AI in Agencies
Agencies are often bleeding edge adapters of tech (we are nerds who like to mess around). Ad platforms, of course, have been using machine learning and AI forever, and the advances from OpenAI, Anthropic, and bunches of others has been nothing short of breathtaking.
There are workflow innovations – AI is built into Notion, Monday, ClickUp, Google Workspace, & Adobe creative tools. There are targeted platforms like Jasper.ai, Copy.ai specifically for content development. There are ad creation tools like AdCreative.ai. There are ad targeting tools like GoAudience. HubSpot has gone all in on AI with Breeze. And there are roughly 12 million AI-powered “Sales Development Representative” platforms in the world (none are listed because I have no idea if any are good). And I am sure that it is just a matter of time before Beehiiv is better at writing my email newsletters than I am.
As an agency leader, you must have a clear view of what is happening in the market, and a vision of how market changes can impact your team and your clients. That is part of what it means to be an entrepreneur in the marketing arena. You are constantly looking over the horizon for opportunities and threats.
It’s your job to be at the edge here.
Strong Signals
At the end of Q3 2023, Dale Bertrand from Fire & Spark shared with me that the AI tools he’s been using had saved his agency 500 labor hours a month (30%), and overall the quality of output was equal to or improved compared to their baseline output. (Note – its not that any AI generated output was better than the baseline, but the workflow efficiencies gave humans the time to make the output better…) At this point, I am sure that he is at or near his 50% time savings goal.
Sam Wright of Blink SEO has corralled his 2 decades of SEO work into a powerful platform that allows he & his team to do complex SEO optimizations for ecommerce sellers on Shopify. They estimate that they can perform SEO work in less than 20% of the time it used to take them. That’s EXTRAORDINARY. While Blink does incredible stuff by leveraging SEO and behavioral data to power multiple marketing channels, they have peeled off their SEO toolset as a standalone SaaS – Macaroni.
Taylor Holiday of Common Thread Collective shared that CTC had doubled their operating margin because of AI.
This is crazy stuff – in fact, this is the biggest shift in agency economics that I have ever witnessed. AI, in the short run, can save your team tons of time so that you can handle more work with the same staff. It can help them focus on value development rather than workflow issues.
Power Performance – How To Minimize Risk & Maximize Impact
All of this AI wonderfulness comes at a cost. You’ve probably seen the number of your monthly SaaS subscriptions explode over the last 12 months. Further, integrating AI into any proprietary platforms, or building data lakes/warehouses to augment LLMs is likely going to require tech resources that are and will be in short supply.
And all of this AI feels like it comes with a risk, too, right?
If your client knows that you are using Claude or ChatGPT to write copy, and you are using Midjourney to make a promo video (btw, check out this game trailer someone made – WOW!), and Go Audience for ad targeting, and Hive Science to find winning messaging & creative, why do they need you?
It’s a good question.
After all, agencies are experts in channel execution, right? They get hired because they can execute all of the tasks needed to launch and manage a campaign, right?
But the answer is really simple.
AI, at this point, simply follows models and makes adjustments based on feedback. (BTW, we do that too. When you write copy that you think is going to work, you follow a model based on other stuff that has worked. If the results aren’t what you expected (feedback), you adjust and try again.) Of course AI is going to change, and perhaps there will be times when AI generates content or a process that seems unique. But it won’t be truly unique or creative. It will simply be iterating according to its ever adapting model.
AI is not a threat to agencies because AI mostly removes the drudgery involved in the output of marketing. It can whip up a serviceable copy draft based on a few inputs. It can generate and adjust images in seconds that might have taken a human hours to create. It can make math models to maximize campaign ROI faster than we can.
Luckily, that stuff isn’t really what marketers get paid for…
We get paid to understand what resources we need to create the desired impact. We get paid to imagine what should be in the product’s hero shot to show it off in the most desirable way. We get paid to say, “What if we tried…?” or “Do you remember how it felt when you first did…?” or “You know, I think we can do better…” or “Hmmm. Let’s think about why that didn’t work…”
Agencies aren’t at risk because marketing isn’t at risk. For the most part, clients hire agencies not because they CAN’T do the marketing stuff, they hire agencies to explore and find the most effective marketing stuff for that business.
It’s OK (in fact, it’s desirable, IMHO) to let your clients and prospects know that AI is part of your toolset. You don’t hide that you use FB Ad Manager or AdWords to generate results, do you? AI is the same – it is a tool that helps you deliver what the client wants and needs. When you tell them that you are using AI, they might offer data that can impact AI output via RAG (Resource Augmented Generation), or they may offer you an API endpoint where you can extract data that will make AI significantly more relevant to the client’s needs.
Further, the tooling and technology in generative AI is changing SO QUICKLY that your prospects and clients will rely on your experience and insight as to which tool stacks are best, what data should be used, and what prompts and workflows create the best outcomes.
AI isn’t here to take away your agency.
In fact, AI has just given you a series of promotions, in a weird way.
AI can make your internal workflows better so your economics improve. Now your agency has more capacity without additional overhead. You were just awarded a better agency.
AI relieves some of the hard initial generation work from your team so they can focus on projects that use more of their high impact skills to make things better rather than just make things. That means you’ve just become a better boss.
AI forces you to become an explorer so scour the market for the best solutions. So you’ve become a trusted advisor to your clients. More than that, you can become a curator for them, someone who can make the complexity and dynamic nature of the AI market into an approachable set of options perfectly suited to them.
Editor’s Note: This article first appeared on The Agency Inner Circle, a free Slack community, newsletter and podcast built to help agency owners grow. It’s free. Join the agency growth fun here. And we may receive compensation if you purchase goods and services from some links. That never influences our decision to link to a resource or not. We only link to products and services with which we have direct experience.