Define the constraint before the capability list
Write the business outcome, current evidence, affected customer journey, internal owner, timeline, and known constraints. A request for SEO may actually be a positioning problem. A paid-media request may be a weak landing page or tracking problem. An AI request may be an undefined workflow.
A good partner should help refine the diagnosis, but the selection becomes clearer when every candidate is responding to the same business problem rather than inventing a different project around their preferred service.
- The desired outcome is observable
- Current performance and uncertainty are stated
- Internal ownership and decision rights are clear
- The project boundary and dependencies are visible
Match the partner model to the shape of the work
Use a specialist when the problem is narrow, the surrounding strategy is stable, and the internal team can coordinate the dependencies. Use a freelancer for focused execution with a clear brief and close internal ownership. Build in-house when the capability is continuous, strategic, and central enough to justify the hiring and management cost.
An integrated growth team is useful when the constraint crosses message, design, development, channels, data, and operations. Integration should reduce handoffs and conflicting incentives, not simply bundle more billable services.
Ask questions that reveal how the work will actually run
Ask who diagnoses, who delivers, how senior people stay involved, what the first two weeks look like, how decisions are documented, which assumptions are unproven, and what the team needs from you. Review relevant evidence, but do not accept unrelated logo walls as proof of fit.
A strong proposal connects the problem to a sequence of decisions, deliverables, responsibilities, and measures. It should distinguish what is known, what will be tested, and what can change after discovery.
- Who will do the work after the sale?
- What evidence supports the diagnosis?
- Which dependencies could delay or weaken the outcome?
- How will quality and performance be measured?
- What happens if early evidence changes the plan?
- What access and time are required from the internal team?
Compare risk, ownership, and learning, not only price
A lower fee can become expensive when the internal team must rewrite strategy, coordinate several vendors, repair tracking, or rebuild poor work. A larger engagement can also waste money when the problem is not yet clear. Compare the complete cost of coordination, delay, rework, tools, media, and internal time.
For uncertain problems, start with a diagnostic or focused sprint that creates a shared evidence base. Expand after the team has seen the quality of thinking, communication, delivery, and measurement.
Notice promises that remove useful uncertainty
Be cautious with guaranteed rankings, guaranteed revenue, instant AI transformation, proprietary frameworks that cannot be explained, generic channel prescriptions, and measurement plans built after launch. Serious partners can be confident without pretending the market is controllable.
The best early signal is often the quality of the questions. A partner who understands the customer, economics, system dependencies, and decision process is more useful than one who starts with a long list of tactics.