Marketing to AI Agents: Get Ready, GPGers!

For most of us, today's AI is experienced either as an application (ChatGPT, Dall-E, etc.) or as an enhancement to other applications (Spotify, Snapchat, etc.). This is about to change - radically. Microsoft, Apple, Google, and others are all investing billions to integrate AI seamlessly into our operating systems as agents instead of as apps. Agents won't just do, they will anticipate your needs and plan on your behalf. How we experience our digital lives will never be the same.

New AI Agents will have a profound impact on shopping, especially on routine grocery trips. Humans will increasingly outsource their planning, research, and buying to their digital assistants. How will marketers influence brand decisions when an AI is at the wheel? How will price promotion and loyalty marketing change when humans are largely "out of the loop?" These are a few of the big questions that we will need to address.

Here is one marketer's point of view...

How (and When) Bots Will Replace Human Shopping

When you break it down, there are (up to) 5 steps involved in buying a product - you identify a need, you find data about possible choices, you process that information to evaluate the value of each option (benefits divided by price), you make a decision, and then you act (often by hitting "add to cart").

With the dawn of eCommerce, machines took over one of the messiest steps - finding relevant information. Today, GenAI has already made processing that information dramatically easier. You no longer have to look at multiple brand websites and product listings - AI can do that for you and give you a thoughtful summary of findings.

By 2030 (possibly earlier), Your AI agent will combine information about your preferences, your calendar, your relationships, and more to anticipate your needs, make decisions, and execute the buy if you let it. Most people will eventually put their trust in automation.

Influencing Share of Brand Choice

How well is your brand positioned for this future? Here's a little experiment you can try to find out. Open up ChatGPT and have it help you with a category-driven task like a recipe or the selection of a cold medicine. Tell it that you want it to recommend specific brands and be specific about your personal preferences (gluten-free, organic, value-priced, etc.).

Here's what ChatGPT recommended to me for a tuna casserole (one of my favorite comfort meals):

How do you get on the AI's recommended list like Campbell's managed to do consistently in my experiments? This is category-dependent but in the case of food, being featured in other recipe content, having strong reviews, and having widespread distribution that's visible to the AI (eRetailers) are all important. In categories that are more performance than taste-driven, provable benefits and ingredient communication will play a bigger role.

AI will be influenced by the consistency of Internet information. If you want your brand to be recognized as the strongest coffee, having this represented on your website, retail websites, and in trusted 3rd party content will help to convince the AI that this is true.

Finally, each person's AI agent will also learn about what they like and don't like. This data will be analyzed to inform future decisions. Over time, for example, your AI will know precisely what level of "spiciness" you prefer and will shop accordingly.

What to Expect: Pricing, Promotion, and Distribution

When humans are out of the loop, shopping transactions will happen very differently...

• Bidding for Baskets: When shopping becomes autonomous, the most logical interaction model between Shopper and Retailer AI agents will take the form of a reverse auction. The Shopper Agent might say "Hey local retailers - I need these 37 items delivered to 1200 Sandy Pines Road." Retailer AIs, assuming sufficient in-stock, will respond with custom pricing - possibly multiple prices reflecting different times, delivery modes, etc., in a way that optimizes their profitability. Too much milk in the warehouse today? Shopper Agents will get a better deal on that item if they buy now. The Shopper AI will choose to buy from the retailer offering the best price and conditions at that specific time.
• Deal Clearinghouses: The AI-assisted shopper may never see an ad or the inside of a store but brand marketers will need a way to incentivize their trial and their loyalty. Like the coupon clearinghouses of old, the "Deal Clearinghouse," will represent brands offering specific and conditional incentives through Retailer Agents to Shopper Agents. Heinz Ketchup will, for example, be able to offset item (and therefore basket) price at retail as they bring new flavors to market. Offers will be infinitely targetable using geo-demographic, behavioral, and preference data about the consumer. In all likelihood, leading brands will have to rely on AI to design their complex micro-targeted promotional strategies.
• Efficiency Retail: The retailers who succeed in this new world will have the most efficient and reliable supply chains and delivery options. Shopper experience, which has been a prime success driver throughout recent history, will be demoted at least for a while. Tomorrow's top retailers are already getting ready for this by investing in autonomous delivery and footprint optimization (I see you, Walmart!).

Addressing Security and Privacy Issues

For all of this to work, consumers and regulators will need assurance that their security, privacy, and funds are all properly protected. There are a few procedural and technological solutions that will help.

• Transaction approvals: In the near term, anything that involves money or a legal commitment will require the AI to get human approval before spending money or entering into any legal agreement ("I've got this week's groceries ready for Thursday delivery and saved $12.29 versus our weekly average - total is $148.15. May I proceed?"). As the reliability and trust in these systems increases, humans will be able to selectively over-ride the approval process if they so choose.
• Private shopper profiles: Deal clearinghouses will interact directly with Shopper Agents but will keep sensitive consumer data from retailers and brands. The trusted clearinghouse, acting as intermediary, will match the shopper's profile to relevant offers and release brand funds to the retailer. Neither the brand nor retailer will know how a specific shopper qualified for those funds.
• Local processing: As consumer devices get more powerful and AI gets more efficient, processing of sensitive consumer data will happen on your own device instead of somewhere in the cloud. This is already happening - later in 2024, Apple will release "Apple Intelligence" which will move the AI onto your phone so that you can use Siri even for advanced tasks even without a network connection.

Preparing for the Future

Brand marketers need to start preparing now to thrive tomorrow. Here are some recommendations:

• Experiment, experiment, experiment: Instead of shopping the way you used to, shop the way you will tomorrow - even if it's more work. Personal experience with the latest and soon-to-be-greatest ways to buy everything and anything will help you to adapt your professional approach. Invest in the latest and greatest AI toys for yourself and for your family - it's just job security.
• Focus on data quality and content distribution: We don't know exactly how to influence the bots of the future but a safe bet is that clear, reliable, and consistent information distributed broadly will make all the difference to AIs. This is the new SEO.
• Invest in training and capability: Things are going to happen quickly - very quickly. Make sure your organization is staying up to speed on changing AI, shopper technology, and retail last-mile-delivery options. It's much better to be too early than too late on this.

The future of shopping is fast approaching, and it's going to be a fun ride. Embrace it with curiosity, creativity, and the skills you've already built as a retail expert. The rise of AI shopping is a "once in a lifetime" opportunity to dramatically change your brand's fortunes. Make it count!