The Rebel Marketer Method: Challenge Assumptions and Test One Decision

The Rebel Marketer method is to question an inherited assumption, design a bounded experiment and publish what the evidence actually supports. Rebellion is a position about how to work, not evidence that a tactic will succeed.

Updated 3 October 2026. Correction: the former introduction claimed years of work with startup and fintech clients without supporting records. That personal track record is removed. This article explains the project's proposed working method.

Start an assumption ledger

For each campaign, write the decision, underlying assumption, evidence currently available, cost of being wrong and next test. Keep assumptions separate from facts.

Example decision: publish another tutorial. Assumption: readers understand the product but need onboarding help. Evidence: three messages mention setup difficulty. Limitation: three messages do not establish the needs of every visitor.

A more useful next action may be asking a few readers to attempt the task and record where they become confused. A new channel is not always the answer.

Choose tactics by the uncertainty they address

Affiliate and referral content can explain a product decision, but affiliation must be disclosed and eligibility verified. SEO can make a useful answer discoverable, but access and publication do not guarantee ranking. AI can help draft alternatives, but its confident suggestions are still hypotheses.

Community activity can reveal questions if there is moderation and a relevant audience. Pinterest, LinkedIn and TikTok are candidate distribution routes, not mandatory ingredients in every strategy.

Choose the route that tests the ledger's uncertainty. Do not implement five channels merely to appear comprehensive.

Make the first experiment small enough to interpret

Define one audience, offer, action, change and review period. Specify the cash limit and total effort. Write a reason to continue, stop or call the result inconclusive.

Hypothetical example: compare two explanations of the same checklist with comparable audiences and the same destination. Observe downloads and whether users can complete the intended task. A higher click count without useful completion may indicate a better teaser, not a better resource.

A simple experiment can still have confounders. Keep their likely effect visible.

Challenge the attractive metric

For every dashboard number, ask what event it counts, who qualifies, what denominator is used and whether the business can verify it.

Followers, impressions and views can be useful for specific questions. They become misleading when presented as customers or profit. Even a verified signup may not produce continuing use.

Our traffic experiment guide separates stages. The fintech attribution method shows why different validation rates can describe the same funnel.

Make economics part of the creative brief

Hypothetical campaign: €150 cash plus six hours valued at €25 means €300 economic cost. Five new customers at €45 contribution each generate €225 before campaign cost. The campaign has a €75 shortfall on that basis.

If the campaign also produced learning, describe the learning separately. Do not quietly convert an unsuccessful acquisition result into a profitable case study by valuing impressions.

Use AI as a challenger with a record

Ask the tool to identify assumptions, alternative explanations and missing sources. Verify those sources independently. Preserve your original wording before a rewrite and check for changes in stance, certainty and attribution.

Our AI bias audit provides a reproducible test. Our AI search writing guide explains how evidence and original analysis can make a page worth citing. Neither method promises algorithmic favour.

Keep a source of record

A useful article should make its question, sources, reasoning and limitations visible. Maintain it when terms change. Link to related pages where they resolve a genuine next question.

Google's people-first content guidance provides a relevant primary reference. It supports focusing on useful content; it is not a declaration that this blog has achieved topical authority.

Publish the decision, including the uncomfortable one

A report should say what was tested, what happened, what remains uncertain and what changes next. “Stop this campaign” may be the most valuable result. “We need more relevant observations” may be more honest than naming a winner.

The method works as a discipline only if the evidence can change the decision. Our Rebel marketing philosophy explains the wider position behind that discipline.

Next, apply the ledger to worked examples of testing marketing claims. That companion article turns traffic, AI-efficiency and referral slogans into testable questions.

— The Rebel Marketer