Are AI AORs next? It’s not long before we start to see companies hiring AI AORs, claiming to the latest and greatest with AI. But is this a good thing, do we need them and what exactly would these agencies do?
Agencies and AI
As more articles get posted on AdWeek and Digiday, and SXSW presentations are overloaded with AI, eventually marketers are going to ask what they should start doing with AI. Some have caught on to this trend by introducing either new AI specific offerings, voice platforms like “Marcel” or integrating AI into digital transformation.
AI offerings are limited
But, up until this point, most of the agency AI/ML launches and offerings have been tied to voice, chatbots, or some combination of the two. According to a Digiday profile on the Wunderman AI division, “Wunderman’s AI division focuses on using AI to create text-based or voice-activated chatbots and virtual assistants.” While the article does mention it is investing in predictive analytics as well, the author does not cite any other AI/ML case studies Wunderman has taken on.
Another voice example is the announcement from JP Morgan that they were hiring a “Voice” AOR. While the announcement was met with some profanities about the legitimacy of this idea, most companies are still in the exploratory phase of how AI can impact marketing, with voice being at the center.
Why are agencies narrowly focused?
Considering that AI/ML has made strides in so many other areas, it’s puzzling the focus has only been only on voice and chatbots. The likely explanation for this is chatbots/voice skills are the low AI hanging fruit. It’s very easy to just deploy a solution using a number of cloud providers and you don’t need your own data to get started. You may just need a couple of APIs and then you are off to the races.
Now comes the hard part
Unfortunately, the next round of AI implementation may feel like it’s a climb to the top of a redwood. After marketers have exhausted off the shelf APIs, the next step is building custom solutions which is going to require sophisticated talent, significant investment in computing power and most of all data organization. For example, in that same Digiday article, the price Wunderman charges for building a AWS/Microsoft/Google enabled chatbot is $150,000 but creating proprietary solutions can cost millions of dollars.
RFP for an AI-AOR?
Let’s say you have already implemented off the shelf solutions and are looking to take the next step. Given these challenges I have presented, you decide to hire an AI AOR to take your marketing to the next level. The next question becomes:
What exactly would these AI AORs do?
Ideally, the first thing an AI agency should do is figure out what type of data you have and where it lives. The amount of data you have will determine the scope of how much AI will be deployed. A follow up question is if you don’t have the data, should you go out and acquire data to build custom solutions or settle on using off the shelf solutions? If data lives in different silos, how easy is it to combine this data and get everyone behind the initiative?
Then, let’s start doing AI
The AI AOR would then need to determine which business challenges AI could solve or how AI could increase customer engagement. The AI AOR shouldn’t just be doing AI for the sake of AI. Ideally, the AI AOR would have a range of tools and skills to solve these problems and should be able to do things like build a programmatic bid optimizer, recommendation engine or an ad tagging system using image recognition.
But, isn’t AI part of digital transformation?
One of the things I struggle with is trying to figure out where in a marketing organization AI should live. Originally, I thought it would live in the analytics department as they were the ones that could write code to actually implement AI/ML. However, while the analytics department certainly uses these methods, AI/ML should really be at the center of any digital transformation strategy. AI/ML can improve efficiencies and help make better decisions across so many lines of business, that it should be at the core of digital transformation. It’s no surprise that the digital native companies (Netflix, Google, Facebook) are all leaders in the AI space.
At the end of the day
There is a niche role for AI AORs to play to build custom solutions or bring creative ideas on how to apply AI to marketing. These agencies should bring a thorough understanding of the space and not just implement off the shelf solutions. The biggest challenge for AI AORs will be the as the lines blur between tech and advertising companies, I suspect the Google/Amazon/IBMs of the world will start creating custom tools for its clients.
More on that next week…