How do I choose a social media marketing agency?

Define the outcome you are buying before you look at agencies, then evaluate each one on five things: who will actually do the work day to day, how they measure success and what they will report, whether they own distribution or resell it, how their pricing is structured and what falls outside the fee, and what they do when a campaign is not working. Ask each shortlisted agency for a plan for your specific situation rather than a credentials deck, and ask to speak to a client whose engagement ended.

What questions should I ask a social media agency before hiring?

Who is on my account day to day, and what else are they staffed on? What will you report, how often, and what does the raw data look like? Which parts of this do you do in-house and which do you subcontract? What is not included in the fee? What would make you tell me this is not working? Can I speak to a client who left? The answers to the last two are the most revealing, because they are the ones agencies rehearse least.

How long should I sign an agency contract for?

Long enough for the work to produce a readable result and short enough that you are not trapped if it does not. For most social and content programmes early performance signal appears within weeks while compounding outcomes take months, so a three-month initial term with a defined success measure, followed by a rolling arrangement, protects both sides better than an annual lock-in. Be wary of any agency that will only sell twelve months up front, and equally wary of judging a programme on three weeks of data.

What are the warning signs of a bad marketing agency?

Guaranteed results, follower or view counts presented without any business outcome attached, an inability to explain how a number was measured, the senior team appearing only at the pitch, reporting that arrives as a slide deck you cannot interrogate, vagueness about what is subcontracted, and pricing that hides ad spend or creator fees inside a single headline number. None of these is proof of bad work on its own, but each one is a question you should get a straight answer to.

Should I hire an agency or build an in-house social team?

In-house tends to win when your output is steady, deeply product-specific, and needs to be always-on — the knowledge compounds internally and the cost per unit of content falls over time. An agency tends to win when you need capability you do not have, need to move faster than hiring allows, or have peaks such as launches and releases that would leave a permanent team underused between them. Many companies end up with both: a small in-house team owning brand voice and community, and an agency supplying volume, distribution reach and media buying.

Guide

How to choose a social media marketing agency

Most agency selection processes compare the wrong things. Three agencies present three decks, all of which contain impressive logos and large numbers, and the decision comes down to which room felt better. That is a reasonable way to pick a supplier you will speak to twice a year, and a poor way to pick one that will spend your money every week.

This guide is the set of questions worth asking instead. WVE is an agency and has an obvious interest here, so it is written to be useful even if you conclude you should hire someone else, or nobody.

Start by working out what you are actually buying

The most expensive mistake in agency selection happens before any agency is contacted: briefing a capability instead of an outcome. “We need someone to run our social” describes an activity. It gives every agency licence to sell you what they happen to be good at, and it gives you no basis on which to compare their answers.

Before you shortlist anyone, get specific about which of these you are solving for, because they lead to genuinely different agencies:

Awareness in a category that does not know you exist: You need reach and distribution — the ability to put content in front of large numbers of relevant people who are not already following you. Prioritise agencies whose strength is distribution and paid media.
Demand you can attribute to revenue: You need measurement discipline as much as creative. Prioritise agencies that can explain their attribution approach without reaching for a platform dashboard screenshot.
Volume you cannot produce internally: You need a production and content operation. Prioritise agencies that can show throughput — how many assets, at what cadence, at what unit cost.
A specific moment to land: A launch, a release, a fixture, a funding announcement. You need campaign planning and the ability to move quickly. Prioritise agencies with relevant category experience and a track record of hitting dates.
A positioning problem wearing a marketing costume: If your content underperforms everywhere and always has, more content will not fix it. You need strategy and positioning work first. An honest agency will tell you this; it is also the least profitable thing for them to say.

Write the outcome down as a sentence with a number and a date in it. You will use it to judge every proposal you receive, and you will notice quickly which agencies engage with it and which restate their own capabilities back at you.

The five questions that actually separate agencies

Credentials decks converge — everyone has logos, everyone has a case study with a big number. These five questions do not converge, because the answers are structural rather than presentational.

1. Who does the work, and what else are they doing?

The most common disappointment in agency relationships is not incompetence, it is substitution: the senior people who won the business are not the people who run the account. This is not always wrong — good juniors supervised well produce good work — but you should know which model you are buying.

Ask for the names and seniority of the people who will be on your account day to day, how many other accounts they carry, and who covers when they are away. Ask how much of the work is done in-house versus subcontracted. Subcontracting is normal and often sensible; being evasive about it is the problem.

2. How will they measure it, and what will you actually see?

Ask what they will report, how often, and — the important part — whether you get access to the underlying data or only a summary of it. An agency that reports into a dashboard you can interrogate is making a different commitment from one that sends a monthly slide.

Then ask them to define their terms. Impressions, views, reach and engagements are routinely used interchangeably in pitches and mean different things: an impression is content served, a view is content watched, reach is unique people. An agency that cannot cleanly explain the difference between the numbers in its own case study is either careless with data or relying on you not asking.

3. Do they own distribution, or are they buying it?

There is a real difference between an agency with genuine distribution — established creator relationships, community and meme page access, seeding infrastructure — and one that will source it on your budget after you sign. Both can work. The first is usually faster and cheaper per unit of reach; the second adds a markup and a coordination risk.

Ask directly how content will reach people beyond your existing followers, and how those relationships were built. Vague answers here usually mean the answer is “we will find out once we have your money”.

4. How is the pricing built, and what falls outside it?

The headline fee is rarely the whole cost. Ad spend, creator and talent fees, music licensing, production, and third-party tooling may all sit outside it. A quote that appears cheaper than another is often simply drawing the boundary in a different place.

Ask for the total cost of the first three months including everything you will be invoiced for or asked to fund directly. This is covered in more detail in the guide to what a social media marketing agency costs.

5. What happens when it is not working?

This is the single most useful question in agency selection, and the one least likely to have a rehearsed answer. Every programme has a period where results are below expectation. What you want to know is what the agency does then.

Good answers are specific and uncomfortable: here is the point at which we would tell you to stop, here is what we would change first, here is a time we told a client their product was the problem. Bad answers reframe the question as being about optimisation and reassurance.

What to ask for before you decide

Two requests will tell you more than any amount of additional presentation.

Ask for a plan for your situation, not a deck about theirs. Give each shortlisted agency the same brief and ask for how they would approach the first ninety days, what they would need from you, and what they would measure. You are not evaluating whether the plan is right — you cannot know that yet. You are evaluating whether they engaged with your problem or reached for a template.

Ask to speak to a client whose engagement has ended. Every agency will supply a happy current client. A former client will tell you how the relationship worked when it was under strain, whether reporting stayed honest, and how the offboarding went. An agency that cannot produce one is telling you something.

Where possible, structure the start as a paid pilot with a defined scope and a defined measure. It converts an argument about credentials into evidence, and both sides learn quickly whether the working relationship functions.

Red flags

Guaranteed results. Platform distribution is not deterministic. A guarantee is either meaningless or it is being met with low-quality traffic that satisfies the letter of the promise.
Vanity metrics with no business outcome attached. Views and followers are inputs. If nobody can draw a line from them to something the business cares about, they are being used as decoration.
Numbers with no defined method. If an agency cannot tell you what a figure counts, over what period, and from which source, treat it as unverified.
Bundled pricing that hides pass-through costs. You need to know what is fee and what is spend, or you cannot judge efficiency later.
No point of view on your business. An agency that agrees with everything in the brief has not read it critically, and will not push back later when it matters.
Reluctance to define an exit. Clear notice periods and data-handover terms are a sign of confidence, not of expected failure.

When you should not hire an agency at all

There are situations where the honest recommendation is to spend the money elsewhere.

When the product or offer is the constraint. Distribution reveals the strength of what you are distributing. If the offer does not convert the traffic it already gets, more attention mostly produces a faster, more expensive version of the same result.

When nobody internally owns the relationship. Agencies need decisions, assets, approvals and access. Without an internal owner with authority, you will pay for capacity you cannot use.

When the work is genuinely always-on and product-specific. Steady, deeply technical, community-heavy work often belongs in-house, where the context compounds instead of being re-briefed.

When the budget only covers the fee. If the entire budget is consumed by the retainer with nothing left for media, creators or production, the agency has no fuel and the programme will underperform through no fault of anyone.

A short checklist

Write the outcome as one sentence with a number and a date.
Decide which of the five problem types you are solving.
Shortlist three agencies whose strength matches that type.
Send all three the same brief and ask for a ninety-day plan.
Ask the five questions: who works on it, how it is measured, whether distribution is owned, how pricing is built, and what happens when it is not working.
Get the all-in cost of the first three months in writing.
Speak to a former client, not only a current one.
Start with a scoped pilot and a defined success measure.

Frequently asked questions

How do I choose a social media marketing agency?: Define the outcome you are buying before you look at agencies, then evaluate each one on five things: who will actually do the work day to day, how they measure success and what they will report, whether they own distribution or resell it, how their pricing is structured and what falls outside the fee, and what they do when a campaign is not working. Ask each shortlisted agency for a plan for your specific situation rather than a credentials deck, and ask to speak to a client whose engagement ended.
What questions should I ask a social media agency before hiring?: Who is on my account day to day, and what else are they staffed on? What will you report, how often, and what does the raw data look like? Which parts of this do you do in-house and which do you subcontract? What is not included in the fee? What would make you tell me this is not working? Can I speak to a client who left? The answers to the last two are the most revealing, because they are the ones agencies rehearse least.
How long should I sign an agency contract for?: Long enough for the work to produce a readable result and short enough that you are not trapped if it does not. For most social and content programmes early performance signal appears within weeks while compounding outcomes take months, so a three-month initial term with a defined success measure, followed by a rolling arrangement, protects both sides better than an annual lock-in. Be wary of any agency that will only sell twelve months up front, and equally wary of judging a programme on three weeks of data.
What are the warning signs of a bad marketing agency?: Guaranteed results, follower or view counts presented without any business outcome attached, an inability to explain how a number was measured, the senior team appearing only at the pitch, reporting that arrives as a slide deck you cannot interrogate, vagueness about what is subcontracted, and pricing that hides ad spend or creator fees inside a single headline number. None of these is proof of bad work on its own, but each one is a question you should get a straight answer to.
Should I hire an agency or build an in-house social team?: In-house tends to win when your output is steady, deeply product-specific, and needs to be always-on — the knowledge compounds internally and the cost per unit of content falls over time. An agency tends to win when you need capability you do not have, need to move faster than hiring allows, or have peaks such as launches and releases that would leave a permanent team underused between them. Many companies end up with both: a small in-house team owning brand voice and community, and an agency supplying volume, distribution reach and media buying.

Ready to talk about your campaign?

Tell us the launch, the release or the moment you need to land, and we'll come back with how we'd build it.

Start a conversation