Episode 5: Mapping the Prompt Gap

Mike Downer: Hey everybody, it’s Mike Downer again, your chief storytelling strategist with Jar Consulting. I’m here with Kevin, the president and CEO of Jar Consulting Group. How are we doing today, Kevin?

Kevin Wosmansky: Doing great, Mike. Welcome back. Good to the camera, my man.

Mike Downer: It’s been a long time. Last time, we talked about how to close the gap and win brand citations in 2026. We’re going to talk a little bit more about that today. So, Kevin, let’s expand on what we were talking about last time. Could you tell me the difference between what we were discussing with keywords and what the prompt gap actually is, and how to map it?

Kevin Wosmansky: Yeah, so last episode, we really dug into the death of the two-word keyword search, right? You’ve got to understand that people are evolving, and they’re starting to have conversations with their AI. That’s such a different strategy and mindset from today’s world, because the only thing we’ve ever really known is searching a few keywords and sifting through information.

Mike Downer: Yeah, that’s what we do, right?

Kevin Wosmansky: Right. So when we talk about that, we have to talk about what this prompt gap is. Traditionally, you search keywords using SEO search. With AI, we prompt AI. Okay? This is part of that transition away from where we were to how things work now.

When you talk about a prompt gap, it’s a fundamental difference between a search query, which is literally about matching, and an AI prompt, which is all about contextual reasoning. For instance, with Google behavior, users typically type two, three, or four words, and they expect to get a treasure trove of documents that they can scan themselves. That’s normal. That’s what we’re used to.

But when you think about what’s happening with AI behavior, and as we’re learning and slowly changing—some people much quicker than others—take Claude or ChatGPT, for instance. Users write prompts that average 23-plus words. That includes constraints, personal context, nuance, and specific goals.

So this is when we start talking about the gap. The gap is what we’re used to doing with keyword search versus now being able to just share everything with AI.

When people talk about the strategic gap, traditional page-one rankings on Google now overlap with AI mentions. Right now, we’re in this in-between phase where people are still Googling things, but AI is creeping into it. AI is mentioning things, or search engines are starting to ingest AI into some of these searches.

Google has Google AI Mode. It’s amazing how many people I talk to every day who don’t realize Google has already changed. When I say, “Google already changed, you just don’t know it yet,” they ask, “What do you mean?” I tell them to pull up Google Search. Right in the search bar, it says AI Mode.

So what’s happening right now is this: there’s this gap where people are searching keywords, looking for something, and Google, with AI Mode, is ingesting AI into it. It’s a way to bridge people over. Really, the prompt gap is just humans learning how to prompt AI to get the exact answers they’re looking for.

When people talk about needing to bridge that gap, it’s really more about optimizing large language model seeding—LLM seeding. It’s about creating context that answers very narrowly scoped, complex technical questions that a generic search engine would typically overlook, but AI will find highly relevant for its synthesis.

That’s all a fancy way of saying that people are starting to evolve away from two- or three-keyword searches, and now they’re asking really complex questions. Most of my clients don’t even realize they’re doing it. The more questions they ask, or the better prompt they give their AI, the better AI synthesizes the data and gives them the answer in a matter of seconds.

The Tire Purchase Example

Kevin Wosmansky: Here’s the example I want to give. Just this last weekend, I was looking to buy new tires for my wife’s vehicle. I stopped in at Costco, a national brand, and since it was the weekend, I asked them what they had for tires. They said, “Hey, we have these Michelin Defender tires,” and explained why they recommended them. Okay, great.

Then I went home. That night, I jumped on Discount Tire’s website, another national brand. I’ve bought tires from them before, and I like that they have a lot of selection. I found a different set of tires that I liked—Coopers, I think they were. They had their little ratings and stars and all this different stuff, and I thought, “Oh, I’m really interested in this. I think I’m going to call Discount Tire and buy these tires.”

But here’s the crazy thing, Mike. The next day, I thought, “Why don’t I ask Gemini? Let me have Gemini compare this.” So I said, “Hey, Gemini, I need to buy new tires for my wife’s vehicle. Here’s the type of vehicle I have. Here’s how we drive it. It’s mostly highway, not off-road, whatever. I was looking at these tires at Costco—here’s the price. I was looking at these tires at Discount Tire—here’s the price. Can you tell me which is the best set of tires for what we’re looking to do?”

And here’s what’s crazy: I had already made the decision that I was going to go to Discount Tire to buy those tires.

Mike Downer: Right, right.

Kevin Wosmansky: Gemini sold me on the reasons why the Michelins at Costco were the overall better value. They were a little bit more money, but it answered the question and said, “These are the top three tires. This is the best one for your money.” It blew me away.

I’m sitting here thinking: AI just sold me on doing something against what I would typically do—go search websites, gather the information, and make the decision myself. So needless to say, Michelin tires got purchased.

Mike Downer: Yeah, and I think you just answered a big question about why this is important in today’s business climate: to get on top of this and not fall behind, because Michelin just won a huge war in your household.

Kevin Wosmansky: Well, you know what? They were all three great tires. This is what’s important. They were all three great tires, right? Based on what I told Gemini, our primary purpose was highway travel. We wanted them to perform really, really well in snow and rain. So it said, for those reasons, this is the one you should go with.

It also said, “This is a great tire, and this is a great tire, but specifically for what you’re looking for, this is the one.” Mike, it did all the research for me. It actually made me realize something: I went and researched those other tires afterward, and I found out that I was wrong.

So I think that’s the power of AI. As humans—myself included—and even though I’m pretty heavily involved in this, helping businesses implement AI, helping businesses with SEO and GEO, and doing all of this stuff for clients, when you start to experience it yourself, you’re kind of like, “Wow, this is pretty amazing.”

Mike Downer: Very, very cool, Kevin. You answered a lot more questions about how to close that gap. I think we’re going to delve a little deeper into this in the very near future. So for today, thanks for giving us a deeper explanation and for sharing your story with us.

Kevin Wosmansky: Have a good one, Mike. Take care, buddy.

Mike Downer: Thanks a lot, Kevin. We’ll talk to you soon.

Kevin Wosmansky: Alrighty. Bye.