The same applies to the UK calendar and the US Q4 run if you sell into those markets.
Prospecting versus retargeting. Retargeting always looks better per rupee because it is selling to people who already came. A plan weighted heavily towards it will report excellent numbers while the business stops growing. Ask what share of spend is reaching people who have never heard of you.
Creative production. If there is no line for it, somebody is assuming your existing assets will last the quarter. They will not. Ask how many new concepts the plan assumes per month and who is making them.
Fees. Agency fee, tool costs, and anything percentage-based should be visible separately from working media. A plan that buries the fee inside the channel lines makes it impossible to see what is actually reaching the auction.
The assumptions nobody writes down
Every forecast of leads or sales rests on three numbers: a cost per thousand impressions, a click-through rate, and a conversion rate. The plan shows you the output. Ask for the three inputs.
Then ask where each came from. "Benchmark" means somebody used an industry average that may not resemble your account. "Your historical data" is far better, and you can check it. If the forecast assumes a conversion rate higher than your site has ever achieved, the plan is a wish.
The second hidden assumption is budget sufficiency. Each campaign needs enough volume to clear the learning phase — roughly fifty optimisation events a week on Meta. A plan that spreads a modest budget across six campaigns in four markets has guaranteed that none of them stabilises. Count the campaigns and divide.
Five questions before you sign
- What are the three input assumptions behind the forecast, and where did each come from?
- Which line would you cut first if the budget dropped by a third? The answer tells you what they actually believe in.
- What has to be true for this to work? Tracking in place, creative delivered on time, the page fixed — these dependencies belong in the plan, not in a later excuse.
- How many optimisation events a week does each campaign expect? If nobody has calculated it, the structure is a guess.
- At what point do we stop? Agree the figure and the timescale that would mean this is not working, before any money moves.
What a good media plan looks like
Fewer lines than you expect. Concentrated spend rather than a sprinkle across every channel someone has heard of. Phasing that reflects the real calendar. A visible creative line. Stated assumptions. And a short paragraph about what the first month is trying to learn rather than earn.
Once it is running, the same discipline applies to the report that comes back — how to read an ads report covers the other half of this, and how to split a marketing budget in India covers the allocation question underneath it.
Questions people ask before hiring
What should a media plan include?
Channel split with reasoning, monthly phasing that reflects real seasonality, a visible prospecting-to-retargeting ratio, a creative production line, and fees shown separately from working media so you can see what actually reaches the auction.
How do I check whether a forecast is realistic?
Ask for the three inputs behind it — assumed cost per thousand impressions, click-through rate and conversion rate — and where each came from. A forecast built on industry benchmarks rather than your own historical data is a guess with a decimal point.
What is wrong with spreading budget across many campaigns?
Each campaign needs enough conversions to leave the learning phase, roughly fifty a week on Meta. A modest budget split across six campaigns in four markets guarantees none of them stabilises, so none of the results can be read.
What should I ask before approving a plan?
Which line they would cut first if the budget fell by a third, what has to be true for the plan to work, and what result at what date would mean it is not working. Agree the stopping point before any money moves.