How to Choose a Digital Marketing Agency: 15-Point Framework

Choosing the wrong digital marketing agency costs more than money. It costs a year — sometimes two — of momentum, plus the indirect damage of a brand that’s been said the wrong way in too many places. The downstream cost of a bad agency relationship usually exceeds the actual retainer by 3-5×.

The flip side: the right agency, hired for the right scope, compounds value year over year. Many of the highest-growth companies we know have had the same agency partner for 5+ years.

This is the framework we share with prospects evaluating us — or evaluating other agencies — even when they don’t ultimately choose us. A clean evaluation process protects you against bad choices and helps you make the right one.

Business decision meeting

Before you start: define what you actually need

The most common mistake in agency selection isn’t picking the wrong agency — it’s not knowing what you need before you start the conversation. You can’t evaluate fit if you can’t articulate the problem.

Write down, in one document:

Business outcome you’re after. Not “more leads.” “Generate $400K in new pipeline by Q3, of which 40%+ is from inbound.” Specific numbers, time-bound.
Disciplines required. Paid search? SEO? Content? Brand design? Video? Tag every discipline that’s in scope.
Internal capabilities you keep. If you have a content writer in-house, the agency shouldn’t be writing content. Be clear about what they don’t do.
Budget range. Not exact — range. “We can spend $7K-$12K/month all-in including media.”
Timeline. Are you in launch mode (need work in 30 days) or building durable systems (12-month strategic relationship)?
Decision criteria. Will you choose primarily on price, expertise, team chemistry, or proven results? Be honest with yourself.

Most agencies you talk to will not push back on poorly-scoped briefs. They’ll quote against whatever you ask for. A well-scoped brief filters out 70% of bad fits before you even have a kickoff call.

The 15-point evaluation framework

For each agency you’re considering, score these 15 dimensions. Anything below “acceptable” on a critical point is a deal-breaker.

1. Specific case studies in your vertical

You need at least 2-3 case studies of work done for companies similar to yours — same vertical, similar size, similar challenge. Generalist agencies usually pivot to “we work across many verticals” — that’s often code for “we don’t have deep expertise in yours.”

Ask: “Can you walk me through 2 case studies where you worked on a [B2B SaaS / DTC ecommerce / law firm / specific vertical]? What were the starting metrics, what you did, and the outcome 12 months later?”

A good agency answer references specifics: starting MRR, ending MRR, what shifted in their funnel, what worked, what didn’t.

2. Reference calls with current clients

Ask for three current clients you can call. Note: current, not testimonials on the website. Talk to them on a 20-minute call without the agency present.

Questions to ask references:

How long have you worked with them?
What’s the one thing they’ve done exceptionally?
What’s the one thing that frustrates you?
Would you hire them again? Why or why not?

Two reference calls is fine; three is better. If the agency can’t produce three current clients willing to talk, that’s a yellow flag.

3. Named team members on your account

Who specifically will be doing the work? You should know the strategist’s name, the paid media specialist’s name, the content lead’s name, etc. — and ideally have met them.

Common bait-and-switch: a senior partner sells the deal, then hands you off to a junior account manager you’ve never met. Pin this down in writing before signing.

4. Defined deliverables and cadence

The contract should specify, at minimum:

What’s delivered each month
What metrics will be reported
Cadence of meetings (weekly, bi-weekly, monthly)
Who you talk to for what (strategy vs. tactical questions)
Response-time SLAs for ad-hoc requests

Vague deliverables (“we’ll grow your traffic”) are the contractual equivalent of a blank check.

5. Transparent reporting

How will they report? Ask to see a sample report from a current client (anonymized).

What to look for in a good report:

Business outcomes (revenue, pipeline, leads), not just activity (impressions, posts)
Comparison to previous period and goals
Narrative explaining what happened and why
Next-period plan with specific actions

What to be wary of:

40-page PDFs of every metric in the platform with no synthesis
“Look at all this activity” reports without business outcome columns
Only reporting at quarter-end (you’ll find out about problems too late)

6. Pricing model and scope clarity

There are three common pricing models. Each is fine if structured right:

Retainer: monthly fee for defined scope. Best for ongoing work.
Project: fixed fee for fixed deliverable. Best for one-time work (rebrand, audit, site build).
Performance: percentage of revenue or per-conversion. Best for high-volume, easy-to-attribute channels (paid media at scale).

Hybrid is common: retainer + media spend pass-through + small performance bonus. That’s fine.

Watch out for:

Open-ended hourly billing for “strategic work”
“Spend management fee” exceeding 15% of media spend
Project fees that balloon with change orders

7. Contract length and termination flexibility

12-month lock-ins are agency-favorable, not client-favorable. The market has moved toward 60-90 day notice periods on otherwise indefinite contracts.

If the agency insists on 12 months minimum, ask: “What confidence do I have that you’ll perform, if you’re not willing to let me leave?” Their answer reveals a lot.

Reasonable contract terms:

30-90 day termination notice
First 90 days as a “trial” with reduced commitment
IP ownership clear (everything they produce is yours, including raw files and accounts)
Access transfer obligations on exit (they hand back your Google Ads, Meta, analytics access cleanly)

8. Tool stack and infrastructure

What tools will they use? Some questions:

Do they have professional licenses to Ahrefs/Semrush/SimilarWeb? (Not free tiers.)
Do they have a reporting infrastructure (Looker Studio, Klipfolio, custom dashboards)?
Do they have process tools (Notion, Asana, ClickUp) for managing your account?
Will they use their tools or yours?

A boutique with 4 clients using free Trello is one thing. A growth agency with 50 clients on the same setup is a process risk.

9. Specialist vs. generalist team composition

For most engagements, you want specialists working on each channel: SEO specialist on SEO, paid media specialist on paid media. Generalists who “do everything” are usually mediocre at everything.

Ask the agency: “Who specifically writes the ad copy? Who specifically optimizes the SEO? Who specifically does the design?” Different people should be the answer.

10. Track record of specific results, with proof

“We grew client X by 300%” is meaningless without context. From what to what? Over what time? With what spend?

Ask for one case study with:

Starting metric (e.g., $50K MRR)
Time period (e.g., 12 months)
Spend during period (e.g., $120K total media + $90K agency fees)
Ending metric (e.g., $180K MRR)
Contribution attribution (how much of the lift came from their work vs. organic growth or other factors)

Good agencies have this data. Bad agencies hand-wave it.

11. Strategic capability beyond execution

Anyone can execute campaigns. The differentiator is strategic thinking: deciding what not to do, identifying which channel matters now vs. later, anticipating where your funnel will break.

In the proposal phase, ask: “What would you NOT do for our company, that some other agency might suggest?”

The honest answer reveals strategic thinking. The agency that proposes 6 channels for a $5K/month budget doesn’t have it.

12. Communication style and responsiveness

You’re committing to spending hours per week talking to these people. Do you actually like the conversation? Does it feel productive?

Specifically test:

How quickly do they respond to a non-urgent email during the proposal phase? (If they take 4 days now, they’ll take 4 days during the engagement.)
Do they listen, or do they pitch over your questions?
Do they push back on bad ideas, including yours? (You want this. Yes-people don’t help you.)
Do you understand what they say, or do they hide behind jargon?

13. Cultural and ethical alignment

This sounds soft until it bites. Specific things to check:

Are they candid about what won’t work, or do they say yes to everything?
Will they do things you find ethically questionable to hit a metric (fake reviews, dark patterns, misleading copy)?
Do they treat your competitors with respect or trash them in the pitch?

A pitch that includes “we just stopped working with [competitor brand] because they wouldn’t follow our advice” is a yellow flag. Talk about their losing accounts that way and they’ll talk about you that way too.

14. Industry expertise and trends literacy

The 2026 marketing landscape moves fast: AI search, third-party cookie deprecation, iOS attribution, Performance Max, server-side tracking. An agency that doesn’t know what’s changed in the last 6 months is a year behind.

Quick test: “What’s changed about [SEO / paid media / paid social] in the last 6 months that affects our strategy?” A current agency has a thoughtful answer.

15. The “talk to ops” test

The CEO or partner sold you. Now ask to spend 30 minutes with the actual person who will do the work day-to-day. If you can’t, that’s a yellow flag. If you can, do they impress you?

The pitch deck reflects the agency’s marketing. The operator reflects the agency’s actual delivery.

Team strategy whiteboarding

Red flags to walk away from

If any of these come up, end the conversation:

Guaranteed results. “We guarantee #1 rankings” or “we guarantee 50% more leads in 90 days” — anyone promising specific results without seeing your funnel is selling vapor.
Pressure to sign before doing diligence. “This pricing expires Friday.” Legitimate agencies aren’t running this play.
Mandatory long lock-ins. 12+ month minimums with no performance escape hatch.
No questions from them. Good agencies should be asking you 30+ questions about your business before they propose. If they pitched without asking, they’re selling a generic service.
Owns your accounts. Some agencies set up Google Ads / Meta accounts in their name, not yours. You walk away and you lose all your data and ad history. Unacceptable.
Junior-only team. No senior strategist on the engagement. You’ll be paying for OJT (on-the-job training) without knowing it.
Refuses to share references. Or only shares 1-year-old “former client” references.

How to run the evaluation process

A clean process protects you from sales pressure. Run this:

Week 1: Define your brief (see top of article). Email 5-7 agencies. Don’t shortlist before they respond — see who answers thoughtfully vs. who blasts a generic deck.

Week 2-3: Initial discovery calls with the 4-5 that responded well. Each call is ~45 minutes. Goal: understand their capabilities and approach. Take notes.

Week 4: Request proposals from your top 3. Set a clear proposal due date. Be explicit about what you want answered.

Week 5: Compare proposals. Schedule references from each finalist. Run the 15-point checklist for each.

Week 6: Final meetings with the top 2. Meet the actual operators, not just sales. Negotiate contract terms.

Week 7-8: Decide, sign, kickoff.

That’s an 8-week process. Faster is possible but risky. Slower is fine if you’re paying due diligence.

After signing: the first 90 days

The first 90 days set the tone for the entire engagement. Bake these into your contract:

Days 1-30: Onboarding and audit.

Agency does deep audit of current state (account, funnel, competition, content).
You provide access to everything (analytics, ad accounts, CRM read access).
Define joint KPIs with clear targets.

Days 31-60: Strategy and quick wins.

Agency presents 90-day strategic plan with prioritized initiatives.
They identify and execute 2-3 quick wins (typically fixes to existing accounts that yield immediate ROI).
Weekly reporting starts.

Days 61-90: Execution and first review.

Full execution of agreed strategy in motion.
90-day review meeting: did they hit milestones? Are KPIs trending right?
If yes: continue. If meaningfully off-track: hard conversation, not termination yet.

If by day 90 they haven’t delivered what they promised in the contract or in early kickoff, that’s the moment to renegotiate or exit — not month 6.

The questions to ask before signing

Print these and bring them to your final meeting:

Who specifically will work on my account, in what roles?
What is the cadence of meetings, reports, and strategy reviews?
What happens if the lead strategist leaves your agency in month 4?
What’s your process for monthly reporting? Can you show me a sample?
What does success look like at 90 days? At 1 year? At 2 years?
Who owns the work product, accounts, and data?
What happens to my accounts if we part ways?
How do you handle scope changes?
What’s the termination clause?
How do you measure your own performance against client objectives?

A confident, well-run agency answers all of these in writing without hedging.

Frequently asked questions

How much should a digital marketing agency cost? For SMBs (under $5M revenue): $3K-$10K/month for serious work. For mid-market ($5M-$50M): $10K-$30K/month. For larger enterprises: $30K-$100K+ depending on scope. Below $3K/month, you’re likely paying for a freelancer wearing an agency hat.

Should I hire a local agency or a remote one? Mostly irrelevant in 2026. Quality of work and fit matter more than geography. Time zone overlap with your team matters more than physical proximity.

Can I terminate an agency early if they’re not performing? Depends on your contract. If you have a 30-90 day notice clause, yes. If you signed 12-month lock-ins, you have to fight harder — which is why you don’t sign 12-month lock-ins without performance clauses.

How often should I re-evaluate my agency? Quarterly reviews against KPIs. Annual full review where you assess whether to continue, renegotiate scope, or look for alternatives. Don’t change agencies just because you’re bored; do change them if performance trends are clearly downward.

What if my agency is doing fine but not great? Have a frank conversation about what “great” would look like and whether they’re willing/able to deliver it. Most agencies prefer to retain a client by leveling up rather than lose them. Give them one quarter to improve, then decide.

Choosing the right agency is more like hiring a co-founder than buying a service. The evaluation process should feel like dating — multiple conversations, real questions, references, and a slow build of trust. Companies that rush this end up paying for the rushing for two years. Companies that take it seriously end up with a partner that compounds value for half a decade.