How Strong Is Your Product Discovery Stage?
How Strong Is Your Product Discovery Stage?
This article draws on Episode 3 of the DTC Revenue Rocket Podcast, where ECD CEO Zak sat down with Sebastian Brown Glad from Clear.io, which builds search and product discovery tools for Shopify brands. Sebastian’s central argument: the discovery stage — what happens between “I’m looking” and “this might be the one” — is the most underused customer data a brand already owns. Search terms and filter behavior show intent earlier than the cart, earlier than anything the retention stack picks up. And a shopper’s interest is strongest right at that first search, fading a little more each time they can’t find what they came in for.
The post explains what search terms, filter behavior, zero-result reports, and collection page engagement are each telling a brand, and how that data should inform email segmentation, paid landing pages, merchandising decisions, and product page content — not just site search itself. The main takeaway is that discovery is not a website project. The brands converting the traffic they already have are the ones whose email, ads, site, and merchandising all know the same thing about their customer.
- Search terms are the customer’s own language, handed over for free Someone who types “shade tolerant perennials” or “gluten-free gift box” has just described your catalog in the words they would use talking to a friend. Sessions that include site search typically convert at two to three times the rate of those that don’t. That is not a site detail — it is a revenue channel that most brands are not staffing.
- The first filter someone uses is the attribute their entire purchase decision hangs on Apparel shoppers filter by size. Skincare shoppers filter by ingredient or concern. Garden shoppers filter by sun exposure. Whatever people filter by first is what they are actually buying on — not the feature the homepage leads with. If filter labels sound like they came off a spreadsheet instead of a conversation, shoppers are not clicking them.
- Discovery data should reach the entire downstream system, not just the site Browse and search behavior make sharper email segments than opens and clicks ever will. Zero-result searches reveal either a synonym problem or a buying opportunity. High-volume search terms are proven demand in the customer’s own phrasing — the next landing page and ad copy are already written.
The biggest brands in ecommerce spend a good chunk of change figuring out who their customer is. Focus groups. Consumer panels. Surveys. Hiring agencies to build half-baked persona decks founded on some insights and a dream. Then even more money goes out the door, into restructuring, campaigns, and ad creative built on top of it.
And yes, a good customer profile gives you direction. The messaging angle, the tone, a starting point. But it’s broad by design. It tells you who you’re selling to. It can’t tell you what’s happening on your site once they get there.
So how should the site be organized? Which parts of the experience deserve the most attention? Those answers rarely come out of a research deck.
Now, what if we told you you’re already sitting on a gold mine of customer information?
It’s your website. And no, we aren’t suggesting you chase a single shopper around the site in real time.
On Episode 3 of the DTC Revenue Rocket Podcast, our CEO Zak sat down with Sebastian Brown Glad, who heads up partnerships at Clear.io, where they build search and product discovery tools for Shopify brands. For Sebastian, the discovery stage is the #1 underused customer data a brand already owns. Search and filtering show intent earlier than the cart, earlier than anything the retention stack picks up. And it’s the high point, too. A shopper’s interest is strongest right at that first search, and it only fades from there, a little more each time they can’t find the color, the size, or the answer they came in for.
So, one question worth sitting with. How strong is your ecommerce product discovery stage, and what has it been telling you?
Your Shoppers Are Already Showing You What They Want
Product discovery is the stretch of shopping between “I’m looking” and “this might be the one.” It happens in your search bar, your category navigation, collection pages, filters, recommendation widgets, and the reviews someone scans on the way past.
With a handful of products, it barely registers. There’s nothing to narrow down. A shopper takes in everything you sell in a few seconds and makes a call.
Give them three hundred SKUs across a dozen collections, and it’s a completely different trip. Every filter, every sort, every category click is one more small decision, and small decisions add up. None of them look like a problem on their own, which is exactly why this stage gets skipped in most audits.
And every one of those clicks leaves a record.
Search Terms Are the Customer’s Own Language
Someone who types “wide leg linen pants” just handed you their exact vocabulary. “Shade tolerant perennials” tells you which product attribute is leading the decision. “Gluten-free gift box” gives you the occasion and the constraint in four words.
And nobody prompted any of it. These are customers describing your catalog back to you in the words they’d use talking to a friend, by the hundreds, every month, for free.
They also tend to be your best visitors. Pull the conversion rate of sessions that use site search and set it next to sessions that don’t.
Industry benchmarks tend to put searchers at roughly two to three times the conversion rate of everyone else. Your own number is the one that matters, but if it’s anywhere near that, this stage isn’t a site detail. It’s a revenue channel you’re not staffing.
Start by pulling your top 25 search terms and setting them beside your collection names. Where the words don’t match, add the synonym so the term resolves, and borrow their phrasing for the collection title while you’re in there.
Filters Tell You Which Attribute Decides the Purchase
The first filter someone reaches for is the attribute their whole decision hangs on, and it’s rarely the same one twice. Apparel shoppers start with size. Skincare shoppers start with an ingredient or a concern. Garden shoppers start with sun exposure. Food shoppers start with flavor or a dietary restriction.
Which is worth knowing, because whatever people filter by first is what they’re actually buying on. Not the feature you lead with on the homepage. The one they use to rule everything else out.
It also matters whether your filters were built around how people shop or around how your product data happens to be organized. “Shade tolerant” is a filter. “Light requirement: partial” is a database field. One of those gets clicked.
So go read your own filter labels the way a customer would. Anything that sounds like it came off a spreadsheet gets rewritten in the words people actually use.
Discovery Is Also Where Trust Gets Built
Finding the product is only half the job. The shopper still needs a reason to believe it.
That was Sebastian’s other point on the podcast. When the right product surfaces with a strong rating, real review volume, and enough detail to answer the obvious question, confidence starts building early. The shopper isn’t just oriented at that point. They’re reassured, and reassured shoppers keep moving.
Most brands keep every trust marker at the bottom of the product page, though. A hesitant shopper comparing six results never makes it that far.
So look at what your search results and collection cards actually show. Rating and review count, or a photo and a price? Stock status, or a surprise at checkout? One detail that separates six things that look nearly identical at thumbnail size, or nothing at all?
This matters most in the categories where buying takes a little faith. Fit, ingredients, compatibility, price, and anything headed to someone else as a gift. Make a shopper open four product pages just to compare ratings, and they’ll go do that comparison on somebody else’s site. Get rating, review count, and stock onto the card itself, before anyone has to click.
How Discovery Data Should Inform Your Marketing Decisions
One shopper’s search doesn’t change anything. A few thousand of them should change plenty, and not just on the website.
Email and SMS
Browse and search behavior make for sharper segments than opens and clicks ever will. Someone who filtered for a specific size and left is a different person from someone who skimmed the homepage for forty seconds, and the email marketing they get should reflect that.
Start with the segments you could build today. Viewed three or more products in one collection without adding to cart. Used a specific filter and left. Searched the same term across two separate sessions. Each of those deserves a different message, and none of them are “you left something behind,” which is the version most brands send and tells the customer nothing they don’t already know. Name the category they were actually hunting through and show them what else lives in it.
SMS has the easier job here. It just has to be specific. Back-in-stock alerts should fire on the exact variant someone filtered for, not the product in general. That’s what ecommerce personalization looks like in practice.
Paid Media and Paid Search
Watch for landing pages where a visitor’s first action is a site search. That’s a shopper telling you the ad promised something the page didn’t deliver, and fixing it is one of the fastest wins in ecommerce conversion rate optimization because it costs nothing but attention.
The same reports hand you your next set of landing pages. High-volume search terms are proven demand in the customer’s own phrasing, so build pages for them and write the ad copy in the words people are already typing. And if a term drives plenty of onsite searches but almost no purchases, that’s worth knowing before you spend money bidding on it.
Merchandising and Collection Structure
If the same two or three filter combinations get used constantly, those aren’t filters anymore. They’re collections waiting to be built, with their own page, their own copy, and their own spot in the nav.
Sort order deserves a look while you’re in there. Most collection pages are still running someone’s hand-pinned picks from two seasons ago, which means out-of-stock products are sitting up top taking the best real estate you’ve got. Let what converts drive the order, then re-check it when the season turns, because what sold in May isn’t what sells in September.
Your Site and Product Pages
Whatever attribute people filter for most should be visible before anyone scrolls. If “fragrance free” is your second most-used filter, it isn’t a spec to bury in an accordion; it’s the reason someone buys. Sizing detail, material, delivery timing. Same idea.
Use their words while you’re at it. If shoppers filter for “petite” and your product page says “shorter inseam,” you’re making them do translation work. Your website design and development should follow the questions shoppers keep asking, not internal preference.
4 Discovery Behaviors Worth Checking First
Four things worth pulling up this week, what each one is telling you, and the fix it points to.
Searches That Come Back Empty
The zero-result report is the most useful thing in your store.
Sort it by volume and every term lands in one of three buckets. You carry it under a different name, which is a synonym fix. You don’t carry it at all, which is a buying conversation and worth bringing to whoever owns inventory. Or you used to carry it, and the page is gone, in which case the term should redirect to the closest live product instead of a dead end.
Then fix the empty page itself, because “no results found” is the least helpful sentence in ecommerce. Show nearest matches, the same item in another color, or a notify-me capture. While you’re in there, confirm search is actually visible on mobile instead of hidden behind an icon that takes two taps, and that autocomplete is suggesting your best-converting terms rather than whatever matches alphabetically.
The Last Filter Someone Uses Before They Leave
The first filter tells you what matters to a shopper. The last one before they exit tells you where the experience stopped helping them.
Usually it’s a combination that returned nothing. Size 12 in navy, no results, back button. That person was as close to buying as anyone gets, and the site told them no without offering a single alternative.
Fix the dead combinations first, the same way you’d fix an empty search. Show the nearest available match or the same piece in a different color. Then audit which filters get used at all. The ones nobody touches are just noise slowing down the ones that matter, and cutting them usually does more than adding another.
Products Viewed Again and Again With No Add to Cart
Someone who’s opened the same product four times over two weeks isn’t undecided about whether they like it. They have a specific unanswered question, and it’s almost always fit, price, or timing.
Same goes for repeat searches. A shopper typing the same term across three separate days is showing real intent and getting no help for it.
The answer to what they’re stuck on is usually already sitting in your support tickets and your reviews, where the same question keeps coming up. Find it, answer it on the product page, and then give that group its own follow-up rather than dropping them onto the general promo list with everyone else.
Collection Pages People Enter and Never Click Into
A collection page with strong entrances and weak product clicks is a page that isn’t helping anyone choose.
Usually the cause is one of four things: too many products before a decision is even possible, a default sort nobody would pick, cards that don’t differentiate one item from the next, or out-of-stock products sitting up top.
Pull up your five highest-traffic collections on a phone and look at what’s above the fold. Can a shopper tell these products apart? Are the filters visible, or buried under a button? If it’s a paid landing page, the stakes go up, because you paid for that visit and the page spent it.
None of this is a one-time cleanup, either. Put the search and filter reports on a monthly cadence with a pass before every seasonal push, because the language shifts with the season and so does the demand underneath it.
How ECD Turns Discovery Behavior Into Revenue
Discovery isn’t a website project. What those reports tell you only becomes revenue if it reaches everything downstream: the segments your emails are built on, where your ad spend lands, how collections get structured, what gets ordered for next season. Fix the search bar on its own and you can still lose the sale, because the ads are putting people in the wrong room and the emails are still guessing.
As a Klaviyo Master Elite Partner, one of roughly 35 agencies out of more than 22,000 worldwide, and Klaviyo’s Agency Partner of the Year for the Americas, ECD has audited hundreds of ecommerce stores. The patterns repeat. The brands converting the traffic they already have are rarely the ones with the biggest budgets. They’re the ones whose email, ads, site, and merchandising all know the same thing about their customer.
We’ve helped brands lift conversion by 40 percent after Shopify and CRO work, and drive 49 percent of total revenue through email and SMS built around what customers actually do.
If cart abandonment shows you where a shopper stopped, ecommerce product discovery shows you where they first needed help.
The distance between browsing and buying is usually shorter than it looks. You’re already sitting on the gold mine.Before your next campaign goes out, find out what it’s been telling you. We’ll help you mine it.
Get Your Free Revenue ForecastFrequently Asked Questions
What is ecommerce product discovery and why does it matter?
Product discovery is the stretch of shopping between “I’m looking” and “this might be the one.” It happens in the search bar, category navigation, collection pages, filters, recommendation widgets, and the reviews someone scans on the way past. With a small catalog it barely registers — a shopper takes in everything in a few seconds. With hundreds of SKUs across a dozen collections, every filter click, sort choice, and category navigation is a small decision, and small decisions add up. The discovery stage is also where shopper intent is at its peak: interest is strongest right at the first search and fades a little more each time someone cannot find the color, size, or answer they came in for.
How should ecommerce brands use site search data to improve their stores?
Start by pulling the top 25 search terms and setting them beside your collection names. Where the words do not match, add the synonym so the term resolves, and borrow the customer’s phrasing for the collection title while you are in there. Then pull the zero-result report sorted by volume — every term lands in one of three buckets: you carry it under a different name (a synonym fix), you do not carry it at all (a buying conversation worth having with whoever owns inventory), or you used to carry it and the page is gone (redirect the term to the closest live product). The zero-result page itself deserves attention too — “no results found” is the least helpful sentence in ecommerce, and showing nearest matches or a notify-me capture recovers shoppers who were otherwise going to leave.
What does filter behavior tell a brand about how customers are actually buying?
The first filter someone reaches for is the attribute their whole purchase decision hangs on — and it is rarely the one the brand leads with on the homepage. Apparel shoppers start with size. Skincare shoppers start with an ingredient or a concern. Garden shoppers start with sun exposure. Food shoppers start with flavor or a dietary restriction. The last filter before someone exits tells you where the experience stopped helping them — usually a combination that returned nothing, like size 12 in navy with no results and no alternative offered. Auditing which filters get used at all is equally useful: the ones nobody touches are noise slowing down the ones that matter, and removing them often does more for conversion than adding another.
How does discovery data improve email marketing and paid media performance?
Browse and search behavior make sharper email segments than opens and clicks ever will. Someone who filtered for a specific size and left is a different person from someone who skimmed the homepage for forty seconds — and the email they receive should reflect that. Segments worth building include shoppers who viewed three or more products in one collection without adding to cart, used a specific filter and left, or searched the same term across two separate sessions. For paid media, watch for landing pages where a visitor’s first action is a site search — that is a shopper signaling the ad promised something the page did not deliver. High-volume search terms are proven demand in the customer’s own phrasing, which means the next landing page and ad copy are already written for you.
What are the four discovery behaviors most worth checking first?
The zero-result search report reveals synonyms to fix, gaps in the catalog, and dead URLs that need redirects. The last filter used before someone exits shows where the experience stopped helping — usually a size-and-color combination that returned nothing, with no alternative offered. Products viewed multiple times with no add to cart signal a specific unanswered question about fit, price, or timing — the answer is almost always sitting in support tickets and reviews where the same concern keeps coming up. And collection pages with strong entrances but weak product clicks are pages that are not helping anyone choose, usually because of too many products before a decision is possible, a default sort nobody would pick, cards that do not differentiate one item from the next, or out-of-stock products sitting at the top taking the best real estate on the page.
Written by: Nathan Pitchan
Full-time daydreamer and professional persuader, Nathan is a fearless word alchemist crafting copy that connects, converts, and feels undeniably magnetic. E-commerce. Food & beverage. Horticulture. Destination tourism. And beyond. In a world overrun by AI-generated fluff, his secret sauce is unmistakable: engaging, conversational, and deeply human storytelling. Why? Robot copy gets read. Human copy gets remembered.