Marketing in Practice: Build a Review Loop

The difference between marketing in practice and marketing plans is whether anything verifiable is left behind. A well-written plan, executed properly, followed by "results were positive, keep optimising" teaches nothing. Real capability means finishing a campaign with three reusable judgements in hand.

Here is the simplest loop that fits on one page.

1. Four Stages

Hypothesis, action, data, conclusion. Drop one and the campaign becomes pure consumption.

A hypothesis must be falsifiable: "moving the enquiry button from the footer to the first screen will raise enquiries by thirty percent". "Improve conversion" is not a hypothesis because it cannot be disproved.

An action must limit variables. Change button position, copy and colour at once and a better number tells you nothing about which change worked.

Data needs instrumentation beforehand. Hunting for it afterwards usually yields only totals, without channel or page breakdown.

A conclusion must state what happens next, not what happened. If it cannot produce a next action, it is not a conclusion.

2. A Four-Column Review Sheet

The data column is the one that gets fudged. Always state the baseline: last month, same period last year, or an unmodified control group. Without a baseline you can read a change backwards.

3. Four Break Points

Too broad a hypothesis. One action tries to solve three problems and validates none.

Mixed variables. Structure, copy and visual changed together, so the conclusion becomes "overall improvement" and the next round starts from zero again.

Unpaired data. Counting enquiries without counting quality makes a channel look effective while revenue stays flat.

Conclusions left unwritten. The meeting ends with warm agreement recorded as "lively discussion", and two weeks later nobody remembers the decision.

4. Running the Review Meeting

Send the one-page sheet in advance and ask every attendee to bring one objection. Entering the room with disagreement is more productive than entering with consensus.

Discuss three questions only: which hypothesis was disproved, which number was unexpected, which variable to change next. Park everything else.

Send conclusions within twenty-four hours, each with an owner and a verification date. A conclusion without an owner expires automatically.

Many companies turn this meeting into a reporting session: each department presents what it did, everyone leaves. Restrict the agenda to disproved hypotheses and the meeting usually halves in length while doubling in information density.

5. Three Accelerators

Keep an action log. One row per action: date, variable, expectation, actual. After a year you own an experience curve.

Keep failures. Failed actions carry more information than successes because they eliminate a path.

Review across quarters. Put two quarters of conclusions side by side; the hypothesis that keeps repeating is the one to write into standard process.

Capability in marketing comes from judgements retained, not projects completed. Ten projects without conclusions are worth less than three projects with nine conclusions.

6. FAQ

What if a small team has no time? Compress the review to thirty minutes and fill only the four rows. That costs less than repeating the same mistake.

What if data is unavailable? Do what can be measured. For offline activity, create a dedicated entry point or a distinct script so enquiries can be attributed.

Does personal branding count? Yes. Publishing and client conversations follow the same loop, with softer data.

What if the conclusion contradicts leadership? Publish the measurement basis. Move the disagreement onto definitions, where it is usually resolvable.

Related

• See also: prioritising a promotion plan
• See also: building a content calendar

GrowLab focuses on integrated content marketing, covering brand strategy, website development and search marketing (SEO/GEO). Contact us for an assessment of your own promotion plan.