How AI Assistants Will Replace Ads with Offers

There are a few commercials from my youth that I will never forget. To this day, they still make me laugh when one of my siblings brings them up or quotes them (“He likes it! Hey, Mikey?!”). I always remember ads as being about awareness and brand loyalty.

However, over the years and with the advent of the internet, ads have become an extension of tech surveillance.

Not to beat a dead horse, but most tech companies and many brands are tracking us everywhere we go. They target us all under the caveat that they want to bring relevance to their spend, ensuring the right person receives the right message at the right time. During the pandemic, people realized that simply talking to their significant other would suddenly cause an ad for whatever they had discussed five minutes ago to appear, which felt creepy.

In academic circles, they actually study the creepiness factor. It's something most people aren't aware of, but it's an extension of data privacy analysis. There are many reasons why things feel creepy, and often it has to do with time and proximity. We really don't want the right ad at the right time in the right place, and it gets even worse when it's evident to us that it's due to surveillance.

Having said all of that, I love offers.

I think most people do.

I don’t believe the difference is merely semantic.

Offers are when you feel that your wants, needs, desires, attitudes, and timing align. It is like when you're just finishing dinner and the waiter or waitress comes by and offers a cup of coffee or espresso. Sometimes you feel like you're being sold, but for the most part, offers are contextually relevant in terms of time and location, in a way that feels like a fair value exchange between you and the person making the offer.

The real opportunity with AI assistants is that we will transition from ads to offers.

Ads will return to focusing on awareness and brand, which I believe is a healthy shift for the market.

Offers will be dynamic, contextual, and, for the most part, under the control of AI assistants that choose to share or request offers on behalf of human beings. This occurs when the human's dialogue with the AI or requested research reaches a point where requesting an offer from a brand's AI is sensible.

(I originally introduced this concept in an AdWeek article in October of 2023.)

From Creepy Ads to Contextual Offers

Allow me to give you an example.

If you go today and start having a conversation with an AI about a new car purchase, the AI can now ask you 10 to 20 deep questions to help it understand what type of vehicle you want to purchase. When you do that, the AI will then go out and crawl websites, review sites, and gather various types of information.

Your AI Assistant will provide you with a comprehensive understanding of how all that information is available. It will build a dossier and recommend five vehicles that make sense, from top to bottom.

Now, the next step is that you, as a human, still need to visit a dealership or arrange a test drive. However, we're aware that as we move towards an increasingly agentic future, where AI can take actions because you have granted it certain agency, this is where you can see the AI scheduling a vehicle from one of these dealerships or car brands to be brought to your home for you to test drive. It could also schedule an appointment for you to get it done next week.

AI Agents Mediate a Fairer Market

Upon test driving, there's also now this opportunity for an offer. Offers should really look very different than ads because offers will take into account not only the context, timing, location, and interest level of you as the consumer, but they can also, on the other side, reflect the business dynamics, current revenue, projected profit, and even discounts or manufacturing discounts that the dealer is living with. Those two incentives typically occur either directly between the seller in an auto dealership and the finance department, but now they can also happen via AI.

The AI can be granted access to the offer side from the dealer to financial pricing curves with an elastic pricing model. If a dealership is having a phenomenal month, they can hold out for a higher potential price.

Alternatively, if they haven’t hit their quota, they can sometimes become even more aggressive. It depends on the dealer, their inventory, upcoming future models, and the demand for their products.

Supply and demand curves operate very differently in an offer-space market, more economically, and with a focus on value meeting in the middle. So to the consumer, once they've done their test drives, they can go back to their AI and say, "Okay, I think I'm ready to buy. I want one of these three models." Once that's done, the AI will go and request offers from each of the dealerships.

Likely, those offers will not be from a human, but they will initially be analyzed and offered via supply and demand-based pricing elasticity curves in the form of data.

The offer's data feed will enable people to change their offer in real-time, based on inventory and other business factors. From there, I can see the offers from the three different dealerships. They can state the MSRP is versus what they're willing to offer, and in some cases, that might be above or below, depending on my interest in the vehicle and their current business environment.

Empowering Everyone by Removing Friction

This is a great opportunity for both the consumer and the business. Not because it takes much to envision, but because it eliminates a lot of the unnecessary bickering and less-than-exciting elements of the sales negotiation process.

It also gives businesses greater control over how they operate. Some people will choose to employ perfectly inelastic pricing strategies, while others will work much more aggressively to build loyalty or move inventory.

Again, there are products, goods, and services that never offer a discount. In fact, you can only buy their content at premium levels, as I've heard about some designer handbags in the market.

However, other businesses are far more adaptable in how they must respond to their own business environment. An offers market is going to spring up very, very quickly around AI agents. AI is remarkably good at identifying, from a supply and demand perspective, where the price is likely to be set for a good or service on that curve.

Humans on both sides will ultimately decide, once an offer is extended, whether they agree to proceed with the transaction.

Anything that happens between an AI assistant, except for very low-priced goods or commodities, is likely to require final authorization from a human on both the consumer and business sides.

However, much of that will essentially be a confirmation, because neither will want to assume the responsibility of damage to their reputation if they fail to follow through on something the two AI systems have agreed to, based on the context of each other's perspective in the interaction.

The Agentic Future is Already Arriving

When will this happen? The reality is that it already is. Everything you're seeing with OpenAI, as well as Perplexity and others, is beginning to offer the ability for consumers to buy directly “in the chat,” which means that commerce is moving to dialogue.

The dialogue in the chat enables people to find the answer. Still, it's also an opportunity for businesses to incorporate far more effective pricing and elasticity curves into their business models. I am very excited about the move from surveillance to empowering both sides with almost no middleman in between.

Prepare Your Business for an Offer-Based World

Start by building your offer data model now. Consider the business factors that influence your pricing. Document the rules surrounding inventory levels, sales quotas, and seasonal demand so that an AI agent can understand your pricing elasticity and generate a dynamic, fair offer on your behalf.

Then shift your content from persuasion to information. In a world mediated by AI agents, your job is not to create a clever ad but to provide the structured, factual data an AI needs to recommend your product. Focus on developing comprehensive, accurate content that positions you as a trusted source an AI can rely on for its users.

Finally, prepare for agentic commerce. This means structuring your product, inventory, and pricing data so it is machine-readable. The businesses that thrive will be those that make it easy for AIs to understand what they sell and under what conditions, allowing them to participate in this new, offer-driven market seamlessly.