What to Decide Before You Launch a Campaign

Most campaigns are lost in the brief, not in the execution. Before any budget moves, four things have to be settled in order: the buying decision you are trying to change, the one segment that decision matters to, the proof that makes your claim credible, and the number that tells you to stop.

Everything else, the channel mix, the creative concepts, the flighting calendar, is downstream of those four. Teams that skip them end up running six channels at once and learning nothing from any of them. Teams that settle them first can argue about channels productively, because there is finally a standard to argue against.

Here is the uncomfortable part. Analysis of the IPA Databank 1998 to 2025 by Les Binet and Will Davis, published by the IPA in 2025 as "Go Big or Go Home," found budget explains 89% of the variation in incremental profit across award-winning campaigns, against 11% for ROI. If the level and concentration of spend is the dominant lever, then the priority-setting that decides where that spend lands is worth more than every optimization that comes after it. You cannot optimize your way out of a budget split six ways.

1. Start with the decision you want to change, not the channel list

A channel plan presented before a decision statement is a budget request wearing a strategy costume. It looks like thinking. It is really a list of places money can go.

The gate is one sentence: who changes their mind, from what, to what.

"Shopify operations leads at 8-figure DTC brands stop believing their retention problem is a discounting problem and start believing it is a post-purchase experience problem." That is a decision statement. It names a person, a current belief and a target belief. Every creative brief, every landing page, every channel choice can now be tested against it.

Compare that to "increase awareness among our target audience in Q3." Nothing can be argued with. Nothing can be disproven. Any channel qualifies, which is exactly why these briefs end up with six of them.

Write the sentence first. If the team cannot agree on it in a room, the campaign is not ready for a budget, it is ready for another working session. That is a cheaper problem to have in week one than in week nine.

2. Which segment actually has to move for this to pay back?

One campaign, one segment. Not because other segments do not matter, but because a message tuned to three audiences is tuned to none of them, and you will never know which of the three responded.

Pull the segment from first-party evidence, not from a persona deck written eighteen months ago. Your own data already has the answer in it:

• Which companies are returning to your pricing and services pages without booking anything
• Which contacts opened three emails in a row and then went quiet
• Which deal types close fastest, and which drag for two quarters
• Which inbound forms produce conversations your sales team actually wants

That behavioral evidence beats a demographic sketch every time, because it is a record of what people did rather than a guess about who they are. A persona says "marketing directors at mid-market B2B firms." Your account data says "four people from the same company read the AI visibility page this month and nobody booked." Those are not the same quality of input.

If a segment cannot be identified in your own records, that is a finding. It usually means the campaign is being built for a market you have not sold to yet, which is a different and more expensive project than the one being briefed.

3. What proof makes the claim credible, and do we have it?

Creative polish ranks below proof. Our position: a beautifully produced spot making an unsupported claim will underperform a plain one showing a real outcome, because the buyer's first question is not "is this nice" but "is this true."

Inventory the proof before you brief the creative:

• Named outcomes with numbers attached
• Demonstrations of the thing working, not descriptions of it
• Reviews and their current quality, including the bad ones
• Customer evidence you have permission to use

Then audit the foundation that the proof will land on. Positioning, offer, page conversion, review quality. More media accelerates a broken foundation faster than a good one. If the landing page converts at a third of what it should, paid traffic will simply reveal that failure at a higher velocity and a higher cost.

This is the step most briefs skip, and it is the one with the highest return. Fixing a conversion path before launch changes the economics of every channel in the plan at once. Fixing it after launch means you paid full price to learn what an audit would have told you for free.

4. How many channels can this budget actually afford?

Fewer than the brief asks for. On a lean budget, one primary capture channel plus one content engine will outperform six underfunded platforms, and the reason is mechanical rather than philosophical. Paid platforms need a threshold of conversion volume before their optimization has anything to learn from. Split a modest budget six ways and none of the six crosses it. You do not get six experiments. You get six inconclusive ones, plus six sets of reporting to maintain.

Concentration is not a compromise forced by a small budget. It is the correct strategy for a small budget.

So the brief should name what gets cut, not just what gets funded. A plan that adds channels without removing any is not a plan, it is a wish list with a media schedule.

5. Demand creation and demand capture are two budgets

Search captures intent that already exists. Video, social, category education and thought leadership create the intent you will capture next quarter. Blending them into one "full funnel" line item makes both impossible to judge, because capture always looks more efficient in the short window and creation always looks wasteful.

For sizing the creation half, excess share of voice is the most defensible input available. Analysis of IPA databank effectiveness data by Les Binet and Peter Field, published by LinkedIn, found that a 10% excess share of voice leads to market share growth of 0.7% per year for B2B brands, against 0.6% in B2C. The relationship runs both ways: brands setting share of voice above share of market tend to grow, and those setting it below tend to shrink, at a rate proportional to the excess.

The same principle now extends to how AI engines assemble shortlists. Our read: answer engines synthesize from the sources that cover a topic with substance and consistency. A brand that only ever buys capture has nothing for those systems to draw on, and gets left out of the consideration set before a human ever sees it. Treating AI search visibility as a demand creation line, rather than a technical afterthought, is the current version of this problem.

6. Write the number that tells you to stop, before you launch

Three lines in the brief, agreed by whoever holds the budget:

1. The metric. Qualified pipeline, cost per qualified opportunity, new customer revenue, contribution margin. Not impressions, not reach, not engagement rate.
2. The threshold. The specific number that counts as working, and the number that counts as failing.
3. The time to learn. How long until there is enough volume to judge either way, decided from expected conversion volume rather than from the length of a quarter.

Then name the decision that follows each outcome. Hits the threshold, we scale and here is the next budget. Misses it, we stop and here is what we try instead. Lands in between, we run one more cycle with one variable changed.

Written after the first report, kill criteria become a negotiation about whose work is being questioned. Written before launch, they are a shared agreement that makes stopping a normal outcome instead of a failure. That single change does more for campaign discipline than any dashboard.

The order is the argument

Decision, segment, proof, stop number. Each one gates the next. You cannot pick a segment without a decision to change, cannot choose proof without a segment to convince, and cannot set a threshold without knowing what the proof is supposed to do.

Run them in order and the channel plan writes itself in about an hour, because most of the options have already been eliminated on the evidence.

At Payani Media, this is the work we do in discovery and audit before anything gets built: the decision statement, the segment pulled from first-party behavior, the foundation audit across positioning, offer and conversion path, and the governance framework that sets thresholds before spend starts. If your next campaign brief is currently a channel list, that is the conversation to have first.

Frequently asked questions

Should the channel plan ever come first?

No. A channel plan presented before a decision statement is a budget request wearing a strategy costume, because any channel qualifies when there is no standard to test it against. Settle the buying decision, the segment, the proof and the stop number first. The channel choices narrow themselves once those four are fixed.

How many channels should a lean budget run?

One primary capture channel plus one content engine. Paid platforms need a threshold of conversion volume before their optimization has anything to learn from, and splitting a modest budget six ways means none of the six crosses it. That makes concentration the correct strategy for a small budget rather than a compromise.

Why separate demand creation from demand capture in the budget?

Because blended into one line item, neither can be judged fairly. Capture always looks more efficient in a short window and creation always looks wasteful. Search harvests intent that already exists, while video, social and category education build the intent you harvest next quarter, so they need separate budgets and separate success definitions.

When should kill criteria be written?

Before launch, in the brief, agreed by whoever holds the budget. Three lines: the metric, the threshold that counts as working or failing, and the time to learn based on expected conversion volume. Written after the first report, kill criteria turn into a negotiation about whose work is being questioned.

What should be audited before any spend starts?

Positioning, offer, page conversion and review quality. More media accelerates a broken foundation faster than a good one, so a weak conversion path simply fails faster and more expensively under paid traffic. Fixing it before launch improves the economics of every channel in the plan at once.