Intent marketing existed before AI
eCommerce has been chasing intent signals for years.
Search queries that show what people want. Cart abandonment signals hesitation. Previous purchase history that indicates preferences. Payday campaigns timed to moments of readiness.
We have always known intent matters.
We just never built our sites to respond to it properly.
The intent signals we already have
Right now, your site receives visitors with wildly different levels of purchase intent.
Someone clicks a display ad while reading an article (low intent). Someone searching "best wireless headphones under £100" (medium intent). Someone returning after abandoning a cart with three items (high intent). Someone arriving from a ChatGPT recommendation after ten minutes of research (very high intent).
And what does your site show all of them?
The same homepage. The same offer. The same experience. This has never made sense.
It just became impossible to ignore once AI traffic started arriving.
A brief history of intent recognition
Early 2000s: Search
Google proved that intent could be understood through queries.
When someone types "buy red Nike trainers size 9," you know exactly what they want. eCommerce sites adapted. Product pages optimized for search terms. Paid ads matched to queries.
This was intent recognition. We just called it search marketing.
Mid 2000s: Email segmentation
Email platforms allowed sending different messages to different segments.
Previous buyers got different offers than first-time visitors. People who browsed but never purchased got abandoned browse emails.
This was intent recognition. We called it lifecycle marketing.
Early 2010s: Retargeting
Display ads started following people around the internet based on browsing behavior.
Someone who viewed hiking boots saw ads for hiking boots. Someone who abandoned a cart saw ads featuring that specific cart.
This was intent recognition. We called it behavioral advertising.
Mid 2010s: Abandoned cart recovery
Platforms like Klaviyo made it easy to trigger emails when someone left items behind.
These emails worked because they responded to clear intent. Someone nearly bought. Something stopped them. The email addressed that hesitation.
This was intent recognition. We called it cart recovery.
Late 2010s: Exit-intent technology
Pop-ups that triggered when someone moved to leave the site.
The logic: if they are about to leave, they need more persuasion.
(This was intent recognition, too. Just often badly executed.)
2020s: Post-purchase upsells
Someone just bought running shoes. Show them socks. Show them running belts. Show them those GPS watches.
This was intent recognition. We called it upselling.
The pattern nobody noticed
All of these things work for the same reason.
They recognize visitor intent and respond accordingly.
Search shows people what they are looking for. Email segmentation sends relevant messages based on behavior. Retargeting reminds people of products they already considered. Cart recovery addresses known hesitation. Post-purchase offers build on demonstrated interest.
But here is what we never did:
We never applied this same logic to the on-site experience for most visitors.
Where intent recognition stopped
eCommerce got good at recognizing intent in specific contexts.
- Search (we can see the query)
- Email (we know the segment)
- Paid ads (we control the targeting)
- Cart abandonment (we know what they nearly bought)
But for everyone else? Generic treatment.
If we could not identify you through one of these specific channels, you got the default experience. The homepage banner everyone sees. Pop-up, everyone gets. Same messaging for all.
It did not matter if you arrived from:
- A blog post about sustainable fashion (indicating values-based purchasing)
- A comparison article listing pros and cons of different brands (indicating research phase)
- A Reddit thread asking for recommendations (indicating active consideration)
- An Instagram post showcasing a specific product (indicating visual appeal)
You all got the same homepage. The same offer. The same popups.
Because we had no system for understanding intent beyond our controlled channels.
Why we stayed stuck
Three reasons eCommerce never solved this properly:
1. Data was messy
Referral URLs contained some information, but not enough to act on confidently.
Landing pages indicated interest but not intent level. Session behavior showed what people did, but not why they did it.
We had signals. Just not clear enough ones to build experiences around.
2. Personalization was hard
Creating different experiences for different visitor types is required:
- Technology to identify the visitor type
- Content variations for each type
- Testing to validate the approach worked
- Ongoing optimization to improve performance
Most retailers focused on simpler wins. Email segmentation. Search optimization. Paid ad targeting.
Things where the intent signal was obvious, and the personalization was straightforward.
3. One-size-fits-all was easier
Running the same experience for everyone is simple.
One homepage. One set of offers. One messaging framework. Treating different visitors differently requires complexity. And complexity requires justification.
(The ROI of intent-based personalization was always there. We just never prioritized building it.)
What AI changed
AI did not invent intent-based marketing.
AI made the cost of ignoring it visible.
When Adobe reports that AI traffic converts 42% better than regular traffic, what they are really saying is that high-intent traffic performs better than low-intent traffic.
The difference now is that AI traffic arrives with intent so obvious it cannot be ignored:
- Someone spent ten minutes discussing options with ChatGPT
- The AI asked clarifying questions about their needs
- They compared multiple brands based on specific criteria
- They were recommended your product for specific reasons
- They clicked through, knowing exactly why
This person has higher intent than almost any other traffic source.
Higher than display ads. Higher than social. Higher than most organic search. Sometimes even higher than branded search. And when they land on your site and see the same generic experience everyone else gets?
The gap between their intent and your response becomes impossible to miss.
Intent signals that existed all along
Here are intent signals eCommerce has had access to for years but rarely acted on:
Time on site before first action
Someone who immediately clicks a product has different intent than someone who browses three category pages first.
Depth of navigation
Viewing one product indicates interest. Viewing six products in the same category indicates comparison shopping.
Referral source context
Someone from "best budget laptops 2026" has different intent than someone from "laptop buying guide for beginners."
Device type
Mobile browsers often have lower intent than desktop browsers. (Though this has been changing.)
Previous visit behavior
Second visit after viewing products indicates reconsideration. Fifth visit without action indicates hesitation.
Page sequence
Someone who goes:
homepage > category > product > checkout
Has different intent than someone who goes:
homepage > about us > returns policy > product.
Search terms used onsite
Someone searching your site for "size guide" has different intent than someone searching for "student discount."
We have had access to all of this.
We just never built systems to respond to it at scale.
What good intent recognition looks like now
The retailers who will win with AI traffic are not doing anything new.
They are applying the same intent-awareness they have used for years in search and email.
Just extending it to all visitors.
When someone arrives from a product comparison article:
Do not show them a generic homepage.
Show them the comparison tools they were already using. Help them finish the evaluation they started.
When someone arrives from a recommendation for a specific use case:
Do not show them your full catalog.
Show them products that match that use case. Reinforce why they were recommended.
When someone lands on a product page directly:
Do not interrupt them with a pop-up asking for their email.
They already know what they want. Remove friction. Make purchasing easy.
When someone returns after abandoning:
Do not show them the same offer as you do to first-time visitors.
They have already seen your site. Address whatever stopped them last time.
This is not revolutionary.
This is just applying basic intent recognition beyond the narrow channels we currently use it in.
The framework that works
Three questions to ask about any visitor:
1. What brought them here?
Referral source. Search query. Ad creative. Social post. Email link.
This tells you the awareness level and initial interest.
2. What are they doing now?
Pages viewed. Time on site. Products considered. Actions taken.
This tells you the current intent and decision stage.
3. What should we do about it?
- Remove friction for high-intent visitors.
- Provide information for mid-intent researchers.
- Build interest for low-intent browsers.
(And stop showing everyone the same generic experience regardless of where they are in this spectrum.)
Why this matters beyond AI
AI traffic is growing fast. But the bigger opportunity is not AI-specific. The bigger opportunity is finally building the intent recognition systems that eCommerce should have had all along.
Because every traffic source has intent signals:
- Organic search shows what people are looking for
- Referral traffic indicates what content resonated
- Social traffic reveals what caught attention
- Direct traffic suggests brand familiarity
- Paid traffic reflects targeting criteria
Right now, most sites collapse all of this into "anonymous visitor" and treat everyone the same.
AI traffic has just made this approach obviously insufficient.
What to do now
Start with the traffic you already have. Pick one high-intent segment. Something clear and measurable.
Maybe it is:
- People arriving from comparison articles
- Visitors who viewed 3+ products in one session
- Return visitors who previously abandoned
- Anyone landing directly on product pages
- Traffic from your highest-converting referral sources
Create a different experience for this segment.
Not a different offer necessarily. Just a different response that acknowledges their intent. Test whether it performs better than your default experience.
Then expand to the next segment.
The advantage is temporary
Right now, most retailers are focused on getting AI mentions. Very few are thinking about what happens when that traffic arrives. This creates an opportunity.
The brands that build proper intent recognition now will capture a disproportionate share of high-quality traffic.
Before competitors realize what is happening. But this advantage will not last forever.
Eventually, everyone will figure out that high-intent traffic needs different treatment. Eventually, intent-based onsite experience will become table stakes.
The question is whether you build it now or wait until eighteen months, when you are playing catch-up.
Intent has always mattered
AI did not invent intent-based marketing. eCommerce has been using intent signals for decades. In search. In email. In retargeting. In cart recovery.
We have always known that treating different visitors differently produces better results. We just never extended this logic beyond our controlled channels. AI traffic has made that gap impossible to ignore.
Because when visitors arrive with documented intent, clear preferences, and specific needs, treating them like anonymous traffic is not just inefficient; it's disrespectful.
It is obviously wrong. The opportunity now is not to learn something new.
It is finally applying what we have always known across all traffic, not just the easy parts.
