The Intent Gap
The billion-dollar difference between what customers type and what they mean
There's a moment that happens in every good electronics store. You walk in, maybe a little overwhelmed, and say something like "I need a laptop for work." A good salesperson doesn't immediately point you to the laptop aisle. They pause. They ask what kind of work. They notice you're carrying a nice bag and wearing business casual. They ask if you travel much. Within two minutes, they understand something crucial: you don't just need a laptop. You need a lightweight machine that won't embarrass you in client meetings, with enough battery to survive a cross-country flight, and you're probably willing to spend more than you initially said if it means getting something that lasts.
Now try that same query on any e-commerce site. "Laptop for work." You'll get thousands of results. Gaming laptops. Chromebooks. Refurbished machines from 2019. The site has no idea if you're a software developer who needs 64GB of RAM or a real estate agent who mostly uses a browser. It treats your words as a search string, not as the beginning of a conversation.
This is the Intent Gap—the space between what customers say and what they actually mean. And closing it might be the biggest unsolved problem in online retail.
The Salesperson in Your Head
Think about what that salesperson is actually doing. They're running a sophisticated mental model of you as a person. Every piece of information updates their understanding:
When you said "for work," they filed away: professional use case, probably needs reliability.
When you hesitated at a price, they noted: budget-conscious, but not cheap—looking for value.
When you mentioned your old laptop was "fine but heavy," they heard: portability matters more than raw power.
They're not just listening to your words. They're building a theory of your mind—what you know, what you value, what you're worried about, what you're not telling them.
This is something humans do automatically. We can't help it. We're social creatures who evolved to understand each other's intentions, and we bring that skill to every interaction. A good salesperson just does it more consciously and applies it to helping you find what you need.
Websites have never been able to do this. Not because the people building them don't care, but because the technology simply wasn't there. Search engines are pattern-matching machines. They find products that contain your words. They rank by popularity, reviews, profit margin. What they can't do is understand that your words are just the surface of something deeper.
The Many Meanings of "Under $2000"
Let's get specific about how this plays out.
When someone says "I need a laptop under $2000," that sentence contains almost no actual information about what they should buy. Consider the range of people who might type those exact words:
Person A
Has exactly $2000 in their checking account earmarked for this purchase. Going over is not an option. For them, "$2000" is a hard ceiling, and they should probably be looking at $1500 machines to leave room for a case and accessories.
Person B
Just did some research and saw that good laptops seem to cost around $2000. They're not actually price-sensitive—they're just anchoring to what seems normal. If you showed them a $2200 laptop that was meaningfully better, they'd probably buy it.
Person C
Types "under $2000" but actually means "I don't want to look stupid by paying too much." Their fear isn't spending money; it's making a bad decision. They need reassurance more than discounts.
Person D
Is comparison shopping. They already know which laptop they want. They're just checking if your site has a better price than Amazon. They don't want recommendations at all.
Same words. Four completely different needs. And current e-commerce treats them all identically: here are laptops under $2000, sorted by best-selling.
"Good for Travel" and the Specificity Problem
Here's another example that shows how badly current systems fail.
"Good for travel" seems straightforward until you think about it for thirty seconds. Travel means different things to different people:
A consultant who flies weekly needs a laptop that's light, has phenomenal battery life, and is easy to pull out at TSA. Screen size matters less because they'll use external monitors at hotels.
A digital nomad working from hostels needs durability, a good keyboard for long typing sessions, and the ability to handle video calls in unpredictable wifi conditions.
A photographer traveling to shoot needs color-accurate display, fast storage for large files, and maybe an SD card slot.
A parent traveling with kids needs something that can handle Netflix downloads, survive being dropped, and maybe shouldn't cost too much in case it gets juice spilled on it.
Each of these people might search "laptop good for travel." They'll all get the same results. And those results will be sorted by some algorithm that has no idea what "travel" means to any of them.
This isn't a failure of any particular website. It's a fundamental limitation of how search has worked for thirty years. The technology could only match words to words. It couldn't understand that "travel" is a proxy for a whole lifestyle, a set of constraints, a collection of anxieties and priorities that vary wildly from person to person.
Why the Gap Exists
It's worth understanding why e-commerce ended up this way, because it wasn't inevitable—it was a reasonable response to technological limitations.
When the first online stores launched, the options were simple: browse a catalog or search by keyword. Search engines got better at handling variations (laptop vs. laptops vs. notebook), but they remained fundamentally text-matching systems. If you searched for "blue shirt," you got shirts that someone had tagged as blue.
Filters helped. You could narrow by price, brand, size. But filters assume you know what you want—that you can translate your fuzzy need ("something professional") into specific attributes (brand: not Alienware, color: silver or black, weight: under 3 pounds). Most people can't do this translation. They know what they want when they see it, not how to specify it upfront.
Recommendation systems tried to bridge the gap by learning from your behavior. If you looked at three Dell laptops, maybe you'd like a fourth. But these systems are backward-looking. They know what you've done, not what you're trying to do. They can't distinguish between "I'm researching laptops seriously" and "I clicked on that laptop by accident while scrolling."
The fundamental problem was that understanding intent requires actually understanding language—not matching keywords, but grasping what someone means. That capability simply didn't exist until very recently.
The Real Cost of Misunderstanding
The Intent Gap isn't just an inconvenience. It has real consequences that ripple through the entire shopping experience.
Customers waste time.
The average person researches a major purchase for hours, reading reviews, comparing specs, trying to figure out what matters. Most of that time is spent compensating for the fact that the store can't help them figure out what they need.
People buy the wrong things.
How many products get returned because they didn't fit the customer's actual use case? How many sit unused because the customer didn't know to ask about some critical feature?
Anxiety replaces excitement.
Buying something should feel good. Instead, major purchases feel stressful because customers are never sure they're making the right choice. That hesitation comes from shopping in a system that can't reassure them.
Good products get lost.
Smaller brands with genuinely better products for specific use cases get buried. The system can't connect "this is the perfect laptop for a traveling consultant" to the consultant searching for it.
Trust erodes.
When customers feel like a site doesn't understand them, they stop engaging. They search on one site and buy on another. They rely on Reddit threads instead of product pages. They develop an adversarial relationship with e-commerce—assuming stores are trying to push whatever's most profitable rather than what's best for them.
What Becomes Possible
Imagine shopping that actually worked the way that salesperson worked.
You type "laptop for work, under $2000." Instead of drowning you in results, the site asks one smart question: "What kind of work—mostly documents and video calls, or something that needs more power like coding or video editing?"
You answer. Now the site knows something real. It doesn't just filter—it reranks. The MacBook Air surfaces for the first person. The ThinkPad with 32GB RAM surfaces for the developer.
The site notices you've lingered on a laptop that's $2100. It doesn't hide it—it explains the tradeoff. "This is slightly over your budget, but it has 50% more battery life than similar options. Worth considering if you travel frequently."
You didn't have to learn what specs matter. You didn't have to translate your needs into filters. You just described what you're trying to do, and the system met you there.
This changes the fundamental relationship between customer and store. Instead of an adversarial dynamic where customers try to extract information and stores try to push products, you get something collaborative. The store becomes an ally helping you make a good decision.
And when people trust that a store understands them, they come back. They spend less time comparison shopping. They feel confident in their purchases. They recommend the experience to others.
The gap between what customers say and what they mean has existed as long as e-commerce has existed. We've just accepted it as normal—worked around it with filters and reviews and hours of research.
But it was never an inherent feature of online shopping. It was a limitation of the technology available.
That limitation is ending. And when every store can understand customers the way the best salespeople always have, online shopping will finally feel like it was supposed to: like having an incredibly knowledgeable friend who just wants to help you find the right thing.
We're building the intent layer that makes this work. If you're curious about what's possible, we'd love to hear from you.