IKEA's website is a warehouse with a search box

Type "ivar" into IKEA's search box and you get exactly what you'd expect: the shelving unit, its accessories, and nothing else. Type "bookshelf for a small room with kids" and you get almost nothing useful, because IKEA's search doesn't try to interpret that sentence. It matches against product names, categories and a controlled set of attributes, and if your query isn't shaped like one of those, the system shrugs. Most retail teardowns would call that a failure. It's worth asking first whether it's a decision.

The naive shape for a teardown like this is checkout friction: how many clicks to buy, how many form fields, where the drop-off probably happens. That framing gives IKEA's site a mediocre grade for reasons that are mostly beside the point. IKEA is not selling like a DTC brand selling one hero product to a browsing audience. It's running a catalogue of thousands of SKUs against a real, physical network of stores, distribution centres and delivery zones that don't carry the same stock or charge the same price two towns apart. Everything about the interface that reads as awkward when judged against a slick single-product storefront reads as correct once you notice that the website's job is to represent that network honestly, not to hide it.

That distinction matters more broadly than IKEA. A single-person site has none of this to manage — no stock, no store network, no delivery zone — which is exactly why a tool like reach can afford to skip an entire category of problem IKEA can't. reach turns a CV into a live one-page site: upload a résumé and a photo, pick a look, and a finished page is generated in about twenty seconds, live at a free address in under two minutes. That works because there's nothing location-dependent to resolve — one person, one page, no inventory sitting in a warehouse three postcodes away. It also means reach has drawn its own hard line in the other direction: no e-commerce of any kind on a site it generates, no cart, no checkout. IKEA's entire site exists to manage exactly what reach has ruled out by design. Neither choice is a compromise; they're both a company reading its own constraint correctly and refusing to build past it. IKEA is worth reading closely for the same reason — not for what it botches, but for the fact that almost none of its "friction" is actually friction once you know what problem each screen is solving.

Location and stock are the objects everything else hangs off

Open IKEA's UK or US site without setting a store and you'll notice prices and delivery estimates are often provisional or missing entirely until a postcode is on file. That's not a dark pattern designed to delay you past a payment screen — it's a straightforward consequence of a pricing and logistics model where a sofa's price, delivery window and even its packaging count can differ between two distribution centres forty miles apart. The postcode gate isn't decoration on top of the catalogue. It's the catalogue's actual index.

Once a store is set, the availability panel on every product page becomes the site's real centre of gravity. It doesn't just say "in stock" — it typically breaks the item down by warehouse quantity versus showroom quantity, by collection point versus home delivery, sometimes by which specific store within driving distance has it and which doesn't. A competitor selling one warehouse's worth of stock nationally has no reason to build this. IKEA has no choice, because the alternative is telling a customer an item is available and then failing to deliver on that promise at the point of collection — the single worst failure mode a network this size can have on its site.

This is the piece that a checkout-friction reading misses entirely: the postcode gate and the availability panel aren't obstacles between the visitor and the product. They're the product information a visitor with a warehouse-scale catalogue actually needs, arriving before the parts that would matter for a simpler catalogue — because for IKEA, whether an item exists where you can get it is a genuinely open question in a way it never is for a retailer with one distribution model.

The product page answers logistics before it answers desire

Scroll an IKEA product page and the order is consistent enough to be a template: a hero image, then dimensions, then availability, then assembly information, and only after all of that does anything resembling lifestyle photography or "how people use this" content show up, usually near the bottom or on a separate linked page entirely. That's close to the mirror image of a DTC furniture brand's product page, which typically leads with a styled room shot and buries the tape-measure facts near checkout.

The ordering makes sense once you accept that IKEA isn't trying to create desire on the product page. Desire, for IKEA, is created elsewhere — in a physical store people already walk through by the million, in the room-set photography discussed below, in decades of brand recognition that means most visitors arrive at a product page already having decided, or close to it. What the product page has to do is confirm the decision survives contact with reality: does this actually fit the alcove, will it come flat-packed in a box that fits a hatchback, does it need two people and forty minutes or one person and ten. Dimensions first, assembly information prominent, and a link through to full assembly instructions as a PDF — that's not a site under-investing in persuasion. It's a site that has already priced in where persuasion happens, and it isn't here.

The assembly information deserves its own note, because it's unusually specific for a retail product page: estimated time, number of people needed, sometimes tool requirements, occasionally a warning about wall-mounting. That level of pre-purchase disclosure is the kind of thing a company builds only after enough support tickets and returns from people who bought something they couldn't physically get up two flights of stairs or into their existing wall structure. It's evidence of the physical world pushing back on the interface until the interface adjusted.

Search and category navigation hold up structurally and break on intent

For a catalogue this size, IKEA's category navigation is genuinely well organised: room type, then furniture type, then a set of filters built from actual product attributes — width, material, colour, price band. Because IKEA's own attribute data is clean (it has to be; the same data drives warehouse logistics), those filters return exactly what they claim to. Filter for "under 120cm wide" and you get items that are under 120cm wide. That sounds like table stakes, but plenty of retailers with messier back-end data can't promise the same thing.

Where it breaks down is intent-based search — the "bookshelf for a small room with kids" query from the opening. IKEA's search is built for people who already know roughly what they're looking for and can name it, or at least name its category and rough attributes. It is not built to interpret a described problem and propose a product, the way a marketplace search increasingly tries to. Given the range's actual complexity — dozens of shelving systems that look superficially similar in a thumbnail but differ meaningfully in depth, weight rating and mounting method — a search engine that guesses too aggressively would routinely put people in front of the wrong item, which for furniture with real installation and return costs is a worse failure than search returning nothing. Refusing to guess is defensible caution dressed up as a limitation.

Room-set inspiration content is a small fraction of the site, on purpose

IKEA's room-set photography — full living spaces styled around a look, browsable as "shop the look" collections — is genuinely good, and it is a small footprint on the overall site by page count and by how deep in the navigation it sits relative to the product catalogue and stock tools. That ratio tells you something. Inspiration content is IKEA's homepage promise, in the sense the method for reading a company's public surface uses the term: the one thing worth the most prominent real estate. But prominent placement of a small number of pages next to a catalogue running to thousands of SKUs is itself the signal. IKEA isn't trying to be a lifestyle magazine with a store attached, the way a brand-forward retailer covered in the Patagonia site teardown leans hard into editorial and story content because brand narrative is close to the entire product. IKEA's inspiration content exists to season a catalogue that does the actual selling through range, price and physical stores. It doesn't need to be large to do its job; it needs to exist and be well made, and then get out of the way of the thousands of product pages behind it.

This is also, notably, the one part of the site where IKEA is not location-bound. A room set looks the same in every market it's translated into, unlike a product page, which cannot escape a postcode. That split — brand content that travels, logistics content that doesn't — is a cleaner architecture than it might look from outside, and it maps onto a broader question worth asking of any company's site: how many pages a small company website actually needs to do its job well is rarely "as many as possible." IKEA's inspiration section answers that question for itself by staying deliberately small next to a catalogue that has no such option.

What a constrained business correctly refuses to optimise

Read against a clean, minimal, single-product storefront, IKEA's site looks over-engineered in some places and under-designed in others. Read against what the business actually is — a company that has to tell the truth about stock across a real physical network, at a scale where getting it wrong means a customer standing at a collection desk with nothing to collect — almost every one of those apparent rough edges turns out to be the correct answer to a constraint the comparison set doesn't have.

The postcode gate, the availability panel, the assembly disclosure, the intent-shy search: none of these are failures of craft. They are the visible cost of running an interface honestly on top of a warehouse network, and IKEA has clearly decided that cost is worth paying rather than smoothing over with a friendlier interface that occasionally lies to someone about whether a sofa exists where they can get it. The inspiration content stays small because it doesn't need to carry the business; the catalogue tools stay heavy because they do. That's not a company that failed to modernise its checkout. It's a company whose actual product is stock and location, correctly refusing to spend its design budget pretending otherwise.