Digital agencies: change your model or die!

I was running Adyax, a 350 people digital agency based in Paris, France, for 10 years.

The terrible January.

There is one moment in the year that is different from any other: the last and the first month of each fiscal year. In the last month, December in our case, we finally knew what was our exact revenue and margin. During this month we were chasing for the old, unpaid invoices, working hard to get all purchase orders and cleaning up our financial reporting to get an estimation of the yearly margin. I remember the excitement and sleepless nights. All 10 of them.

Then, once the New Year holidays were finished, January was the month of plans. We worked on projections. We carefully reviewed ongoing projects, to see how much revenue we’ll produce during the first quarter, then we go through all low hanging fruits, projects about to start and maintenance contracts. Finally, we would make long term estimates about our commercial pipe, playing with future RFP probabilities and potential growth of each account.

The final stage of the process was to see the gap between the revenues we actually see and the revenues we would need this year to make a 20–30% growth.

And, each January was terrible for our morale. We could see something like 2M€ in January, while we would need 10M€ this year. It was terrifying. Making 8M€ seemed impossible. Each single January out of 10, I was thinking: we’ll never gonna make it. We’ll shrink, we’ll have to lay off people, we’re bad, I’m bad.

Don’t be jealous.

Over the course of our adventure, we often met and talked with other entrepreneurs. One couple of them, ultimately became our friends, Maxime and Sebastien, founders of a hosting services company, Oxalide. They provided hosting services for large companies with 3 years-long contracts.

They started each year in a very different mood — only caring about what new business they will sign this year and how to please their customers enough to avoid churn.

I was jealous.

Then, we both sold our companies within a year from each other. And we had a launch all 4, Yann, my partner, Maxime, and Sebastien. While celebrating our success, I discovered that, while having almost similar revenues around 15M€, they’ve sold for twice the price.

I was jealous.

I’ve worked then for one year for our acquirer, Smile group, a much bigger company, with more than 1500 people and 140M€ of revenues. Surprisingly, despite their size, they had their January. We did our January battle plans in Excel, they had their in-house ERP, I was the only sales-rep in Adyax, they had around 100 sales guys reporting their pipe and projections, but at the end of the month, Marc, CEO of Smile had it’s own terrible January. I wasn’t jealous anymore.

No matter how big your agency is, if you don’t have recurring revenues and sell only time, you start each year from scratch. And if you sell your company, you will only get around 10 times your EBITDA.

That’s why I decided to create code.store. A unique platform to build, host, re-use and sell microservices.

Let’s see how you could change your business model to avoid terrible Januaries and enhance your company valuation.

Time is not enough.

First, let’s see how you generate revenues on a project, with an example: you just won a competition for a B2B e-commerce site.

You decide to build a classical monolith, using Magento. Magento isn’t the best solution for B2B e-commerce and you hesitated long between doing it from scratch using some frameworks like Laravel or Symfony, but you ended up selling Magento Enterprise.

Let’s see how you will generate revenues over time.

Depending on the phase of the project, you’ll produce more or less value based on the team working on the project. In the beginning, you’ll produce less, as you’ll work on discovery, produce some user journeys, architecture, designs. Then, your development team starts to work and produce more. You’ll have a spike in the end of the project where you intensively test the MVP, produce content, make security and SEO audits. Finally, once the site goes live, you’ll have a rapid decrease in the size of your team, proportional to the number of remaining bugs.

What happens next?

Then comes the most complicated part of your relationship with the client and the one where you can introduce innovation and create more value: the run.

Once a project is built, there are several services that may be provided by your agency, with different levels of margin and added value.

Usually, you’ll end-up providing some kind of complex contract (what SLA ? guarantee periods ?) and detailed invoices with difficulties to grasp your margins and project revenues. With questions, from your clients like: “Why should I pay for bugs ?”

But more importantly, as you add features, because of the complexity of your monolithic project, each of them costs more, takes more time and generate more bugs.

After a while, you’ll get a meeting with your client who will ask somehow guarantee you the quality of your deliveries. Maybe he’ll ask for a fixed price for new features, after all, the uncertainty of the beginning is the past, you should be able to make precise estimates for new features.

As time goes, your margins on new features fall dramatically near 0. You try to be more accurate in your estimates, but the client negotiates every single line (“how it’s possible that a LinkedIn Connect feature costs 10K€ ?! It’s just a button to add !!”)

After a couple of years, the relationship will ultimately deteriorate and you’ll be replaced by a new vendor.

Let’s see what we got at the end of the journey :

• Customer lifetime value: 1M€ over 3 years
• Overall net margin: 80-150K€ (more in the USA, less in EU)
• Value for your enterprise: around 1M€ for the period (x10 your net margin)

Because you sell only the time of your employees, each new project adds value to your company only while it’s active and only by a multiple of the net margin generated.

All the experience, processes, reputation, clients, know-how worth almost nothing when it comes to evaluating your company. They will take your EBITDA for last year, multiply it by 10 (actually it will be anything from 5 for generic outsourcers providing on-site or remote resources to 12 to agencies with some unique technological edge) and that’s it.

So to generate a bigger valuation, you’ll need chasing more and more active projects, hire more and more people. But as you grow, you get into more troubles :

• You start to be less and less attractive for young talents, your recruitment costs grow too
• The overall quality will decrease, as you lose control over each client
• You’ll need larger projects, which get scarce as large clients setup internal teams for the most critical projects
• You start to need middle management that produces no value but costs you a lot of money, just to make your machine work

And you start to be jealous of SAAS editors who have half of your revenues and headcount but are worth 10 times your company.

How could you do better?

What are the main differences between your agency and a SAAS editor? You both have a well-established brand, loyal customers, talented developers, architects, UX and UI guys, product owners. You build digital products for your clients every day.

• Predictable and recurring revenues
• Intellectual property

Let’s go back to the B2B e-commerce site and tell a different story.

While bidding for the project, you go through existing assets, built by your teams on previous projects, and identify a set of reusable API components. You discover that you already have B2B organization management, shopping cart, products catalog, authentication, checkout, and partial shipping. You’ll need to build a new front-end, integrate an order management system and a new search system.

You offer those existing features as a subscription which includes features, their support and hosting for a 10K€ / month.

You can do that because you build your project as headless e-commerce, based on a macro-services (like microservices, but bigger :) architecture and not a monolith. You can do that because you use a platform that enables you to build, host, re-use and sell macro-services.

During the project, you identify that you could re-use later their order management system on other projects. You build it as an independent and isolated component.

The client is happy because your bid is 30% lower and 20% quicker to go live. As you split your monolith into macros-services, adding new features is not that costly and maintenance costs are kept steady.

With the help of a platform like ours, you are able to get a foot in the SAAS business. You’ll probably never become the next SalesForce, but let’s see how this approach enhances your finances :

• Customer lifetime value: 1M€ over 3 years (30% less on the build, but with a 12x3x10 = 360K€ of recurring revenues)
• Overall net margin: 55–100K€ from the build but also a 288K€ from your re-usable components
• Value for your enterprise: around 3,5M€ for the period and even more for all projects in the future that will re-use your OMS module.

Clients are happy with OPEX vs CAPEX too. The absolute rise of SAAS VS lifetime license in software editors' business proves that. It’s way easier for the finances of your clients to manage yearly OPEX than estimate CAPEX amortization.

As time goes, you have a more and more competitive advantage during initial bids and you might even build new projects for free. You could wipe out your local competition entirely.

Think of it like Nespresso or printer hardware companies who sell their machines for nothing while getting a tremendous pile of money from selling disposables.

Want to try?