The core mistake: intent is a signal, not a market

Ecommerce teams keep making the same expensive mistake, and it has the emotional sophistication of a shopping cart with one wheel missing. They see a search query, read it as demand, then build a page as if the query itself proves a market exists. It does not. A query proves that someone typed words into a search box. That is a signal, useful and often valuable, but still just a signal. If a person searches for “black linen trousers,” that tells you something about intent. It does not tell you whether there are enough people like that person to justify a dedicated page, indexation, internal links, and ongoing maintenance. Search behavior is evidence of curiosity or need. Market size is a separate question.

This distinction matters because a single person looking for a product and a market large enough to support a page are very different things. One shopper can create a spike in query volume, especially in categories where language fractures into size, color, material, season, fit, and use case. That is why keyword tools can make a tiny pocket of demand look larger than it is. A term may have 200 searches a month, but if those searches are spread across branded terms, near-duplicates, and long-tail variants, the real opportunity is thin. A page deserves to exist when it can capture meaningful incremental demand, meaning demand you would miss without that page. If it merely collects traffic that would have found another page anyway, you are rearranging deck chairs and calling it merchandising.

Bottom-funnel pages get made because they feel safe. They are easy to explain in a meeting, easy to measure in a dashboard, and easy to defend when someone asks what the SEO team shipped. “We created a page for this query” sounds disciplined. It also flatters the team’s sense of control. Yet the safety is mostly cosmetic. A page built for a narrow intent can look tidy while contributing almost nothing beyond a small amount of cannibalized traffic. In many catalogs, the page becomes a duplicate of an existing category, product listing, or filter state, which means the only thing it really changes is the URL structure. That is a lot of ceremony for very little commerce.

The real question is sharper, can this page capture demand that would otherwise stay out of reach, or does it only reshuffle existing demand across more URLs? That is the test. If the answer is reshuffling, the page is probably a vanity asset dressed up as search strategy. Senior marketers should treat intent the way a good trader treats volume on a thin stock, interesting, but not proof of depth. The market decides whether a page matters, not the query itself.

Why ecommerce teams overproduce bottom-funnel pages

Ecommerce teams overproduce bottom-funnel pages because page creation is one of the few marketing activities that looks like progress in a meeting. A new category page, a new landing page, a new filtered collection, all of it produces visible output that can be pointed to on a slide. That matters inside organizations. Teams are rewarded for activity that feels concrete, especially when the output can be tied to a query and a ranking. A page is easy to approve, easy to assign, and easy to report on. If it ranks, the story is tidy. If it does not, the team still has something to show for the work. That is a powerful incentive, and it pushes marketers toward volume. Humans love a visible artifact. It makes everyone feel less haunted by the spreadsheet.

Search volume tools make this habit worse because they turn fuzzy demand into neat numbers. A keyword with 1,300 monthly searches feels real, even when the number is a rough estimate spread across dozens of related queries, mixed intent, and multiple geographies. In practice, many terms are too coarse to justify a page. A search for “women’s boots” may represent someone comparing styles, someone looking for sizing advice, someone ready to buy, and someone just browsing. The tool hands the team a number, the number creates confidence, and confidence becomes a page. That is how marketers end up treating estimated volume as if it were a demand warrant. It is not. It is a weather report with a haircut.

The bigger mistake is that teams confuse three different things. Category-level intent means the shopper wants a product class, like sofas or running shoes. Modifier-level intent means the shopper adds a descriptor, like leather, wide fit, black, or waterproof. True commercial demand means enough people are actively searching for that exact combination that a standalone page can capture meaningful traffic and revenue. These are not the same. A modifier can signal interest without signaling a page-worthy audience. Yet teams build pages for every adjective they can find, as if “white linen shirts,” “oversized white linen shirts,” and “women’s oversized white linen shirts” each deserve their own destination. That is how catalogs turn into SEO confetti, pretty for a minute, impossible to clean up later.

The SEO playbook trained marketers to think in pages first, because pages are the atomic unit of ranking. One query, one page, one title tag, one set of internal links, one report. That model is tidy, and it has taught entire teams to ask, “What page should we make?” before they ask, “Where is demand actually concentrated?” The better question is usually about aggregation. If a dozen modifiers all point to the same buying job, the answer is one strong page, not twelve thin ones. Search does not reward the most pages. It rewards the page that matches real demand with enough depth, clarity, and authority to deserve attention. The internet is many things, but it is not impressed by paperwork.

The economics of thin intent pages

The first mistake is treating every query as if it deserves its own page. In practice, a cluster of near-duplicate bottom-funnel pages behaves like a tax on the whole site. Crawl equity gets split across pages that say almost the same thing. Internal links get scattered. Editorial attention gets diluted across a dozen pages that each need maintenance, refreshes, and alignment with the rest of the site. Search engines are very good at noticing when a site keeps producing variations on the same theme, and thin pages tend to pull the architecture downward, because the site starts to look less like a clear information system and more like a pile of adjacent landing pages. That pile may be alphabetized, but it is still a pile.

This is where index bloat enters the picture. A bloated index is full of pages that technically exist, technically target a query, and technically earn impressions, yet contribute very little to the site’s overall relevance. That weakens the signal of the stronger pages around them. If a retailer has one solid category page and six thinner variants for slight intent differences, the search engine has to sort through a mess of overlapping relevance. The result is often worse ranking stability, weaker internal linking flow, and a site structure that looks busy without becoming more useful. Search engines reward clarity. Thin intent pages create noise, and noise is where good rankings go to lose their keys.

A page can rank and still fail commercially. That is the part many teams miss, because ranking feels like proof, and it is only proof of visibility. If a query has tiny demand, the page may collect a trickle of traffic and never move revenue. If the conversion path is weak, maybe because the page sits too far from a real buying decision, the visits wash away. If the page cannibalizes a stronger page, it can even steal the better ranking from the page that actually converts. Search Console is full of these false wins, pages that sit on page one for a query with almost no commercial weight, or pages that earn clicks while suppressing a broader, more valuable page. The dashboard applauds. The bank account shrugs.

The real cost is opportunity cost, and that is where the economics get ugly. Every thin page competes for the same editorial hours that could have gone into a category page, a buying guide, a comparison page, or a stronger internal linking structure. It also competes for engineering attention, content review, and maintenance. A site does not have infinite capacity to publish and prune. When teams spend that capacity on weak intent pages, they are choosing marginal gains over compounding gains. That is a bad trade. A page that ranks for a tiny query is not free if it keeps a better page from becoming the page that wins. The hidden invoice always arrives, usually when no one is looking at the right dashboard.

What real demand looks like

Real demand is repeated, commercially meaningful interest that keeps showing up after a single query pattern has been stripped away. One search for “black running shoes” is intent. A steady stream of searches for “black running shoes,” “men’s black trainers,” “all black sneakers,” and “black gym shoes” is demand. The difference matters because intent can be momentary, shaped by one session, one ad, or one product detail page. Demand survives contact with the market. It shows that many shoppers, over time, want the same thing and are willing to spend for it. That is the part that pays rent.

That is why head terms, modifier terms, and demand clusters cannot be treated as the same thing. A head term like “sofa” is broad and messy, often full of research behavior. A modifier term like “corner sofa bed” or “small leather sofa” is narrower, and often closer to purchase. A demand cluster is the full set of related searches and behaviors around one need, including the head term, the modifiers, and the phrasing people actually use. Clusters matter because they reveal concentration. If 20,000 people each search a slightly different version of the same thing, that is a market. If 20,000 people each search something unrelated, that is noise with a spreadsheet attached.

Real demand shows up in more than search volume. It appears in internal site search, where shoppers type the same phrases again because the site navigation did not answer them. It appears in direct traffic, when people come back by name because they already know what category they want. It appears in paid search, where the same terms keep producing efficient clicks and conversions instead of one-off curiosity. It appears in marketplace behavior too, through repeated ranking pressure, review volume, and the same products being compared again and again. Retail media adds another signal, because spend follows proven shopper movement, not wishful thinking. When several channels point to the same need, the market is speaking plainly. Loudly, usually with a receipt attached.

The best pages are built around demand concentration, where many shoppers want the same thing in slightly different language. That is the difference between a page that exists to catch a keyword and a page that earns its place because the market has already formed around it. Think of it like a crowded train platform. One person standing there means little. A hundred people standing there for the same train means you build for that route. Ecommerce pages should follow the crowd, not the dictionary. When demand is concentrated, the page has a job. When it is scattered, the page is decoration.

How to decide whether a page deserves to exist

The first question is whether the query cluster is large enough to justify its own page. That sounds obvious, yet teams still build pages for search terms that show up in a tool because they are visible, not because they matter. A page deserves its own URL only when the cluster has real demand, a distinct shopping job, and enough repeat volume to stay relevant. If the query set is thin, seasonal, or made up of one-off phrasing, the page is a bet on noise. In practice, I want to see evidence from search behavior, internal site search, and customer service language that the same problem keeps returning in a durable way. A page for “waterproof trail running shoes for wide feet” makes sense if the same need keeps appearing across many queries. A page for “blue waterproof trail running shoes with reflective laces” is a merchandising accident wearing SEO clothes.

The second check is overlap. If an existing category page, buying guide, or filter already answers the query well, a new bottom-funnel page is usually a mistake. Overlap creates cannibalization, which means your own pages compete for the same intent and none of them wins cleanly. Google has been clear for years that it ranks pages, not keywords, and when multiple URLs chase the same intent, the result is often instability rather than growth. The practical test is simple, does the proposed page answer a different shopping problem, or is it just a more specific version of something you already publish? If the answer is the latter, the right move is to improve the existing page, not split the demand into another URL and hope search engines sort it out. Hope is a lovely emotion. It is a terrible content strategy.

The third check is fit. Can the page answer the shopping problem better than a broader page can? That is the real bar. A broad category page can handle many intents when the product set is broad and the decision is still early. A buying guide can explain tradeoffs, like fit, material, or use case. Filters can help users narrow by size, color, or feature without forcing a new page for every combination. Editorial content can handle education, comparison, and “how do I choose” questions. A standalone bottom-funnel page only earns its keep when it gives the searcher a cleaner answer than those other formats. If the intent is “best running shoes for flat feet,” a guide or category page may serve it better than a thin page built around one phrase. Searchers want resolution, not a URL tax.

That is why page creation should be the exception, reserved for demand that is both distinct and durable. Distinct means the intent cannot be handled cleanly by a broader page, filter, or guide. Durable means the need will still exist after the current keyword excitement fades. This is where many ecommerce teams go wrong, they confuse a query pattern with a business case. A search cluster is not a mandate. It is evidence. Treat it that way. Build pages when there is enough demand, enough separation from existing content, and enough staying power to justify another entry point into the site. Everything else belongs somewhere else in the information architecture, where it can do its job without multiplying pages for no good reason.

What to build instead of more bottom-funnel pages

The answer is not another page for every modifier someone can type into a search box. The answer is a cleaner category architecture, stronger internal linking, and fewer pages that each do a clearer job. A site with 200 thin pages aimed at tiny variations usually creates more crawl noise than revenue. A site with 40 well-structured pages, each connected to the others with intention, gives search engines a map and shoppers a path. That matters because Google has said for years that internal links help it understand which pages matter, and in practice the pages with the best structure usually win the traffic, even when they are not the most exact keyword match. Search is many things, but it is not sentimental about your favorite URL.

Editorial content should sit beside commerce content, not be treated as a separate hobby. Buying guides, comparison pages, and refinement pages can absorb demand that would otherwise be split across a dozen weak pages. Think about how a shopper actually behaves. Someone searching for “best running shoes for wide feet” is expressing a need, and that need can be met by a strong guide that explains fit, cushioning, stability, and return policy, then points into a well-built category page with filters and sorting. That single page can satisfy the query, earn links, and move people toward product. A swarm of pages for “wide running shoes,” “running shoes for wide feet women,” and “wide toe box running shoes” only dilutes the signal. More pages, less clarity. A classic modern tragedy.

Better merchandising usually does more work than another narrow page. A category page with clear sorting, visible filters, strong product grouping, and copy that explains the range can serve both broad and specific intent. Taxonomy does the same job behind the scenes. If the product structure is clean, a shopper can move from “trail shoes” to “waterproof trail shoes” to “winter trail shoes” without the site needing a separate article for each micro-intent. Retailers have spent decades learning this lesson offline, where a better shelf arrangement sells more than a new aisle sign. Ecommerce is no different, except the shelf is indexed by search engines and judged by people who have the patience of a caffeinated squirrel.

The pages worth building are the ones that can earn links, answer broader questions, and support more than one conversion path. A strong category page, a serious guide, or a well-written comparison page can attract editorial links because publishers want something useful to point readers toward. It can satisfy a shopper at the research stage and still work for the shopper who is ready to buy. A thin page built for one modifier rarely does either. The business case is simple, build pages that compound, not pages that split demand into fragments and call that growth. Compounding is boring in the best possible way, which is usually how durable revenue behaves.

The editorial discipline senior marketers should enforce

The operating rule is simple, and senior marketers should enforce it without apology, every new page must earn its place with demand, differentiation, and a clear job inside the site architecture. Demand means there are enough people searching, comparing, or arriving with that need to justify the page. Differentiation means the page says something distinct, because if it reads like a copy of five other pages, it is dead weight. Clear job means the page has a role, such as capturing a specific query, supporting a category, or resolving a comparison that matters to buyers. If one of those three is missing, the page is a nice idea, not a business asset.

That is why page creation should be reviewed like capital allocation. Every page consumes editorial time, search equity, internal linking attention, and maintenance. Those are real costs, even if they do not show up on a P&L line. McKinsey has long argued that disciplined capital allocation separates strong operators from average ones, and the same logic applies here. A page that gets approved should pass a simple investment test, what demand exists, what gap does this fill, what gets displaced if we build it, and how will we know it earned its keep. If the answers are vague, the page is a vanity spend dressed up as strategy. The internet has enough vanity already, thank you very much.

The same discipline should be applied to the archive. Audit bottom-funnel pages by asking four blunt questions. Does this page attract meaningful demand, or does it sit in the long tail with no traffic and no links? Does it rank for a query that deserves its own page, or is it cannibalizing a stronger asset? Does it offer unique information, or does it repeat the same claims in different clothing? Does it still reflect how buyers search and compare, or has the market moved on? Pages that fail this test should be consolidated into a stronger page, pruned if they add nothing, or repurposed if they can serve a broader informational or comparison role.

This is where senior marketers earn their keep, by saying no to page sprawl. A site with 40 thin bottom-funnel pages is usually weaker than a site with 12 pages that each have a reason to exist. Search engines reward clarity, users reward usefulness, and internal teams benefit from fewer assets that actually matter. Intent is useful because it tells you what people are trying to do. Demand pays the bills because it tells you whether enough people are trying to do it. The brands that win treat pages as investments, not as decorations for every possible query.

Frequently asked questions