A Gulf-based e-commerce marketing agency was selling to everyone, and struggling in two places. It attracted clients with structural business problems who expected ads, social campaigns or a new website to fix them, and who asked for results no campaign could deliver. And its proposals stalled: discovery calls and first meetings went well, then deals went quiet.
The founder came with a brief: reposition from a generalist agency to an e-commerce specialist. I joined on a four-month contract to build the strategy.
The brief carried two assumptions worth testing rather than executing. That e-commerce was the right ground to stand on — plausible, but asserted rather than evidenced. And that positioning was the problem to solve. If proposals were dying after good meetings, something in how clients made decisions was breaking, and positioning alone couldn't fix what nobody had understood yet.
The founder also assumed his clients were business owners. I held that loosely too.
The engagement as a double diamond. It started from the brief: reposition from generalist to e-commerce specialist. Discover widened to three perspectives. Define narrowed them to four patterns and one problem statement: help a decision-maker commit to work they can defend internally, before they fully trust us. Develop widened again into two decisions and six requirements. Deliver narrowed to one offer, now live: a fixed-scope, time-boxed diagnostic engagement.
I built the research from three perspectives.
The agency. I interviewed the founder and his senior team on how the agency won work, delivered it and lost it.
The clients. I designed the interview methodology; the founder conducted the interviews with his clients, and I synthesised them from the recordings.
The market. I reviewed competitors' strategies: what they offered, how they delivered it and what they competed on. They were full-service generalists offering the same broad range of services, with no stated methodology, competing on team size, client logos and awards. Their messaging followed from that: interchangeable, built on the same industry language. None of the competitors reviewed articulated an e-commerce specialisation or a point of view on how growth actually works.
The founder had been selling to business owners. The research showed the buyer was a system.
| Role | Position in the decision | What they needed |
|---|---|---|
| Business owner | Final authority, cost-sensitive | Control they could justify on cost |
| Marketing manager | Carries the risk | Protection from internal failure |
| E-commerce manager | Accountable for performance | Metrics that make sense |
| Operations manager | Influencer | Confidence on feasibility and resources |
The owner decides, but the manager carries the risk and makes the case internally. The manager is the relay; the owner is the target. Any positioning had to work on both at once.
I synthesised the three sources into recurring patterns. Four shaped everything that followed.
Demand was triggered by failure, not ambition. The clients interviewed didn't arrive looking for growth. They arrived when something had visibly broken, under pressure and afraid of internal judgement. So the agency's entry role had to be stabilisation first, growth later.
Strategy was needed, but it felt unsafe. Clients needed clarity, yet heard "strategy" as delay plus cost — which to them meant risk. So strategy had to arrive as the fastest route back to control, not as a separate phase to be paid for.
Trust collapsed at the point of commitment. Trust held during conversation and broke the moment a decision became explicit. This became the insight the rest of the work turned on — traced in full below.
The managers carried the risk. Even when the owner signed, the manager's credibility was on the line. So the agency wasn't really selling outcomes. It was selling decision safety.
Put together, the reframe was this: the agency wasn't selling marketing services. It was selling decision safety to a buying system at a moment of failure.
I mapped the buyer's decision rather than a sequence of touchpoints, because the decision was where deals were being lost.
Along that journey, trust was built by questions that showed real understanding, honest diagnosis instead of a pitch, clear structure and sequence, and numbers that made sense. It was destroyed by condescending expertise, overpromising, complexity without clarity, and attention that dropped after signature.
That gave the problem statement the strategy had to answer:
How might we help a decision-maker commit to structural work they can defend internally, before they fully trust us?
The engagement had to answer two separate questions. Where the agency should compete, and how it should compete once there. They rest on different evidence.
The founder arrived wanting e-commerce specialisation. My job was to test it rather than execute it, and the market review supported it: the competitors reviewed were full-service generalists competing on scale and credentials, with no stated methodology, leaving space for a clearly articulated e-commerce specialism. The capability was there too: the founder had started several e-commerce businesses of his own, and the leadership sessions confirmed the agency could deliver in that space. The client problems surfacing in the research were e-commerce problems. The direction held.
What the research did change was the basis for choosing clients. The founder had been selecting by industry, size and whoever was willing to pay. I recommended defining the agency by the moment it's needed — when performance has visibly stalled and the team is under internal pressure — because the research showed demand was triggered by failure, and different kinds of business shared that same moment.
E-commerce defined the capability boundary. The moment of failure defined when the agency was relevant.
That meant deciding who not to serve: business models that don't fit — marketplace-first, low-margin commodity, trend-dependent — and behaviours that don't, above all equating growth with ad spend and resisting structural fixes. A developmental audience, who believed ads were the answer, was to be educated rather than sold to.
Much of the market the agency was already talking to sat on the exclusion list, and the founder resisted turning that work away because he expected to lose revenue. I ran working sessions with him and the people responsible for marketing to work through it. I was explicit that strategy is a set of trade-offs, not a guarantee of success. The argument that carried it was that saying no protects the outcomes: a client who wants ads to fix a structural problem gets neither, and the agency is left owning a result it could never have delivered.
Where to compete came from the market. How to compete came from the buying behaviour.
From the insights, I set six requirements any engagement model had to meet, each traced to what went live below. Qualify before commitment is the hinge between the two decisions: where the choice of who to serve stops being a statement and starts being enforced.
After the strategy was complete, I ran a facilitation session with the senior team. I took them through how the strategy was constructed and how to apply it across presentations, proposals and client conversations, and the team used it, with my input, to design their engagement model against the requirements, then carried it into the site's structure, copy and social advertising.
For the requirements to hold up in how the team behaves, I also translated the agency's values into working behaviour: evidence over intuition, holding the line when pressure demands compromise, predictable delivery and communication, clarity that survives internal scrutiny, and strengthening the client's own authority.
The agency relaunched around a fixed-scope, time-boxed diagnostic engagement — the point where both decisions land in one offer.
| Requirement | What went live |
|---|---|
| Qualify before commitment | Clients apply rather than buy, against explicit fit and not-a-fit criteria — including clients who want ads without structural change; the agency also commits to telling clients when they don't need further work |
| Follow a fixed sequence | Diagnosis first, then implementation, then handover |
| Reduce the risk of strategy | Fixed scope and duration; price scoped after qualification rather than displayed |
| Protect the decision-maker | The site speaks directly to the problem of justifying spend internally, and to decision-makers and those who sit beside them |
| Transfer ownership | The client's team receives the change record and a roadmap it owns |
| Clarity before confidence | The site sets out the problem and what happens if it's left alone, before offering one clear route forward |
I reviewed the live site against the strategy. Two things drifted. The capabilities list reintroduces the generalist breadth the positioning set out to remove, including technology services the strategy explicitly ruled out. And the developmental audience is now excluded firmly but not visibly educated, which leaves out a group the strategy treated as central to shaping the category.
I owned the research design, the stakeholder interviews, the synthesis, the strategy, the engagement-model requirements and the facilitation. I didn't design the final packaging of the engagement, build the site, or produce any client results that appear on it.
I'd agree success measures before handover — proposal-to-close rate, where in the client's approval chain deals stall, the share of enquiries that fit — so adoption could be judged by results, not only by what went live.