Is hiring a marketing agency worth it?

What are you comparing when you ask if an agency is worth it?

In my client work, owners all want the same thing in the end: results. Price is a hard place to start the comparison, because the three prices are set in different ways.

A freelancer sets their own. An employee's pay is often set by HR, from the going range and from where you hire. The same role costs more in one city than in another. An agency's fee covers a share of a team's time.

So the three numbers describe three different amounts of work. Before you compare them, write down the job. Which tasks need doing every month? How good does each need to be? Then ask what each route would cover, and what it would leave with you.

Why does an agency's fee look bigger than the alternatives?

Because it arrives as one invoice. The costs of the other routes arrive in pieces.

An employee costs a salary. Recruitment, tools, training, your time managing them and cover when they're away can follow. A freelancer costs their fee. The skills they don't have can cost extra. So can your time joining their work to everyone else's.

An agency's invoice isn't the whole cost either. Your time briefing and approving sits outside it, and so does the ad spend.

Leaving things as they are is free on paper. If the work would have brought customers in, that is its cost. Nobody sends an invoice for what a quiet day costs.

There's research on a related point. Christopher Hsee ran four experiments on judging options one at a time or side by side. In one, 112 students judged made-up job candidates. Judged alone, a quality that was hard to assess counted for little. Side by side, it counted for more. These were classroom tasks, so apply them to a real purchase with care.

The practical step is to put your options next to each other, on the same rows. The table near the end does that.

When does an agency beat a freelancer or an in-house hire?

The larger the scale, the more the choice tilts towards an agency. I see that tilt when marketing starts to touch several areas at once. Ads, design, video, copy and tracking are different skills. The tilt comes if you want expert level in each.

I see it again when the market is highly contested. If several businesses are after the same customer, winning that customer can't be left to chance.

The other routes have their own conditions. A single, well-defined task can go to a freelancer. Daily work that depends on knowing your business deeply can sit with an employee.

Research on large advertisers fits the employee case. Birger Wernerfelt, Alvin Silk and Shuyi Yu surveyed senior marketers at 79 large companies, each with more than 1,000 staff. Where managers said they could hire people as good as an agency's, more of the work stayed in-house. The authors read that as company knowledge mattering more than specialist skill. Every company kept some work inside, 58% on average. It was one survey of big firms, at one point in time. It can't tell a small business what to do. Among those firms, though, a mix was normal.

I buy in specialist work myself. I pay other people for graphic design, video and real photography, even with AI tools getting better.

Does paying for marketing help pay for itself?

An experiment published by the World Bank came close to testing it. Stephen Anderson and David McKenzie followed 753 small firms, each with 2 to 15 workers. About 600 were offered help, which the programme subsidised. The help was training, a consultant, a hired worker or an outside specialist. The rest were offered nothing, as the comparison.

By the second year, firms with an outside specialist had higher sales and profits than the comparison group. The result held on four of the five measures. The estimate for firms with a hired worker was also positive, but smaller and not statistically clear. The study couldn't rule out that the two were equal.

The researchers then did a rough sum. Their estimate of the extra monthly profit was imprecise, and on its own it wasn't statistically clear. Taken at face value, it would repay the subsidy in just over a year. It would repay the programme's full cost per firm in under two. They called it a return that could pass a cost-benefit test, "with a large degree of uncertainty still".

One more result matters here. More than a year after the subsidy ended, the researchers checked who was still buying. Take the firms given a specialist, a hired worker or a consultant. Against the comparison group, they were 14 to 19 percentage points more likely to be paying for outside professionals. They were spending their own money by then.

Keep the limits in view. The study ran in one country, with firms that had all applied to a business-support programme. The help was subsidised. The providers were screened, and each began with a diagnosis of the firm. The specialist was one professional, in marketing for about three-quarters of the firms and accounting for the rest. So the study can't forecast your return.

It does show something about owners who tried outside help. More than a year on, they were more likely to be paying for it than owners who hadn't. The study speaks to paying for specialist help at all. Which kind fits comes down to the conditions above.

Why do owners hold back from paying for help?

The same researchers asked owners why firms like theirs don't buy this help. For marketing, nearly half pointed to a lack of funds. Over a quarter pointed to an uncertain return. About a fifth said it was hard to judge a provider's skill. Only 12% pointed to a low return from marketing help, and only 11% to poor marketing providers.

So money and proof came up most. A belief that the help doesn't work came up least. The second answer fits the invoice problem. The cost is certain and visible. The return is neither, and trying is the main way to learn more about it.

Information alone didn't settle it. In a second study, Anderson and McKenzie ran an experiment with more than 1,000 small firms. About three-quarters were shown a directory of providers, and some of those also saw quality ratings. The ratings changed which provider the firms preferred. They didn't measurably change how many firms hired one over six months. Read that with care. The follow-up reached about two-thirds of the firms, as the pandemic began. The authors' view is that owners often learn what this help is worth only by trying it.

One way to try it is a small first job. A defined piece of work, with a clear end, tells you more than any brochure.

Why does paid help sometimes fail to pay?

Sometimes the work was poor, or it was the wrong work for the business. Still, when I'm told that help didn't pay, I ask two questions. What was the goal? Which measures were tracked? The answers decide what "paid" means.

Some marketing isn't aimed straight at a sale, and its reward is hard to count. Take a billboard on a busy street. Did it pay for itself? On its own, it may not have. Add the effect of several touch points over time, and the balance tends to tilt.

Owners who are burning cash want results faster. But marketing still comes down to human nature, influence and emotion, and those take time to work. Before you stop partway, read what happens when you stop advertising.

So agree the goal and the measure before the work starts. Agreeing them is part of being responsible for your marketing, whoever does the work. How to judge progress after that is a subject of its own.

When should you not hire an agency yet?

I've told owners to wait, or to buy something smaller. Two situations come up.

You can't serve the demand. There's no need to create demand you can't meet. A nationwide campaign makes no sense if you can't deliver on that scale.

The business isn't proven yet. If you're still finding out whether customers want what you sell, find that out first.

When to scale depends on a few things. Where is the business now? What have you tried? What worked, and what failed? How fast do you need to grow? How big is your war chest?

The war chest matters for a plain reason. A fee goes out every month, and the customers it brings pay later. A plan can look profitable and still leave you short in between.

If your current arrangement is working, keeping it is a fair answer.

How do you compare your options side by side?

Fill in one column for how things are now. Add a column for each real option. Use actual quotes and actual workloads, over the same number of months.

Count only what changes from today. Count profit after the cost of delivering the sale, because revenue overstates it. Keep ad spend apart from fees, so you compare the same plan each time. Where you don't know a figure, leave the gap. A gap tells you what to ask next.

There's no score. An agency may come out ahead. So may a freelancer, an employee, a smaller job or waiting. If an agency does come out ahead, choosing which one is the next decision.

Questions owners ask

Is it worth it to hire a marketing agency?

It is when the work spans several skills or your market is contested, and you can serve the extra customers. In one experiment with small firms, those given an outside specialist had higher sales and profits two years on. They were also more likely to be paying for outside help more than a year after the subsidy ended. The help was subsidised, so read it as evidence and nothing firmer.

Is an agency or in-house better?

It depends on the work. An employee suits daily work that needs deep knowledge of your business. An agency suits work that needs several specialist skills at once. In one survey of large companies, every one used both.

How much does it cost to hire a marketing agency?

Prices vary too much by place and scope for one figure to help. Compare the full cost of each route over the same months. Count fees or pay, setup, tools and your own time. Keep ad spend separate.

What is a disadvantage of an in-house marketing team?

One person rarely covers every skill, so gaps can appear in design, video or tracking. The cost also runs past the salary, into recruitment, tools, training and management time.

Can a freelancer replace a marketing agency?

For one defined task, often yes. When the work spans several skills, someone has to join the pieces together. With freelancers, that someone is often you.