Way too many years ago, during a family reunion, we went white water-rafting. It was fabulously fun. The rapids were barely that - these were Class 2 rapids.

A few years later, I went on a boys' trip where the rapids were rapids for sure (Class 4 rapids). I don't have any pictures of that trip because somewhere along the way, we lost my dry bag with my phone and wallet when my raft undertook an unexpected “vertical inversion” as my nerdy engineer friend opined.

Right now, agency growth is being hampered by a lot of external factors. But almost all of them, like economic issues, political issues, global conflicts, and the like, are all known territory. These are Class 2 or 3 rapids - they can be challenging, but most everybody is going to make it through by doing what they know how to do… Agencies will have varied levels of getting soaked, but few are going to end up in the drink.

We've all been through technological changes before. Some have been hard, and some have been easy to navigate, but AI is unlike anything that we have ever been through before. AI presents itself horizontally across every single agency type that I can think of. In many cases, AI seems like it has been specifically designed to disintermediate agency businesses.

In 2020, I had a great agency (the team at R11 Solutions is terrific - you can see my old site on their portfolio page) rebuild my site. They did an amazing job - and it cost a lot of money - 100% worth it at the time. I was switching some things up this year, and I was futzing with AI…and I built a brand new site - all by myself - it didn’t take all that long and it was 5 figures LESS expensive.

Is my site as pretty or as “clean” as it once was? No. I don’t have the same aesthetic sense or talent as R11. Is it way better for me? HELL, YES. I can build pages like my Agency Research page or any of my Agency Growth Tools pages without need for a designer, developer or account manager.

In fact, I’ve used AI to build tools like the WTF Agency Assessment & more complex tools like Call Lab Pro. AI has allowed my to build things that I have always wanted to build - like Agency Field Guide - an evidence-based agency directory that isn’t going to sell “premium placements” or any other BS like that. I don’t know quite yet how I am going to cover my costs quite yet, but I’ll figure it out. Right now I am primarily using it to learn about what agencies say about themselves. Yesterday I published my 1st data finding:

In the past, building a site, complex tools & an agency directory (BTW, want your agency listed? Submit your agency.) would have taken something like $150K in agency fees. Anthropic (want a free week of Claude Code - I have 1 gift week left) & OpenAI have built them for me for about $2.3K in subscriptions & token usage.

That’s something like $147,700 that DIDN’T go into the agency ecosystem. That’s why this is a Class 5 rapid.

Every customer signal is captured in one shared context layer, always current and compounding. Agents and workflows build pipeline, chase every buying signal, and move deals forward, an always-on revenue engine running alongside your team.

With Attio, you’ll get:

Teams like Parallel, Turbopuffer, and Wordsmith build on Attio. Are you one of them?

Get started now

Well, TBH, yeah. AI has freed me to build things that I've always wanted to build. Are they all perfect and bullet proof? Fuck no.

A a ton of your clients are building things on their own right now. And many of you are using Claude Code or Codex for building apps and data analytics, or Perspective for funnels or HeyGen for videos.

Many of us are bringing stuff “in-house” that we used to send to someone else (the “out-house”?). That’s the turbulence. For about 110 years, agencies built their value on what they DELIVER or BUY. That’s why for so many years, agencies billed time & materials (how many hours a thing takes) OR on a commission basis (percentage of the media buy). They gave away creativity and strategy as a sweetener to entice the prospect to buy the work.

Today, as evidenced by my own actions, the work isn’t the win. It’s being able to decide what to work on, design the output & know how to determine success. While the deliverable work is still REALLY important (and it will be for the foreseeable future), your ability to charge for it is being compressed more rapidly than anything I’ve witnessed in my 30 years of agency experience.

While it has always been your truest value, your understanding of what is the right thing to do for the client, based on the client’s circumstances, resources and market conditions is now your best pricing (and perhaps, only) lever.

First of all, the goal should never be to use AI solely to do stuff that you already do. So many agency owners use AI to accelerate the amount of work that they do, knowing that they can generate more work by using these tools. People are thinking about AI as a capacity enhancement.

There is there anything wrong with using AI to make you & your team more efficient. But the most important question is what do you do with those efficiency gains? There are two options:

Option 1 keeps you in the current state value proposition. And since everybody is pretty freakin’ sure AI is magic, the pricing power of the current state is going to decline. Over time, the market will going to be willing to pay less and less for the deliverables associated with the current state value proposition.

Option 2 sounds pretty flippin' fluffy, doesn't it? But the emphasis on the planning and measurement and all of that thinking actually puts you in a position to be more outcome-focused.

That’s where we want to be, my friends.

Blu Dot used Roku Ads Manager to drive incredible results for its furniture sales event. Its strategy hinged on custom audiences and retargeting, where intent was strongest.

“Roku has been a top performer,” said Blu Dot’s Claire Folkestad. “We have seen…CPMs lower than any other CTV partner we've worked with.”

Learn More

I just talked to an agency owner today who wanted my opinion about his SEO/AEO offer. Lately he has been getting questions like “How many hours will be spent on my site every month?” He wanted to defeat that objection by being deliverable-focused, and he was super excited to, in his words, “do some margin-mogging” by focusing on perfecting his agents that can generate the output.

His offers were identical except for the number of deliverables. The “basic” package was 10 well-written & schema-rich articles for $X, the “better” package was 25 well-written & schema-rich articles plus some better reporting for $2*X. There’s no indication that a ladle of 25 is better for your business than a dollop of 10. They’ve never sold the “better” package.

Ugh.

A client of mine who offers a similar service, tossed away her number of deliverable-based approach after reframed her value to her client around trackable pipeline. By tying herself a measurable outcome, it almost doesn’t matter what she delivers in a month or a quarter as long as her team’s efforts connect defensibly to pipeline growth.

This client, BTW, is an AI-driven MACHINE. There are strategy agents, research agents, pre-production agents, outline agents, draft agents, review agents, human writers and editors, production & post-production agents, data agents…SO MANY AGENTS.

Since the switch to pipeline measurement, her clients have never once asked about number of deliverables, number of hours, or even pushed back on strategy because her team & her client have the same measurable goal. Her margins, customer satisfaction & retention rates are all higher with her pipeline clients vs her legacy deliverables clients.

Yes, but also, no. The outputs have a similar shape and, in many respects, a similar scale, but the level of trust and alignment is so different that her team now has the freedom to make changes to what outputs are created based on the outcome.

The contractual obligation is actions connected to outcomes, not deliverable made or pages touched. Leveraging AI has unlocked her ability to measure those outcomes in better ways, and AI has freed up so many human resources in terms of data analysis and research, that the humans on these accounts can do better work. In fact, the work output varies every month based on the impact of previous work. This allows for better focus. This allows for midstream adjustments, and most importantly, it allows her team to adjust to strategic changes or reprioritization from the client, without requiring change orders or rescoping.

AI has made the strategy planning and the human contribution to content so much more valuable that now her team has been empowered, essentially, to do whatever generates the right result for the client. At at some point, I am sure that there will be a conflict between what the client requires in order to hit their goals and the amount of work and/or tokens that need to be used by this agency. That is a completely different discussion from having to defend the value of what was originally scoped.

Check out Agency Field Guide & let me know what you think, and if your agency isn't one of the 5,000+ we already have published, please submit your agency, and we will get it included.

TLDR Marketing is the free daily brief that 330K+ growth marketers, performance marketers, and CMOs actually read. The most interesting stories in marketing, curated and summarized in 5 minutes.

Subscribe for free

Our sponsorships are ads (obviously - duh). But some of the companies or services linked in our content may pay us a minimal commission if you buy something from them. Potential payment is never a factor in what we link to, and we have direct experience with every company or service where you can buy something that we mention in our content.