Most founders don't realize they've hired the wrong marketing agency until three months and $30,000 in. The agency promised growth, delivered brand decks, and couldn't explain what CAC meant for your business model. That pattern is predictable - and avoidable.
This guide covers what separates a startup marketing agency that actually moves metrics from one that burns your runway on deliverables that don't convert.
If you are specifically weighing a partner who is paid on pipeline rather than hours, our performance marketing agency for startups guide explains how to choose and vet one.
Related: For the hiring and evaluation side of the same decision, see our growth marketing agency for startups guide.
Separate Strategy from Execution in the Scope
The agency that writes the plan and runs it should be judged on the plan's result, not its activity. Insist the scope names the metric - pipeline, CAC, activated users - so the work points at growth.
Ask for the Rep'S Actual Track Record
Case studies on the site are curated. Ask for two references in your stage and category, then call them. The real signal is what a peer in your shoes experienced, not the deck.
Keep a 30-Day Off-Ramp
Good agencies accept a short prove-it window; bad ones lock you in. A 30-day out protects the runway while you learn whether the engine actually turns.
What a Growth Agency Is Actually Hired to Do
A growth marketing agency is hired to build and run the system that turns attention into pipeline. That means channel strategy, campaign execution, and measurement - not a deck about strategy you then have to staff yourself. The good ones bring both the plan and the operators. Before you sign, name the single number they own: qualified demos, activated trials, or pipeline. If the contract lists activities instead of an outcome, keep looking. The outcome is what protects your runway.
How to Scope the Engagement
Scope should separate strategy from execution and name the metric. A clean scope reads: the agency proposes the plan, runs the channels, and reports against pipeline weekly. It should also state what you provide - product access, sales follow-up, an internal owner. Engagements that leave the client's responsibilities vague stall when the agency waits on inputs you did not know you owed. Write the handoff both ways before the first invoice.
Red Flags in the Pitch
Watch for a deck with no references in your stage, a plan that is really a generic template, or a lock-in contract with no exit. Also beware the shop that wants to own strategy without letting you see it - you should receive the plan, not just the results. The pitch is a sample of the working relationship; if it dodges specifics, the engagement will too.
Questions to Ask Before You Sign
Ask for two references in your category and call them. Ask which channels they would avoid for you and why. Ask how they report and how often. Ask what happens in month two if the first month misses. The answers tell you more than the case studies, which are curated. A confident agency answers plainly; a vague one is a risk to the budget.
Setting the Success Metric Up Front
Agree on the number before the work starts: cost per qualified demo, activated trials, or pipeline contribution. Review it monthly against the plan. If it moves and the playbook is documented, the engagement worked. If it is flat after two months, change the plan or the partner. Do not wait a quarter to notice the number is not moving.
Keeping an Internal Owner
Outsource the doing, keep the ownership. A founder or hire should own the strategy so the agency executes against your goal, not theirs. Without an internal owner, the work drifts toward whatever is easy to ship. The owner is the tie between the agency's output and your business number. That role is non-negotiable for a good result.
The 30-Day Prove-It Window
Insist on a short initial window where the agency shows it can move the metric before you commit long. Good agencies accept this; bad ones hide behind a year lock. Use the window to watch not just results but how they work - do they report clearly, do they act on data, do they communicate. The window protects the runway while you learn if the engine turns.
When an Agency Is the Wrong Call
If you have no product-market signal yet, an agency spends your money finding it for you at a markup. Fix the message and the channel fit first, then scale with help. Also, if you cannot staff the sales follow-up, more leads just rot. An agency multiplies a working motion; it does not invent one from nothing. Know which stage you are in.
Pricing Models and What They Signal
Retainer, performance, and project are the three shapes. Retainer buys availability; performance aligns incentives but is rare for top-of-funnel; project buys a defined build. The model signals the risk split. Pick the one that ties payment to the outcome you care about, and read the fine print on what 'performance' actually means before you trust it.
Handoff and Continuity
A good engagement ends with you more capable, not dependent. Require the plan, the metric definitions, and the channel notes in a shared doc. When the contract ends, you keep the machine. If the departure leaves a hole, the engagement under-delivered regardless of the pipeline it produced. Continuity is the real deliverable.
Measuring ROI Beyond the Lead Count
Leads are a means, not the result. Track the leads that became meetings, the meetings that became pipeline, and the pipeline that closed. An agency that floods top-funnel but never connects to revenue is a cost, not a growth partner. Tie the report to the money, and the relationship stays honest to what the business needs.
Category Specialists vs Generalists
A shop that lives in your category knows the buyer and the channels cold. A generalist brings fresh angles but learns on your budget. For a first engagement, the specialist usually pays back faster. As you scale, a generalist can widen the mix. Match the firm to the gap, not to the logo on the homepage.
Running the Agency Like a Teammate
Treat the agency as a senior function, not a vendor. Share the roadmap, the sales feedback, and the churn reasons. The more context they have, the better the targeting. A vendor gets a brief; a teammate gets the business. The second relationship is what produces growth instead of deliverables.
Exit with the Asset, Not Just the Report
When you part ways, you should leave with a working system and a hired or promoted owner. The engagement succeeded only if the company can run without the agency. Measure the exit by capability transferred, not by the last campaign's numbers. That is the bar that makes the spend worth it.
Common Myths About Agency Growth
The myth is that an agency alone drives growth. In reality it scales a motion you must own. Another is that more channels equal more growth; focus beats spread early. Drop both myths and the engagement stays pointed at the number. Clarity about the agency's real role is what keeps the budget from leaking into noise.
A Simple Vendor Scorecard
Score the agency monthly on three lines: does the metric move, is the report clear, and do we learn something. A shop that fails two of three for two months is a change candidate. The scorecard turns a vague relationship into a managed one. Use it from month one so the decision to stay or go is data, not discomfort.
Transitioning to an in-House Team
When the channel is proven, move the doing in-house and keep the agency for the hard new bets. That sequence scales cheaply: the agency blazes, your team harvests. Hiring too early rebuilds what you paid to learn; hiring too late caps the engine. The transition is a planned handoff, not a sudden cut.
Final Checklist Before You Sign
One metric owned, an internal owner named, a 30-day window, references called, the plan in writing, and an exit that leaves you capable. If all six are true, sign. If any is missing, treat it as a gap to close before the money moves. That checklist is the difference between a growth partner and an expensive lesson.