Why agencies get stuck at $50k MRR | Hacker News Reader

Why agencies get stuck at $50k MRR

jakobgreenfeld.com

159 pointsTomte89 comments

This is what happens when a Twitter Saas bro reads a few viral tweets from Design twitter and now thinks they understand an entire segment of the economy.

First of all, virtually no agency counts revenue like a Saas company (as “MRR” and “churn”), since the contracts most agencies sign are by definition not recurring. While some do sell retainer agreements, in general, the whole point of hiring an agency is to avoid hiring in-house for a need that isn’t ongoing or core to your business.

Second, there’s quite literally tens of thousands of agencies/consultancies that are over $50k in monthly revenue. All the big professional services conglomerates are constantly acquiring them (Omnicom, WPP, Publicis, Deloitte, EY, Accenture, etc etc) This is not a plateau that exists and the cited numbers are all totally made up.

Again, while it might be trendy on Twitter right now for people to try to sell “productized services,” this represents a minuscule percentage of the agency business. And agencies will never have true Saas economics, no matter how much they’ll try to tell you they do to get viral tweets.

You can see it in his solution which is to acquire more and more clients. Big firms don't sell like a SaaS. It's high-touch, long cycles, highly competitive and pricing is almost always negotiated based on scope. We'd spend weeks doing competitive analysis, writing pitch decks and case studies, spend thousands on travel to do an in-person pitch with a squad of slick experts wherever their office was. It's a tough game and not one that's easy to break into if you don't have the right frame of mind.

You created a story in your head that he wasn’t telling and then argued against that.

Clearly states in the intro this agency is selling a 2k/month package. Typical in a marketing agency this is what you would do.

Places like EY and Deloitte didn’t start in 2019, their success stories probably can’t tell you much about conducting a small growing business today.

Whilst it’s fun to bash people on twitter for being techbros let’s not let that cloud our vision

I directly argued the core points of the article.

There is no magical 50K revenue plateau, nor is there a math formula that dictates anything about agencies (they aren't Saas, humans cannot be replicated at zero marginal cost). And there are tons of agencies over 50K monthly revenue (if you have 3-4+ employees, which literally tens of thousands of agencies do, you're likely over that number). The core of the article is wrong.

It is not typical in a marketing agency to sell a 2k/month package. That would only sound typical, if your entire knowledge of the professional services business comes from reading that DesignJoy guy's viral tweets. If you're repackaging human hours into tiny 2K/month slots, you're going to have a bad time if you aspire to be anything more than an overworked freelancer. That's not a business model that scales -- hence why a vast vast majority of agencies do not sell tiny "productized services" packages.

The reason? Its the worst of both worlds -- tiny ACV (in Saas parlance) with crazy high marginal cost. 99% of agencies go for high ACV with more bespoke project-based or AOR retainer agreements. It's a 500+ year old business, so things are the way they are for a reason.

And no, I'm not comparing to Deloitte. I was simply making the point that hundreds of agencies get acquired every year and rolled up into conglomerates like Deloitte. None of them do this Twitter-trendy productized services model.

It appears the author discovered what any professional services business-owner already knows. Then tried to "thought leader" the space with their 6 months of knowledge and bunch of fake-analytical charts.

I think the OP runs a productized lead gen service, so they may be (inaccurately?) generalizing from their own experiences.

This article hits on something I've noticed before that is perhaps obvious to many but always seems unintuitive to me: past a certain point, generic "management" skill is the limiting factor irrespective of domain. Sure, it helps to have domain knowledge, and it's more important in some areas than others. But for a service-oriented business, hiring, organization, and client management (growth + retention) dominate at scale. Helps understand why people good at those things get paid so much despite "knowing nothing" about the business they're managing (in some cases).

My father-in-law has been slowly retiring from his contractor business for the last several years. He's down to working 40 hour weeks these days. It's really hard for him to step back further because he's still a critical step in so much of the process. He's the only person at his company who can do certain parts of each job.

Prospects want to talk to him, specifically, because his name is on the business and previous customers refer to him by name. He can bring other people into the sales cycle but can't completely delegate it.

He's got to sign off on the quote for the work (rehabbing an old kitchen, for example). The business depends on these quotes being profitable, so he needs to be extremely confident in the quote, and the work crosses several domains within the crew. Recently he's been training a back office employee, who's seen years of work and how much it costs, on how to do this well.

And he's ultimately the one who has to inspect the finished work and ensure it's up to the company standards. Typical manager responsibility there. If someone on his crew does a crappy job, he's the one getting a phone call and has to rearrange the schedule to address rework.

Anyone with interest and aptitude for these tasks is long term aiming to start their own contractirs.

I don’t know how it relates to construction but if you were describing a software development studio I would say it sounds like your father in law hired too many junior and/or incompetent and/or unmotivated employees and then micromanaged the process + done too much of the work himself to compensate.

Question is, now that he's in that hole, how would you go about getting out of it?

The same way you avoid it: Succession planning. You need to do the same thing as before this becomes an issue, but now you have less time. You need/want battle plans before a battle, but if you don't have them before you have to make them during.

I don't think there's a contextualess answer, but broadly you either find a direct replacement that you mentor (or maybe you're lucky and they're already GTG) or you diminish your role and have multiple other roles within your org (or outside, i.e. outsource) take up what used to be your responsibilities until your role is obsoleted.

Probably try to hire someone more senior, preferably with management, sales or bizdev experience. And then start pairing with them until they have his specific workflow down. It may never be anyone as good as he was but the alternative is that the business goes under when he can no longer work.

You can hire great management talent by adopting low-level titles.

In your case he just needs to hire "coordinators" that coordinate the work and quotation and have some P&L bonus.

In construction, former architects are great for this type of role

Lack of domain knowledge usually increases the amount of management you need too.

In software development for example your teams become much larger because you made the wrong decisions by not understanding the scope and requirements, selecting the wrong stack and tools and hiring the wrong people.

So now you need more management all the way from HR to procurement, engineering, support and legal to brute force your way into delivering.

And all these salaries and benefits need to be paid which means now there's less money to hire and retain qualified engineers and to invest in improving your products and services.

Once an organization falls in this cycle is very hard if not impossible to change, it becomes part of the culture.

> "management" skill

You mean sales and marketing? Because there isn't anything about management on the article.

Read a little closer. The thesis of the article is that you need to transition from personally managing sales and marketing to hiring someone else to manage them. It's a difficult transition that requires a different skill set than just domain expertise and hustle.

The article assumes you hire perfect people and manage them perfectly. If you manage them improperly, or your the people you hire are bad, the numbers certainly go down.

But the point of the article still stands even if you do things perfectly, it's the agency business model that is the limiting factor here, I think.

Seems like the author is just working backwards from a conclusion, picking numbers that "work out" to $50k. Obviously growth plateaus when churn matches new + resurrected, that's just basic growth accounting. Why is $50k such a barrier? Is it even?

It's pretty clear by the second sentence of the article that the author has no idea what they're talking about and is just making shit up:

> Hardly anyone ever breaks through that [$50k MRR] barrier and reaches that next level where the agency generates reliably $300k+ in monthly revenue.

Why is $300k the "next level" after $50k? So if I have a $50k MRR business and 5x its revenue, I'm still not at the "next level?" I've never heard a single person say that and they present it as if it's fact and move on.

It's just concrete numbers to help the reader visualise.

And as a reader who's worked at agencies before, he's right.

The only part he's missing is that most agency owners grind on with this because of personal (psychological) limitations, and thereby get stuck. They get mentally capped out, so they can't move forward, and they form the core of the business, so they can't sell out. All they can do it stay locked in grit-your-teeth mode, until age and/or health fixes the situation for them despite their will.

Agencies cap out very much like this, because of their fundamental scalability issues, but that's also why people enter that market. Easier to enter, because no one person can dominate it like a Google or Facebook.

The number they cap out at is very tied at the intelligence, work ethic, etc, of the founder, but it always caps out in that asymptotic graph kind of way.

The only way I've seen them successfully break out is to know this effect, and to use the cashflow to cover a product company offshoot that gets traction - basically a self-funded startup. But it's a rare case, because they we're attracted to the immediate and certain money lure of agency, vs the valley of death gamble of product.

I also imagine a fair share of owners are actually content enough at ~$50k MMR. If you have expenses under control and not too much staff that can work out to a decent profit/salary.

> And as a reader who's worked at agencies before, he's right.

I’ve worked at a load of agencies and I don’t recognise what he's saying at all. What agencies sell $2k/mo packages? They are all either much larger finite projects or they have much larger retainers. What would you even sell for $2k/mo? That’s not going to cover anything beyond absolute basic maintenance to keep the lights on; it’s certainly not where agencies are making their money.

agree. From my experience I’ve also noticed that agencies usually have a shelf life of 10-15 years max. At the start they are the young fashionable agency doing great work. At some point due to a combination of great employees leaving for bigger jobs and the company getting bogged down with stable but boring long term clients, the company dies.

Sometimes they end up cashing out and retiring. Most often they go bankrupt or get bought out by a consultancy looking to brand themselves as fresh.

It’s fun working in ad agencies on the uphill. You’ll be able to do great work and learn how to ruthlessly prioritize quality output. On the downhill it can really corner you with little career growth and little pay.

Yeah, they used a bunch of completely random magical numbers: $2k/mo/client, 2 new clients/mo, asymptotic churn rate of 15%. All they’ve demonstrated is there’s likely a ceiling somewhere for each agency, which is obvious to anyone.

Big thumbs down, save yourself the click.

> that’s just basic growth accounting

And a monad is just a monoid in the category of endofunctors.

This stuff is not obvious to everyone. The arbitrary numbers are called an example.

The point of the article is to say, when you experience a plateau and you begin to speculate why, it’s mathematically expectable.

It's okay to use specific examples to explain a general principle or to make your point more clearly. But when those examples are more-or-less made up, as they are here, you'd better:

1. Be clear that they're examples and not some fundamental, mathematical truth.

2. Use multiple examples with different numbers to show how things could look different, even if only slightly.

The author does neither of these. #1 is violated in the title. It's not clear at all that the author is using these numbers as merely examples.

This use of the term math irks me. Math exists within the bounds of its own rules, and often it explains the world well. Sometimes, math can be used to make predictions about the world. But to infer that such a prediction is valid „just because I can make the numbers line up“ is IMO too hand-wavy and maybe even pseudo scientific. I wouldn’t drag mathematical models into this so early in the process.

He’s expanding on how rates grow in proportion.

It can hardly be controversial to call this math.

But ok, it’s not “the mathematical reason” but rather “a mathematical explanation” - the math is a property of his explanation, and the reason is growth, churn and employee psychology; while well-explained mathematically, they aren’t math.

Somehow this seems to be besides the point, being: here is a common reason for plateau’ing.

Is it common? The point is asserted at the top of the article, but no data is provided to back it.

Reminds me of our introductory course in math at uni. The lecturer was showing a concrete example of how to integrate using the technique we'd just learned.

After a lengthy manipulation of the integral, he said "and then you can just plug this into a calculator and get an answer, but that's not really interesting so I'll stop here".

What I mean is this: Making some anecdotal observations that can be explained by a certain model, doesn't mean the real world follows this model. It might not even be a good approximation.

It's not obvious that when the number if new customers acquired = customers churned, growth flattens?

Founder of a solo dev $500k/year agency here.

10000%.

I topped out around $500k/year. Just naturally settled there after aggressively optimizing my process.

This is with 1 dedicated sales person who's entire job is to minimize and replace churn.

Good to see there is a mathematical proof haha.

I've been exploring productization and have made it a goal for 2024, to possibly breaking this ceiling without hiring any one else.

I have a $2 million/year dev agency w/ 7 employees (50% gross margin) and am curious - how did you land on the salesperson you have? I have not had a single salesperson yet and am wanting to grow, but am afraid of hiring someone who will just waste time.

Hiring a sales person is easier than a developer because they will accept commission as part of their pay package. Commission aligns your goals with their goals. The more they sell the more they make. Hire a few because they will compete and be even more productive. Fire if targets are not met.

Your wasting your own time not hiring. Sales targets are the easiest to measure.

But not in the grandparent's business (solo dev). There is a natural cap on the amount of work the solo dev can complete in a year (point of the article) and as such there is a natural cap on the amount of work the sales person can sell and fulfill in a year, which caps their sales bonus. It might not even be that big.

poached from a competitor, they tried to sell the exact same service i provide to my product company, and they did a great job.

How did you go about hiring your one dedicated sales person and what’s their compensation like?

poached from a competitor, they tried to sell the exact same service i provide to my product company.

80k base, 20% commission

With benefits I hope!

Why 1 sales person? Last agency I worked for had like 25 people dedicated to sales and everyone in leadership of every discipline participated in pitches. We did something on the order of $50MM/year.

I'd strongly caution against productizing unless you have inputs telling you customers want it. It was tried at every agency I've worked at and failed every time.

> So as soon as you have other people take over service delivery, you will inevitably experience a drop in quality. And this will lead to more churn.

IME, this is not correct. Many founders are crap at service delivery; what they're good at is client acquisition. And good developers don't generally make good salesmen. So at the point when they decide to hire real developers, service delivery will improve. The founders can stop monkeying around with code, and focus on selling.

I mean, I used to work at a digital agency that pulled in way, way more than $50k MRR. It did high quality work across a range of services (copywriting, web development, graphic design, social media engagements, online advertising). Mostly it worked by having long term retainers with large clients (I remember being amazed that we wouldn't even really look at jobs under 50k). It had a lot of younger staff but the leadership was experienced and serious.

It was a business that was really doing well, with a huge network of contacts in the region. I'm not saying it was representative of all agencies - probably it was an exception to the rule. But I definitely got the sense that there were ways of making the agency model scale well beyond 50k/month.

The article's implicit assumption is that there's only one grown-up/owner to manage client relationships. If your company has N of those, I guess you can reach $50Nk MRR.

I don't know what precisely the definition of "agency" is, but I've also cofounded more than one service firm that drastically exceeded $50k MRR, and I still think this is a really interesting, insightful post. The number is least important part of the piece.

> The number is least important part of the piece.

Yet the most important part of the title

Long retainers means more retention.

Some agencies and consultants, if they are really good at what they do, can demand revenue sharing. Meaning that they provide a service (SEO, for example), and then get a piece of the revenue - given that they hit some threshold.

In that case, the sky is the limit. I met one consultant that netted nearly $10m a year doing that - he had specialized on SEO in a cash-rich field, and got some pretty big clients. In his words, hourly billing is for suckers - there's a natural limit for how many hours there are, and the only way you get more revenue that way is by increasing the number of employees. Apparently when going from hourly billing / fees, to revenue sharing, he jumped 10x-100x in annual revenue.

Of course, most can't just demand that - and many clients will refuse. But some will, if you're good enough.

This post sounds reasonable. If I could hire a SEO and do profit sharing, it would make good sense.

    SEO in a cash-rich field

Can you share the field?

Credit cards, consumer lending, etc.

Middle ground between serious banking and payday loans.

The article's assumption about the retainer - $2k/month - is its biggest flaw.

What are you giving away for $2k/month per client?

Seems like it should be much, much higher - and therefore less people total to serve. Seems like a win-win.

Most agency owners I speak to (in VC marketing circles) charge minimum $8k a month, not $2k. $2k a month is like an overseas sweatshop.

the secret to making 300k mrr is to operate just like the 50k mrr business but charge 12k instead of 2k.

In that case the real secret is how to convince your clients they are getting 6x the service level of a 2k customer while really giving them the same thing.

Most corporate math I've seen doesn't work like that. If you can sell a product to a corporation for $2k a month, you probably can sell it to about 95% as many customers for 3k a month. Usually the person that is agreeing to the price doesn't directly use the service or isn't able to judge actual usefulness of the product.

For a 2k price level it probably is actually the user making the decision. Maybe that's the solution here! If you charge 20k, now you move into the corporate BS tier where there is no accountability to the user and you can just manage the relationship with the decision maker

That is why paradoxically the higher the price the easier the sell, and conversely the lower the price the more nitpicking the customer will do.

I always assume the bottom 10% of any market is a cesspit that is best avoided whether you are a customer or supplier. I suspect the top 10% is the same but I have little experience of that.

I mean, yes? Yes, there is something to this.

Remember, 50k mmr is what some Googler make slinging code.

Like, I think the most important difference between someone who charges $2500/mo to their customers and someone who charges $15,000/mo is that the latter person asks for more money. There are other differences, but they follow distantly behind, not closely.

I don't think this is a cynical observation. I think some of the same kinds of dynamics this post talks about are at play in selecting the types of organizations that make those asks; the causality is important though, because it starts with asking for more money, it doesn't end up there.

The funny thing is the guy charging $15K a month will get treated a lot better than his cheaper peers. People really do value things by how much they pay rather than some objective worth.

My last employer pulled out of a services business so I left and started contracting for the same customers. The only distinction was I charged three times as much as my employer had been. Not only were they happy to pay but their whole attitude towards me noticeably improved.

- 2k/month is low. Depending on the services you provide. - There is an 80/20 rule on per client revenue.

So this already makes the math wrong with these two assumptions. Actually the whole article reminded me of the Law of Diminishing Marginal Returns applied with wrong assumptions.

I worked at loads of agencies that were millions or tens per month. None of them really used MRR as a metric because work was volatile. Most of the biggies that belong to the major holders like WPP, Omnicom, Publicis are that big.

I'm maybe confused by what OP is considering an agency. A group of people making WordPress themes are technically in the same industry vertical as Accenture or IBM.

I suspect any ceiling on earnings is more about the owners ambitions. I'll go out on a limb and say most people running these businesses aren't looking to get wildly rich, they just want to replace a job for themselves. For whatever reason they don't fit will with the traditional employee role so start a business as an alternative.

So this was all about the dis-economies of scale. Large agencies loose more clients over a set period of time so clearly there is a limit how large they become. What is totally neglected are the economies of scale: Larger agencies have a larger network adding more clients over time. The fundamental flaw is to see the service delivery as purely service delivery and not also in part as a sales job.

The dis-economies of scale usually grow faster than the economies of scale (network effects etc. may kick in and deliver a boost over some limited time) so there is a little truth to the article but it still feels like attacking a straw-man put in place by the author.

The more clients, the more chances that any will leave. That's called variance, aka, risk.

Economies of scale reduce or help the amortization fixed costs, not variable costs.

Economies of scale have more risk. The right approach is to diversify the risks.

> The more clients, the more chances that any will leave. That's called variance, aka, risk.

The less clients you have, the worse impact it can have if you lose the contract with one of them or the client downsizes.

Say you run an agency that has only a few clients, and (almost) all of them are automotive... and then covid hits, leading to budget cuts in the entire sector as the carmakers can't sell cars because they don't have chips to make them with. Now, you're fucked unless you have iron-clad contracts with fixed minimal spending amounts or retainers which is a rarity in the business. This happened here in Germany and it was an awful time.

And now you run the same agency, but have a multitude of small, medium and large clients, among different industries... so what if your 1, 2 automotive clients cut their budget, you just call up the other clients and ask them "hey, we have a bit of capacity left, wanna do <X>?".

> This happened here in Germany and it was an awful time.

Because you didn't diversify your economy. Just like my last sentence said.

Sorry about your country but my ex wife is German and... yeah, sorry. You didn't learn your own history.

It's funny because I'm argentinian and we don't diversity our risks either. And now Germany is comparing themselves with Argentina:

https://www.thetimes.co.uk/article/germany-sick-man-europe-e...

>> This year Germany’s GDP is forecast to shrink by about 0.4 per cent, the worst performance of any large country in the eurozone and any member of the G20, with the exception of Argentina.

What kind of "agency?"

In this context, they're usually web design companies that specialise in Wordpress sites.

The article ends with what feels like a cliffhanger:

>It is very unlikely that you can 10x your acquisition efforts without a complete overhaul of your approach.

So how must you change your approach? Is that next week's topic?

SaaS founder here. I can confirm that the first part of the article is true: you do get "stuck" eventually at a certain level, because your churn rate is a percentage of clients, while your new client acquisition rate is a constant (e.g. it does not depend on the number of your existing clients). So the two will eventually become equal to each other and your growth will slow and then stop, unless you start/increase your marketing and client acquisition efforts.

I don't like the word "stuck", though — I do not subscribe to the idea that endless growth is the Holy Grail, to be pursued at all cost.

I know, we are not supposed to discuss voting. But I'll still say that seeing an informative comment based on real-life experience of being a solo SaaS founder being downvoted is discouraging. It's a signal that I should stop contributing.

The article is about agencies and you are a SaaS founder. These are very different types of business with very different economics.

It’s fair for you to chip in with your experiences, but you shouldn’t start off by saying you are confirming the article is true, because you can’t. If you were speaking from your experience as an agency founder then you could confirm it, but your business is very different to an agency.

The problem a lot of people here have with this article is that it seems to be based on SaaS economics not agency economics. So for you to say you confirm it as a SaaS founder is propagating the problem people here see with the article.

Can we downvote? This guy can't comprehend that with consistent good work, there will be network effects where in you get bigger value clients/projects?

You can use "flag". For articles it acts a bit like a super down vote. (Whereas for comments it means please get mod attention).

This is really well laid out and digestible.

FWIW this is what growth teams think about all day when putting cash to work to grow the business.

Revenue = new revenue + expansion - contraction - churn

Jakob Greenfeld

https://jakobgreenfeld.com/

Some companies solve this with Franchises. Others with performance linked bonus.

What other options do we have?

You can’t scale when you’re working in your business instead of on your business.

TL;DR; Make sure your get 5x-10x clients per month vs churn.

The $2k number isn't the issue here. It's just a number, if you disagree with it replace with with something else. The point is that the author argues there's a ceiling to agency earnings.

That being said, the article is wrong because of other, more fundamental reasons.

The first point essentially argues that churn goes up as the client list grows (because it's a constant percentage), so at some point the new incoming clients will be the same as the ones churning so the agency will be at sales equilibrium. This relies on assuming that the rate of new clients is constant. I would expect agencies to get better at client acquisition as their reputation and experience grows. So you'd have to show that this growth rate is slower than churn rate at some point which isn't obvious/proven here.

The second point states that the owner runs out of attention with too many clients regardless of how many talented employee they hire. This, however, explicitly assumes that owner spends an equal amount of time between each client. A wiser strategy is for owner to concentrate their energy on high-value deliverables and touchy client relations and farm the common/generic work to employees.

The math just makes bad assumptions.

Reply on news.ycombinator.com