Editor's note: revised June 19, 2026. Originally published November 15, 2012. Updated for current AI Communications and measurement realities.
Good marketing comes from good understanding — knowing what makes the brand distinctive and knowing why the customer should care. Five questions structure any marketing plan worth executing. The questions have not changed since the original 2012 version of this piece. The answers have changed substantially as the channel mix and the measurement model have evolved.
1. What's in it for the customer
How does the product or service benefit the customer specifically? What problem does it solve? What advantage does it offer over the alternatives the customer is actually considering? The marketing communicates customer benefit in language the customer would recognize as describing their own situation. Generic benefit statements that any competitor could claim do not move conversion.
What to do: List the customer benefits, rank them by importance, and write the messaging in language the customer uses — not language the brand uses internally. Push the messaging through sales, customer service, and marketing consistently. Anywhere the customer hears the brand, the benefit framing should be the same.
2. What's in it for the company
The largest mistake brands still make in 2026 is launching marketing campaigns without specifying what success looks like. Vague goals — "more sales," "build brand awareness," "drive engagement" — produce vague campaigns. Sharp goals produce sharp campaigns.
What to do: Set specific, measurable targets — number of leads, percentage lift on conversion, revenue contribution by channel. Use prior performance data to set realistic baselines. The discipline of quantification is what separates campaigns that scale from campaigns that get rationalized after the fact.
3. A timeline that matches the work
Marketing campaigns without timelines drift. The category leaders in 2026 run campaigns with phased timelines — short-term tactical executions inside longer-term strategic arcs — and review milestones against the original plan. Open-ended campaigns produce open-ended results.
What to do: Set a campaign duration tied to the goal. Break long-term goals into shorter milestone checkpoints. Treat the timeline as a hypothesis to be tested, not a commitment to be defended — when the data says the timeline was wrong, adjust openly rather than pretending the original plan is still in motion.
4. The audience, specifically
Where does the audience actually research products? Which channels, which voices, which AI engines? The 2012 version of this question pointed at Facebook, Twitter, and LinkedIn. The 2026 version adds TikTok, YouTube, Instagram, the creator-driven discovery layer, and the AI engine retrieval layer above all of those. A growing share of category research now happens inside ChatGPT, Claude, Gemini, and Perplexity before the customer reaches any social or search surface at all.
What to do: Conduct audience research that maps actual behavior, not assumed behavior. Audit what the AI engines say about the brand and the category. Identify which voices in the category have the audience's trust. The brands that win the audience layer in 2026 are the brands that have mapped not just demographic data but the discovery infrastructure the audience actually uses.
5. How the campaign moves the business
Campaign goals that do not connect to business outcomes are decoration. The marketing functions that retain executive credibility in 2026 are the ones that can trace campaign activity to measurable downstream effect — pipeline contribution, customer acquisition cost, lifetime value, retention rate, brand-search lift. The functions that cannot make that connection are the ones being asked to defend their budgets every quarter.
What to do: Map each campaign goal to a downstream business metric. If 500 new newsletter subscribers is the goal, name the conversion path that connects subscribers to revenue. If brand awareness is the goal, name the measurement signal — brand search lift, citation share inside AI engines, share of voice in earned media — that operationalizes "awareness" into something the business can actually track.
The 2026 measurement layer
Three structural shifts have changed how marketing plans get measured since 2012.
Attribution has fragmented. Last-click attribution no longer maps cleanly to multi-channel discovery. The brands operating well in 2026 use multi-source modeling combining first-party data, media-mix modeling, and incrementality testing. The brands still optimizing against last-click are reallocating budget on data that no longer reflects reality.
AI engines have become a measurement category. Citation share inside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews now anchors a meaningful share of category discovery. The marketing plans that ignore the AI engine layer are operating with a measurement blind spot the competition is increasingly closing.
Retention math matters more than ever. The brands that boost durable returns are weighting investment toward retention infrastructure — loyalty, lifecycle email, post-purchase experience — and treating acquisition as the engine that feeds retention. The acquisition-only model is producing declining returns across categories.
The takeaway
Five questions structure any marketing plan worth executing. The questions hold. The answers have evolved as channels, measurement, and AI engine discovery have rewritten the operating environment. The brands that ship plans answering all five clearly are the brands operating with the discipline the category increasingly requires. More across the EPR marketing, GEO, and archives.
Related from the EPR archive: Top 10 Factors that Influence Purchase Decisions · Five Things to Avoid in Email Subject Lines · Five Things That Make a Better Business Leader
