Agencies and Consultancies must be ahead of the curve when leveraging AI
Ad Agencies and Consultancies are at the forefront of creativity and strategy. They should be the early adopters, the first movers, the risk-takers.
Some are. And they take up a LOT of the media cycle.
But the reality is that many aren't. Or at least not in a meaningful, consistent, and sustainable way.
Consider this your rallying cry.
Risks of not using AI
“We’re in a golden age of AI innovation, really, for the media industry... The implications are vast, and the impact on creative processes is particularly striking. AI allows for rapid creation and iteration, which perfectly suits the fast-paced nature of today’s media landscape.” - Mike Bregman, chief activation officer at Havas Media Group
While the big agencies (and agency groups) like WPP, Havas, Publicis, and Huge are investing in massive AI insights engines, most US-based professional services companies (which include advertising, design, digital marketing agencies, and consultancies) have not recently used AI - nor are they ready to use AI in the next six months.
More surprisingly, it seems that there is very little overlap in what types of AI are being used by those companies. While 12% have used AI (for business) in that period, only 5% have used natural language processing, and only 4% used marketing automation.
When we think about the risks of not using AI, it comes down to four things:
1. Working in an insights vacuum
While this has been the way things have always been done, there is a growing divide between agencies working from a small sample size of already bought-in customers and ones leveraging powerful predictive analytics and insights engines – and then building on those insights through customer research. AI can be used to identify trends and causal relationships that we wouldn't be able to on our own. It also allows us to remove our biases when going into a business or marketing challenge and be open to what best shifts consumer behavior.
2. More effort to sell in ideas
I think everyone can agree that AI won’t replace creatives and that consumers don’t want everything to become AI models and actors. However… there are some amazing tools coming out that allow creative teams to prototype their ideas to a level that makes it so much easier to excite clients (or potential clients) about what the idea could turn into. Bringing storyboards into 3D through voiceovers with Wondercraft, music with Google Dreamtrack, and video or images with Runway.
3. Over-relying on "best practices"
As humans, we default to assuming novelty where there isn't any. This isn't an issue if you have a large team of people with whom to iterate and expand on ideas. However, most teams don't have that luxury (whether because of headcount or budget). AI tools afford creative teams the luxury of providing a coach and a sounding board, and most importantly, it's able to tell you if an idea is too common, expected, or trite. Best practices are great for usability - not for creative or innovation projects.
4. You can't build for what you don't understand
Many of the AI implementations that we see today probably won't last. The infrastructure of AI will though. This is the new baseline of what will be possible, and the implementations will only get more useful and more powerful in the near future. As with any emerging technology, you need to understand how it works and how it can be used before you can even start understanding how it can be leveraged for customer benefit or to grow your business. This means starting. Just start. Even getting things wrong gets you closer to getting it right. The big guys are building their own AI systems, and that's great for them. But there are many off-the-shelf AI tools that you can start with to build your own system or identify where you can give your customers (and employees) that biggest value.
Risk of doing AI badly
Let's start with this: If you begin with safeguarded instances of integrating AI into your agency's business and avoid using PII or NDA-covered content, then the risk is very low. This allows your team to understand the technology and have confidence in what they're doing before you level up to more "risky" applications.
I'm not a security specialist, so I'm not going to dive deeply into that realm, but here are the risks that I most commonly speak with clients and peers about:
1. Risk of antagonizing customers
… if you rush a customer-facing AI tool without proper consideration or training. We're still in the early days when customers prefer no AI to bad AI. So take your time. Train on high-quality data. Consider all the ways in which this tool can enhance the customer experience and make your customer feel better cared for.
2. Risk of creating monotone content
… if you over-rely on what the AI output is. It's rarely a silver bullet that outputs something great in one or two iterations. What most AI tools today do well is remove the time-consuming starting or editing point. It's great at finding themes in data or transcripts but not at telling you what to do with those themes or creating moving stories that build business cases. It's great at being a sounding board for creative ideas or iterating on what you've done, but it needs a deeply thoughtful starting point.
3. Risk of changing ownership rights
… if the terms of service change. We haven't yet reached a place where legislation exists around ownership and commercial use licenses for AI-created or enhanced content. Most AI companies include these details in the TOS, but nothing is stopping those companies from changing their TOS. And many of these companies require you to be a current, paid subscriber to maintain that ownership over the output. Treat public-facing AI creative content the way you would any other licensed content to avoid potential issues in the future.
4. Risk of not evolving your process
… if you’re still thinking of data (and the data science team) as an input to your creative process, and not intrinsically linked. With AI, data is part of the experience and is meshed with the user experience. We’re no longer “pulling data” - data is part of what needs to be designed for and with. As strategists and product leaders, we need to understand how data science works just as well as we understand consumer needs.
Agencies should be leading clients into the AI journey
Now that we've talked a bit about what AI should mean to agencies and consultancies themselves... what about their role and responsibility as strategists for their clients?
It's my belief that as agents of emerging trends and consumer behavior, we are also responsible for guiding our clients to the right tools and experiences for their business and their customers.
Agencies and consultancies need to find their way back to being early adopters and first movers, deeply entrenched in the technologies their clients aren't able to take the risk on yet. Then, they need to use their strategic prowess to build the business cases that their clients need to start leveraging innovative solutions grounded in the deep needs of those clients’ customers.
But to be effective coaches, we also need to understand the catch-22 of researching and building with emerging technologies. You won’t know what's effective until you have something for people to try. You have to advocate for iteration, collaboration, and co-creation.
As coaches, we need to constantly ask ourselves: What is the right AI tool for each project? What is the right foundational model that sets your client up to succeed 1... 3... 5 years from now?
You can't use what you don't know. Start now. Start with a purpose in mind.
But start now.
For advice on building emerging technology products:
Brittany Hobbs is a Freelance product insights leader specializing in the adoption of AI at https://ph1.ca and the co-host of the Designof.AI podcast.