What Does AI Mean for Ad Agencies

I have been getting a lot of questions from clients about what impact AI will have on the advertising business.? In speaking with people from across the industry, I have been hearing some consistent perspectives on where we are, and what to expect from AI in the future.

1.?????? It is already having an effect, but it is limited to just making existing processes a bit faster and more efficient.? Everyone is experimenting.

2.?????? Eventually, it will lead to lower staffing, but focused on the junior levels.? Mid/senior staff will still be needed to oversee and refine the output of AI tools that will replace work done today by junior staff.? There will be fewer junior staff and those that remain will need to know how to work with AI tools to create outputs and help it to learn over time.? Think of these staff as becoming more like mini conductors of an AI orchestra rather than playing the actual instruments.

3.?????? Near term, production will face the biggest impact, as AI tools produce artwork and even basic ads/commercials, the need for production staff (or use of external vendors) will decline.? Production is often a pass-through cost, so this should have limited impact on agency P&Ls…unless they have large production departments.

4.?????? Major change is 3-5 years away – until AI output can be fully trusted, it will require 100% review by humans.? As fast as AI is developing, getting to 100% trust still seems years away.? In the meantime, it can do menial tasks and start projects, shortening development timetables and improving the efficiency of current processes and human deliverables.?

5.?????? It will empower competition for large agencies - For smaller agencies and in-house client staff, AI will allow them to produce high quality output for basic projects competitive with larger agencies.? As AI capabilities expand, the cannibalization of simple assignments given to large agencies will grow.? Large agencies will still be looked to for major, strategic, creative initiatives, but total billings could decline over time.?

Specific impact by function:

·???????? Creative - AI will help agencies produce volumes of concepts quickly that can then then be culled, refined, and edited by experienced staff.? For basic versioning, AI will replace work done by humans, say replacing a graphic in a versioned email.? Eventually, creative directors may develop their own personal AI tools that adopt their style and have been tuned to their creative approach over time - those trained, personal AI assistants may replace the portfolios they have today and be the way creatives differentiate themselves in the job market, almost like a personally trained assistant.?

·???????? Production – AI’s ability to create artwork from scratch will replace some real and virtual ad shoots and postproduction.? Directors will need to know how to give direction to the AI tool to prompt it from concept to finished deliverable with AI replacing a production team.

·???????? Standard Reporting – AI has the potential of automating reporting for functions like finance, HR, media, analytics, and account management.? Humans will still need to review and refine any insights the AI may generate, but basic reporting should require fewer human touches. This will depend on linking the AI tool to the data streams needed to create the reports.? Ultimately, it should require fewer people to do this work.

·???????? Media – Basic media planning and execution will become increasingly automated, especially as AI tools are linked to execution platforms and performance metrics.? Humans will still need to produce overall strategies, but AI will be able to propose initial approaches, so a media planner refines that plan rather than creating a new one from scratch.? Like Creative, this will mean fewer entry level jobs.

·???????? Clients- Clients are always looking for ways they can reduce cost or do things internally.? AI will allow clients to do more basic work themselves and this will grow over time.

·???????? Legal – As specially trained AI tools evolve; they will provide the legal and regulatory review currently done by in-house counsel.? The need for final human legal oversight may not result in significantly lower legal staff, but it should allow that staff to be much more productive, reducing the need for a junior lawyer or paralegal.? This raises the interesting question of whether we will trust AI legal to accurately review AI generated content?

·???????? CRM – Eventually, AI will be able to continually generate and evaluate ad versions (e.g., A/B tests) and customize output on the fly in real time.? This will not necessarily change staffing, but it may allow 1:1 marketing costs to become competitive with mass channels and thus see the shift from mass to addressable channels accelerate even faster than it is already.

So what should agencies do today?

Establish an AI steering committee or Center of Excellence - You need a cross-functional team to guide the integration and evolution of AI throughout your agency.? This will be tasked with staying up to date on the latest AI advancements and how they are used.? The following recommendations should originate from the steering committee.

·???????? You need an AI strategy – Agencies need a strategy for where to focus their AI efforts - what roles to prioritize, how will it be implemented and by whom, how will it affect current operating processes and roles, how it will be funded and how it will be evaluated (KPIs).?

·???????? Re-evaluate your current processes – Using AI will require you to do some things differently, it cannot just be plugged into your current processes.? Start with a blank slate and use AI as an excuse to realign your processes to be more efficient…and plan to revisit and revise them at least twice per year.

·???????? Re-align the organization – While AI may allow agencies to eliminate a few junior roles, it will provide some new skills for some roles.? As with process, this is a good opportunity to re-look at your org structure and the roles within it to optimize around AI and how the organization will need to evolve as AI capabilities become more robust.?

·???????? Training is essential – The biggest mistake we hear regarding AI is that companies sign up for a tool, make it available to their staff and assume they will do something with it.? The steering committee needs to make training available, explain how they expect these tools are to be used and focus on those functions for which AI offers the highest near-term potential.

·???????? Start small – To make AI work, start with a small group of people that have an interest in using AI and have them start playing with it.? As they develop successes, build on those, and gradually expand into and across the organization opportunistically.

·???????? AI as a competitive differentiator – AI tools are all trained.? An off-the-shelf AI product, like ChatGPT can be taught to use your own approaches and rules so that over time, it will be a truly unique product based on what you taught it and what it learned from working with you.? Overtime, that could become a competitive differentiator leveraging your best practices and experience in a given field or industry vertical.?

Pricing - Agencies need to reconsider how to price their services.? Agencies historically bill based on hours worked.? AI threatens to allow agencies to do the same or more work with fewer staff hours, and the hours being reduced (junior staff) tend to bill out at the highest margins.? Thus, the efficiency promise of AI threatens the revenue and profit of agencies unless they can find a way to switch to a fixed price or retainer model (like consulting firms) not based on an hours-based formula so they can get rewarded for greater efficiency.? I realize, easier said than done…

Questions?? Drop me a line, happy to help!