How to choose a marketing agency for your business.

Picking a marketing agency is a five or six figure annual decision that most business owners make after two phone calls and a nice looking proposal deck. That is backwards. The agencies that are wrong for you are often the best at the sales process, because winning new clients is the one skill every agency practices daily, whether or not they are any good at running your actual campaigns once the contract is signed. This is the vetting process we would want a business owner to run on us, and on anyone else pitching for the job: the questions that separate agencies fast, the reference checks that actually mean something, the red flags that should end a conversation immediately, and the contract terms that protect you if the fit turns out wrong.

Before You Sign AnythingThe Vetting Process

Why most agency hires go wrong before the contract is even signed.

Most business owners hire the agency with the best pitch, not the best fit. That is not a character flaw, it is just how the decision gets made under time pressure: a handful of calls, a proposal deck with impressive-looking logos, a salesperson who is genuinely likeable, and a gut feeling that this one seems different from the last one. None of that predicts whether the agency will still be answering your emails in month four, or whether the person who ran the pitch will ever touch your account again once the ink dries.

The uncomfortable truth is that pitching and delivering are different skills, and the agencies best at the first one are not reliably the ones best at the second. A strong proposal deck takes a few hours from a skilled writer. A strong campaign takes months of disciplined work from people who may never sit in on a sales call. Separating the two requires asking questions that a slide deck cannot answer, which is exactly what most vetting processes skip.

We say this having sat on both sides of the table. We run paid media, SEO, and web design for Network Automotive, a multi-location Arizona auto repair group, and Network Collision, a Gilbert, Arizona collision shop, and we have inherited more than one account from a previous agency whose reporting was polished and whose actual results had been flat for months before anyone caught it. The gap between the pitch and the work is exactly where this article lives.

Get this nailed down before you take a single call.

The vetting process works far better when you walk in already knowing three things about your own business, because an agency that has to guess at them cannot give you an honest answer to anything else.

• Your actual starting numbers. Current lead volume, rough cost per lead if you are running any ads, and close rate if you track it. An agency cannot tell you what is realistic without a baseline, and one that does not ask for it is guessing at your plan too.
• An honest timeline expectation, by channel. Paid direct response on Meta and Google can produce real leads within 24 to 48 hours of launch and a confident read inside two weeks. SEO and content take months to show meaningful movement, typically four to six at minimum. If you walk in expecting SEO to behave like paid ads, every agency will look slow by week three, including a genuinely good one.
• Who inside your business actually owns this decision. Agencies that spend weeks re-pitching the same material to three different decision makers waste everyone's time, and it is a fair question for them to ask you back.

You do not need a finished marketing plan to have this conversation. You need enough clarity that the agency is responding to your actual situation instead of a generic pitch deck that would work on anyone.

The vetting call: the questions that separate agencies fast.

A handful of direct questions, asked in the first call, will tell you more than an hour of case study slides. The agencies worth hiring answer these specifically and without flinching. The ones worth avoiding get vague, pivot to a different topic, or answer a question you did not ask.

• Who exactly will work on my account day to day? Not the person pitching you. Ask for the name and role of the person who will actually be inside your ad accounts and writing your content next month.
• Show me a client in my exact category, not something similar. "Similar" is doing a lot of work in most agency pitches. A home services client is not the same proof as a retail client, even if both are technically local businesses.
• What is the earliest realistic signal you would show me, and in which channel? A specific answer, tied to a specific channel and a specific week, is a sign they have run this exact playbook before. A vague "results vary" is a sign they have not run it enough times to know.
• How often do I get raw numbers, not a slide deck? Weekly or biweekly access to actual lead counts and cost per lead, not a monthly summary built to look good, is the standard worth holding out for.
• What happens in month one if the early numbers are bad? A good agency has a specific answer here: what they diagnose first, what they change, and when. A vague "we'll optimize" is not an answer.
• If the pitch centers on one channel, why that channel for my specific business? Whether the answer leans on SEO, Meta ads, or Google Ads, a good agency can explain why that channel fits your buying cycle and budget, and whether they run the others too or only sell what they happen to be good at.

Watch for one more thing that has nothing to do with the answers themselves: how the agency handles a question they cannot answer perfectly. "We haven't run that exact scenario, here's how we'd approach it" is a far better sign than a confident answer to everything, because nobody has actually run every scenario, and the ones who claim they have are performing certainty instead of offering it.

How to actually check case studies and references.

A logo on a slide proves an agency once had that client, nothing more. The proof that actually matters comes from a phone call, not a case study PDF built to flatter both parties.

Ask for two or three references yourself, and insist on picking at least one from their client list rather than accepting only the names they hand you. A reference the agency did not choose for you is worth more than two they picked specifically because that relationship went well. When you get someone on the phone, skip the generic "were you happy" question and ask specifics: how fast did they respond when something broke, what actually changed between month one and month six, did the reported numbers match what showed up in their own CRM, and would they sign with this agency again today, not two years ago when the relationship started.

Case studies with real numbers, not just percentages, are worth more than polished ones with no context. "Cost per lead dropped 40 percent" means nothing without knowing the starting number, the channel, and the timeframe. Our own breakdown of lead generation channels by real speed to first result covers what honest, channel-specific numbers actually look like when an agency is willing to show its work instead of a single flattering percentage.

One more test worth running: ask what went wrong on a real account, and how they handled it. Every agency running enough campaigns has had a bad month somewhere. The ones worth hiring can describe one specifically. The ones that insist nothing has ever gone wrong are either new enough to not have hit one yet, or not being straight with you.

Red flags and green flags: what to actually watch for.

Worth Ending The Call OverRed Flags

What a red-flag agency looks like.

• Vague answers dressed up in jargon. "Synergy" and "full-funnel omnichannel strategy" are not answers to "what will you actually do in week one."
• No direct access to your own ad accounts or analytics. This is a control tactic, not a technical limitation, and it should raise real concern on its own.
• Long lock-in contracts with no exit clause. A 12-month commitment with no way out signals the agency expects you to want to leave before month six.
• Resistance to raw reporting. If every answer to "can I see the actual numbers" gets redirected to a scheduled monthly call, that call is being built to manage you, not inform you.
• A proposal leaning on the old playbook. If direct mail, coupon mailers, radio spots, or billboards show up as a real part of the plan, that's the old playbook, untrackable, skip it, and it's fair to wonder what else in the proposal is similarly unmeasurable.
• One person handling sales, strategy, and delivery. Fine for a very small shop, but ask directly what happens to your account when that one person is out sick or leaves.

The agency that is honest about a slow month is more trustworthy than the one that has never had one.

What a green-flag agency looks like.

• Full access to your own accounts, from day one, with no argument about it.
• A named point of contact who is the same person you will actually talk to in month three.
• Specific, honest timelines by channel, including telling you plainly when a channel will be slow to show results.
• References you can call, plus a willingness to let you pick one yourself.
• A written exit clause that does not require a lawyer to interpret.
• A willingness to say "that's not something we're strong at" about at least one channel, instead of claiming expertise in everything.

Contract terms that protect you.

The proposal is where an agency sells you. The contract is where you find out what they actually committed to, and it is worth reading closely instead of skimming to the signature line.

Push for month to month, or a 90-day initial term at the very most. Anything longer than that should come attached to a specific reason, like a website build with a real production timeline, not a default every client gets talked into regardless of scope. A 30-day written notice to cancel after the initial term is the standard worth holding the line on, and an agency confident in its own results has no reason to resist it.

Ownership matters more than most business owners think to ask about upfront. You should own your website, your ad accounts, your domain, your Google Business Profile, and every piece of content produced for you, in your name, from day one, not held in the agency's accounts pending a data transfer if the relationship ends. Get this in writing, because "we'll hand everything over when you leave" is a promise, not a contract term, and promises are exactly what a contract exists to replace.

Reporting cadence belongs in the agreement too, specifically, not as a vague "regular updates" line. Weekly or biweekly numbers, in a format you can actually read without a scheduled call to interpret it, should be a stated deliverable. On pricing, most local business retainers land somewhere between roughly $1,500 and $6,000 a month depending on channel mix and market, though the number matters far less than what is actually included in it. A cheap retainer with no ownership rights and a 12-month lock-in is a worse deal than a pricier one with full access and a 30-day exit.

The first 90 days: how to judge the hire after signing.

Signing the contract is not the end of the vetting process, it is the start of the part where you actually find out if the answers on the call were true. The first 90 days is when the pattern becomes visible.

For paid channels, expect real setup diligence in week one: tracking verified, a clear offer defined, and a specific plan for the campaign before anything launches. Real lead activity should show up within 24 to 48 hours of a Meta or Google direct response campaign going live, with a confident read on cost per lead by day 14. If nothing has moved by week three and nobody flagged it or proposed a fix before you had to ask, that is a real problem worth escalating immediately.

For SEO and content, 90 days will not produce dramatic ranking movement, and any agency claiming otherwise in month one is either inexperienced or overselling. What you should see instead is a clear, published plan, consistent output against it, and honest monthly updates on what is actually happening in the meantime. Our breakdown of the four numbers that prove marketing is working covers exactly what to track once the agency is live, so the answer to "is this working" is a specific number instead of a shrug from either side of the relationship.

The single best predictor of whether an agency will still be a good fit in year two is not the pitch, and it is not even the first month's results. It is whether the numbers you asked for on the vetting call are the numbers you are actually getting once the contract is signed. If those two things match, you vetted well. If they do not, the vetting process was not thorough enough, and it is worth running it again with someone else.

Frequently asked
questions.

How long should I trial a new marketing agency before deciding to stay?

Give paid channels like Meta and Google direct response 60 to 90 days, since the first two to four weeks are a learning phase where costs run high and swing daily. Give SEO or content a minimum of four to six months before judging results, because meaningful movement in organic rankings simply does not happen faster than that. Judging either channel on a 30-day window will make a good agency look like a bad one.

What's a reasonable contract length for a marketing agency?

Month to month, or a 90-day initial term at most, with a 30-day written notice to cancel after that. Anything longer than a 90-day initial commitment should come with a specific reason attached to it, like a website build or a content backlog that genuinely needs runway, not just a default the agency defaults every client into.

Should I hire an agency or an in-house marketer first?

For most local and regional businesses under a few million in revenue, an agency delivers more channel expertise per dollar than a single in-house hire can, since one person rarely has deep, current skill across paid media, SEO, and web design at once. In-house makes more sense once marketing spend is large enough to justify a full team, or once you need someone physically present for daily operational coordination an outside agency cannot provide.

How many client references should I actually call?

Two to three, and pick at least one yourself from the agency's client list rather than accepting only the references they hand you. A reference the agency did not choose for you is worth more than two they picked specifically because the relationship went well.

What's the biggest red flag when choosing a marketing agency?

Refusing to give you direct access to your own ad accounts, analytics, or website. That is a control tactic, not a data limitation, and it usually means the agency is protecting numbers it does not want you to see rather than protecting anything technical. Everything else on this list is worth weighing. This one should end the conversation on its own.

Mitchell Wolfert Founder, Provision Digital. Running paid media, SEO, and funnels for local and national brands from Queen Creek, Arizona.