Agency Economics

November 18, 2021
Michael Taylor

Where does your money go when you hire a marketing agency? What can the economics of an agency tell you about when to hire one? How can you avoid being ripped off when working with an agency?

If you work in an agency, you might similarly be wondering why so little of the revenue you manage ends up in your paycheck. If you run an agency, you’re probably frustrated at how little you make in profit, and every month you struggle to make payroll.

As the agency I founded grew from 2, to 20, and then 50 people, I spent a lot of time with difficult clients, talking to similarly exasperated agency owners and figuring out how agencies make money. So now that I’ve left the agency world, I’m putting a model out there to show how agencies work, what incentives they have and how to use that information.

Feel free to jump straight into the model and play around.

UPDATE: I built a better model in a tool called Causal, which lets you easily forecast a range of outcomes. I find it's a little easier to understand than gsheets.

How Agencies Make Money

There are three major metrics that determine how much money an agency makes. It’s important to understand each one and how it drives incentives.

  1. Revenue — how much do they charge clients?
  2. Variable Costs — how much does it cost to service clients?
  3. Fixed Costs — what overheads do they have aside from client work?


There are lots of ways to charge for agency services, by the hour, based on value or as a retainer. Whatever way you pay, it always comes down to number of clients and average revenue per month per client. 

The way the agency charges will affect their behavior, but only in that it will affect their risk of collecting payment. This always backs out to an expected value based on risk, so if there’s a $20,000 bonus payment with a 10% chance of success, that’s worth ~$2,000 in cash.

Variable Costs

Regardless of how you make money, the number one cost of any agency is staff. They might be freelance or employed full time, but there is a fixed number of hours in a week, and an average capacity in terms of hours of work each client takes each month. Bonus payments, equity or other incentives can change behavior, but you can always put a risk-adjusted price on them and work out the effective hourly rate.

Now this is where you’re probably getting angry. You’re seeing that for the $5,000 a client paid, only $1,505 or 30% of it went to the person actually doing the work!? Worse, the charming, talented founder you wanted to work with in the sales meetings is barely going to be present, what with 39 other clients and a business to run. No wonder working with an agency feels like getting ripped off.

Fixed Costs

This is where the agency model is made or broken. Small agencies have very few fixed costs, so don’t charge a big enough gross margin. Then as they grow, and need to layer on marketing, sales, HR, finance, etc, they are confronted with reality — they either go out of business, stop growing, or learn to charge more.

These are just illustrative costs per client for a 40 client, 20 person agency collecting $200,000 a month. This isn’t exactly what Ladder paid (I don’t have access to those records anymore) but I made realistic assumptions based on my knowledge of the US market. Each cost is relatively small, but in real terms it means we’re down from 70% gross profit to 15% net margin.

Explaining Fixed Costs

In this fictional (but typical) agency, we have two executives (usually the founders) who pay themselves $150,000 a year. There are 3 in management, earning $110,000 each, and 1 person for each department of sales, marketing, finance and HR. 

I also added a $5,000 per month advertising budget (usually a mix of Google and LinkedIn Ads, blog content production, event sponsorship and production of sales materials). There’s also a $5,000 annual budget to advertise jobs (a recruiter costs 20% of first year salary per hire, so assume you can’t afford one). ‘Other’ captures the cost of SaaS subscriptions (easily more than $5k per month at scale), office space and travel & entertainment.

This adds up to an eye-watering $2,730 per month per client, or more than half of every dollar they pay you, just going to fixed admin costs. This is how whole teams of smart, hard-working, talented agency folks can work long hours and still be left in a precarious financial situation. 

You thought they were ripping off their clients, and yet, the agency owners are still only collecting $20 profit per hour from your account — and this is a healthy agency that charges enough! According to Campaign magazine, agencies should target 15-20%, yet the top 50 UK agencies reported an average of 10.4% profit margin.

Why Must Agencies Charge So Much?

Diving into the model, you can play agency owner and see why. Change any of the assumptions in blue on the right to simulate what would happen. Think the management team gets paid too much? Decrease the salary and see how little difference it makes. I hope you don’t lose them when they get offered a 10-20% payrise to go client-side (with less work).

Isn’t it unfair to charge our clients so much? Let’s see what happens when we charge less. Cut the annual deal size by just $5,000 – this is the equivalent of keeping a client for 11 months instead of 12, or giving them a $420 per month discount. This cuts the margin to a teeth-clenching 8% — if just 1 in 20 clients pays you late, you might not be able to cover payroll. Realistically it’s more like 10%-20% that pay late (which is why you need to make 15% margin).

Even at a healthy 15% margin, it would take you 6 ½ straight months of profit to save up 1 month of reserve cash. You can’t easily cut staff in troubled times, but what can you do? Lose 10 clients without replacing them (for example thanks to a global pandemic), you’d be running at below breakeven. With no cash reserves you’d be unable to meet payroll. In reality it’s even tough than that – in real life you have a handful of big clients that drive 80% of your revenue. Lose one of those and you can become unprofitable overnight.

Of course you can argue that these fixed costs are too high — of course that’s what I did often! What I eventually found though, was at scale, cut back on any one of these costs and the agency suffers. Cut your 1 sales person and now as a founder you’re doing 30 sales calls a week, and working late into your evenings and on weekends. Cut marketing and your sales team runs out of leads. Cut management and your team fails to get the training and supervision they need — it's a matter of time before a client asks for a refund due to someone’s mistake.

You might find, as most agency owners do, that the economics get better the bigger the client. The fixed costs don’t really increase by much to service big clients, so they’re spread out across a bigger base. There are also cultural norms around spending 10-20% of media on your agency, which anchor the amount you can charge low. For a small client though spending less than $5,000 a month on marketing, you can’t just charge them $1,000 to $2,000 – there’s no way to make that work in the model with fixed costs. It’s also a fallacy, because when you add in the cost of creative and other services, most CMOs spend roughly the same on agency services as they do on media (and the same again on technology and team).

Finally big clients are happier to pay more for less, because it’s not their money! Their business won’t fail if you do a bad job. That plus they act as sales material to help you close other big clients – deliver for one Fortune 500 company and more people will trust you to deliver for them too. This is why agencies inevitably move upstream when they wise up to these laws of economics.

When to Hire an Agency

You might disagree with my methodology or my numbers, but hopefully by now it’s clear that when you compare the true cost of doing business, you’re getting exactly what the market can bear to offer. Despite all the work they do to differentiate, Agencies are largely commoditized, so they exist under close to perfect competition — if any one agency is making too much margin, it’s only a matter of time before other agencies come in to eat their lunch. 

It should be clear by now that an agency isn’t ‘ripping you off’ – they’re lucky to be barely staying afloat! So they’re probably not ripping you off, they just have a very specific economic structure. Sometimes that structure makes sense to engage with, other times you can get much better results hiring a freelancer or building an in-house team.

So when does it make sense to hire an agency? In my opinion there are three stages to scaling a marketing function, and different structures make economic sense at different stages.

1. Discovery

If you don’t have a lot of experience with a marketing channel or activity, or your business hasn’t done much of it before, then DO NOT HIRE AN AGENCY. This is especially important for small businesses or startups, as well as new product launches for established businesses, because about 9 in 10 products fail. When things aren’t working, the last thing you want is to be locked into an agency retainer, and the last thing the agency wants is to be blamed for a lack of traction. Same goes for an employee — you’re doing them a disservice if you hire them and then need to fire them when it doesn’t work out.

If you’re an agency owner that grows to be successful, eventually you will learn to move on from these clients. They can be great to get started and build case studies, but ultimately they can’t be profitable for you at scale (especially given the thin margins we saw in the model!). They often churn in high numbers, which makes it impossible to grow past a certain point. Most successful agencies will filter out these clients with high retainers ($10k+) or minimum spends ($100k+).

If you really can’t try to do it yourself, this is the perfect opportunity for a freelancer. You need a very tight feedback loop to get new activities right, and dealing with just one person, who’s livelihood in part depends on you continuing to pay them (otherwise they have to go back to networking which they hate), aligns the incentives beautifully. Even if the freelancer costs a large hourly rate, it’s worth it for the flexibility of being able to try something with no long term commitment, with someone who was at least good enough to strike it out on their own.

2. Growth

This is the stage where you’ve proven something is working at a small scale. For example you ran $5,000 of Facebook ads and despite not knowing what you were doing, you saw great results. Or perhaps you contracted an SEO freelancer who got several quick wins over the first year, so now you’re driving 10,000 visits a month and it’s time to ramp up. The key is that you know the fundamentals of the channel, and have validated that it works for your business.

This is the time to hire an agency. You can actually be a good client for them, because you now know the basics of how things work. They have a foundation to build off, and should be able to run an audit to point out where they would improve things. Unless the freelancer you worked with is gearing up to start their own agency, you’ll find you’re outgrowing them. No freelancer should want to work full time on a client (and in many locations it would be illegal), so you’re only ever getting a fraction of their time.

As your activity becomes more serious, you’ll find you need better reporting, more responsiveness in returning emails, and knowledge gaps where you could do with a wider spread of skill sets, as well as a deeper bench. Now you’re ready to go find an agency. Make sure you interview them like you would any employee (and treat them with as much respect!). Three other rules that can help you when making this choice:

  1. Focus on methodology rather than results or flashy logos — it’s easy to get lucky or fake case studies, what’s hard is designing a repeatable process to get results.
  2. Don’t work with an agency ‘above your station’ — you never want to be someone’s smallest client, but as a top 3 client for an up-and-coming agency, you’ll be true partners.
  3. Be a servant-leader for your agency — handing everything over and expecting magic is not delegation, it’s abdication. You have to be an active contributor to get results.
  4. Meet the people you’ll work with — the agency might not like this because they can’t guarantee resources before a sale is closed (I didn’t) but it’s the right thing to do.
  5. Get references from ex-clients — don’t put them in a tough position by asking for references, instead get your own by cold emailing ex-clients or asking around.

3. Scale

So when does it make sense to take an activity in-house? For some channels, the right answer could be never! For example most clients don’t do their own branding, they hire creative agencies. Some clients maintain their agency relationships spanning multiple decades to their mutual benefit. This makes sense when your culture or volume might lead to worse outcomes for your company versus continuing to use an agency. 

For example most companies don’t need a new brand every year, and a sales or engineering-led company might struggle to hire or retain the best creatives. Think about this strategically for your agency — are you doing something non-core to the business, that is needed infrequently, or employing people your client couldn’t easily hire? If not, then you might find client retention particularly difficult, which means you’ll forever be husting on sales to replace the ones that get away.

Where it does make sense to take work in-house, is when you have some unfair advantage in the discipline. If this activity is core to the way your company makes money, or you’re amongst the top 10% in the world at it (or could be), then it’s strongly worth considering. Sometimes if you’re a ‘cool’ or ‘innovative’ company with a great culture, you can get better work out of employees who will work for less just to be a part of your team. However remember that also is true of agency employees, who will fight to be on the agencies’ most illustrious account.

If you do decide you need to take an activity in-house, take it slow. Remember that the efficiency gains will be relatively dubious (it costs the same amount of admin to run any organization of similar size). The chances are you can get more mileage out of switching agencies, than taking it on yourself. However when you do decide to make the switch, give ample time to do so: put the account up for review, give the agency a fair shot at retaining the business, be open about what changes they’d need to make to remain competitive. Give them 3 to 6 months notice if possible — this is enough time enough to close another deal to replace you, or make staff redundant if necessary. 

Most of all I’d implore you to take personal feelings out of it. You might not be getting value for money out of your agency — maybe they’re even genuinely bad at their jobs. Yet they’re just people, working for a living — they don’t owe you anything over and above the contractual agreement you made, which at the end of the day was just an educated guess at what you needed, made under conditions of uncertainty and commercial pressure. If you are unhappy with your agency, don’t be an asshole about it — nobody works more for their least favorite client, and word travels fast. Be straight with them, and if it doesn’t work out, move on.