.avif)
Dear Diary:
OMG. So today has been the WORST and also the best day of my life!
I went to the Metrocentre with mum after school and TAMMY GIRL had the Bag Grl vest in literally every colour but not my size. Why does this always happen to me?! I have wanted this since I saw Kate wearing hers and if I don’t get it before the disco on Friday I’m actually going to cry.
So I’m stood there, trying to hide my deep sadness, and the girl who works there (she’s lush. I think I want to work here too!) says “don’t worry pet, I’ll take your name”. She gets a Post-It and writes “Bag Grl vets - size 8. Joanna” and my mum’s landline on it (I wish she would get me a mobile, everyone else has one) and she tells us that she’ll call when more come in.
But guess what?! She rang on Tuesday night. We were watching The Fresh Prince and mum shouted telling me my top was on hold and ready to collect! I have never felt so important in all my life!
We went back on Thursday and I wore it for the disco, everyone loved it! I felt amazing.
Ok - I’m not 15 anymore. But I probably would have written something like this in my diary (spelling errors included).
Let’s look at this now I’m grown, and capable of more than one feeling at a time.
What actually happened there
I had never before felt as understood by a retailer as I did getting that phone call. Not by an algorithm that I seem to be stuck in the same loop with now. Not by an ad that knew I’d been looking at diamanté vests (it’s research for this article, I promise). But by a girl with a note and a pen and my mum’s landline number.
Technically none of that was sophisticated. There was no CRM record, no automated workflow, and certainly no dashboard. There was just a member of staff who knew what to do with the information I’d given her. The process wasn’t impressive, but the recognition was.
That Post-It note was declared intent. I told a real person exactly what I wanted, in exactly what size, and she wrote it down and acted on it. There was no guessing. No-one had to infer anything from how long I’d lingered by the t-shirt rack.
Inferred behaviour tells you what someone did. Declared intent tells you what they want. Everything else in retail data is really just an attempt to close the gap between the two.
The problem retail’s back in the middle of
Which is, weirdly, the exact problem retail’s back in the middle of in 2026, just with better technology and much worse handwriting.
Retail has never had more data, and somehow it’s getting harder to know what any of it means.
In January (this year, not 2006), at NRF, Google’s CEO Sundar Pichai stood up and announced the Universal Commerce Protocol (UCP), built with Shopify, Target, Wayfair, Etsy, and Walmart and backed by more than 20 other companies besides. The idea is that an AI agent can handle discovery, checkout, and what happens after buying inside a single conversation. Google is starting with checkout: a buy button appearing directly inside Google Search’s AI Mode and Gemini.
Which means some unknown slice of the “customers” browsing your site this afternoon aren’t shopping the way a customer used to. They’re agents, and an agent could be doing any almost anything one someone’s behalf - researching, comparing, checking whether something’s in stock, gathering information for later, acting on a vague instruction, or actually buying. The retailer sees activity, but the activity doesn’t tell you how close that person actually is to a purchase, or whether there’s one coming at all.
Put simply, the behaviour you’re measuring is getting much harder to attribute to actual intent. Whether the click came from a person or an agent running an errand for one, it looks identical on a dashboard.
The same clicks are happening, but the intent behind them is wildly different, and there’s no label anywhere to tell you which is which.
We know there’s no shortage of data. What’s disappearing is the confidence that any of it means what it used to.
What still tells the truth
One thing doesn’t need decoding, whatever traffic looks like: a customer sitting down with a real person and saying “I want this, in this size, for this event, and I need it by Thursday.” Joining a wait list and turning up. Booking a bra fitting, a styling session, or a made-to-measure appointment with their name on it, for a specific reason, at a specific time.
AI agents will get better at connecting digital discovery to physical retail: booking appointments, checking stock, arranging services on someone’s behalf. These don’t change what the booking itself says. An agent might one day make it. It’ll say “this person wants a fitting at 2pm on Saturday,” and that’s true no matter what pressed the button.
If you run a personal shopping suite, a bra-fitting service, or a made-to-measure operation, none of this is new information. You’ve been sitting on this signal for years, long before anyone had a word for it or worried about an AI agent turning up in the analytics. Luxury, beauty, and made-to-measure retail have always run on a real conversation with a real person. The rest of the industry is only now catching up to what that’s worth.
The booking is valuable because of what it tells you: who, what, when, and why, confirmed by the customer rather than guessed by an algorithm. Declared, not inferred. Which is why the humble appointment is becoming more interesting, not less: in a retail world increasingly full of behaviour nobody can quite attribute, a booking is still one of the few moments where a customer simply tells you what they’re there for.
Some things haven’t changed
In 2006, a (very cool) teenager told a sales assistant exactly what she wanted. 20 years on, retailers have vastly more data and, in plenty of cases, a lot less certainty about what any of it means.
Sometimes the most valuable data a retailer can have is still the thing a customer explicitly tells you. It just used to come from a Post-it note. Now it comes with a booking reference.
Quick answers
What is declared intent in retail? Declared intent is when a customer tells a retailer directly what they want (a specific product, size, appointment, or request) rather than the retailer inferring it from browsing behaviour like clicks, dwell time, or footfall.
What is first-party intent data in retail? First-party intent data is information a retailer collects directly from a customer, with their knowledge, through an owned channel (bookings, waitlist sign-ups, in-store conversations, or account activity) rather than data bought or inferred from third-party tracking.
How can retailers capture customer intent in-store? The clearest in-store methods are ones where a customer states a specific want to a real person: appointment bookings, waitlists, styling or fitting requests, and staff-logged preferences, all captured against the customer’s name so the retailer can act on it later.
Why is inferred customer data becoming less reliable? A growing share of website traffic and browsing behaviour now comes from AI shopping agents acting on a person’s behalf, not from the person directly. That behaviour can look identical to genuine customer interest, making it harder to tell how close any single session is to an actual purchase.
