Navigation Shortcuts: When Menus Get in the Way
Most navigation design work focuses on the usual suspects: mega-menus, breadcrumbs, sidebar layouts, and search with filtering. These are all essential, but a handful of patterns tend to get overlooked. Think of them as shortcuts — ways to get users to their destination without routing them through the traditional navigation hierarchy at all. On large sites with thousands of pages, many of which have been gathering digital dust for years, these alternatives can end up being considerably more effective.
Queries Instead of Menus
The typical web experience is a roundabout affair. Visitors arrive with a specific intent, yet they must translate that intent into a sequence of keyword searches, menu clicks, and selections. The answer is rarely served up directly; instead, users discover it through a long-winded journey across pages and sub-navigation items. As Gerry McGovern once noted, more people have stood on top of Mount Everest than have visited the tenth page of Google's search results. Despite this, interfaces tend to present long lists of options rather than taking advantage of context and association. Users are left to survey the landscape on their own, jumping between menu items, iterating on search queries, and scouting tags and footer links.
To shorten the distance between intent and action, we can query users about what they want. The concept of a "query constructor" is not new — variations of natural language forms, Madlib-style prompts, and chatbots have all been used to let users specify intent without typing in a search box or navigating a menu. The pattern works dynamically: present options, and based on each answer, surface the next relevant set of choices, guiding users all the way to the page they need.
On AO.de, the front page immediately queries the type of device a customer is interested in. Once selected, another filter option appears, followed by a third, and finally a slider to set a price range. Customers can reach relevant results without touching the navigation or search at all. The experience works much like drill-down navigation, but users are making one small decision at a time rather than being confronted with the entire navigation structure at every step.
Commonbond.co applies the same idea in a dedicated area beside its primary navigation. A drop-down asks what visitors want to do on the site; when an option is selected, a second drop-down appears to narrow the intent further. Cork Chamber takes this a step further by placing an "I want to" query in a primary spot in the navigation itself, driving users directly to the relevant page — a notably user-centric framing of what might otherwise sit behind a generic menu item.
The pattern also handles situations where the target page is buried deep in the information architecture. Sbahn.berlin, Berlin's public transport site, offers a choose the view that fits you best selector that jumps users straight to fourth-level navigation pages they'd otherwise have trouble finding. The City of Düsseldorf's service portal uses two central drop-downs — one for intent, one for location — to produce a direct link to the right appointment-booking page, bypassing both global navigation and external search engine detours.
Monday.com adopts a similar approach in its onboarding flow. Prospects first select what they'd like to manage with the product, and that choice triggers a relevant onboarding path. It's a clean way to minimize the distance between a stated intent and the moment a user sees value.
Comparison Without the Table Matrix
Classic feature comparison tables demand a lot from users. When comparing a handful of products — headphones, say — the typical interface serves up a multi-column table packed with attributes, expecting shoppers to scan row by row, right to left and back again. That lawn-mower pattern of inspection is tiring and slow, especially when most shoppers have only a few specific attributes in mind.
Nobody wakes up hoping to finally compare products in a matrix. Users actually want to discover which option is better for them. Instead of forcing a table on users, a smarter approach asks them what they care about first.
Productchart.com sidesteps tables entirely. All products are mapped in a two-dimensional space where customers choose the attribute on each axis. Filters trim the field, and products of interest can be highlighted and compared side by side.
Mediamarkt goes further, dropping the table altogether during initial comparison. When users elect to compare products, they are prompted to choose relevant attributes first. Each selection becomes one step in an evaluation journey, where users vote products up or down based on which features they have. When the process is complete, a winning option is presented based on the user's own preferences. The full feature matrix remains available as an optional view, and it can be exported as a PDF — but only for those who genuinely want the exhaustive breakdown.
When Taxonomy Breaks Down: The A-Z Index
As websites grow, they often suffer from what information architects call navigation decay. Content managers add new items wherever space allows, avoiding the hard work of retiring outdated content. Old categories linger alongside new ones, and the result is an architecture full of inconsistencies, mislabeled content, and duplication. The proper fix involves governance and a redesign of the information architecture itself, but that can take years — during which time frustrated visitors turn to search engines, often landing on competitors' sites.
The A-Z Index pattern is a pragmatic stopgap. The approach starts by identifying the top tasks users come to perform and the keywords they associate with those tasks. Tree testing validates that those keywords map to findable pages. Then, a single page catalogs all of it — every keyword a link that pushes users straight to relevant content.
Large public service sites lean heavily on this pattern as a complement to global navigation and search. Usability testing shows that listing keywords on one long page works better than splitting them across per-letter pages, largely because users can rely on in-browser search to scan the list quickly without navigating through multiple views.
The pattern can go further. Instead of merely linking out to individual pages, the index itself can surface information users want — opening hours, booking links, locations. Placing that detail inline is a real alternative to an accordion, which would need an expand/collapse all control to be usable at that scale.
Aarhus University demonstrates another variant: the A-Z index is part of the global header, so visitors can jump from role to letter to specific faculty or department from anywhere on the site. This pattern is unique in offering direct cross-page navigation without the standard series of category-to-subcategory clicks.
The caveat is that keyword selection must be tested rigorously. If the vocabulary shown doesn't match users' mental models, the index becomes another dead end. Some sites pair the index with an in-index search box, which effectively recreates an autocomplete experience.
Tap-Ahead Autocomplete
Autocomplete typically does one job well: it surfaces keyword suggestions as the user types. But the pattern is capable of far more. Instead of offering bare words, it can point users to categories, products, brands, or prepared collections of records. Sites such as Prisma.fi, Hema.nl, and Ikea.com use autocomplete to present category suggestions alongside products and rich details like price — giving users a snapshot of what's available rather than just a word list.
Statistics Estonia 100 extends that idea further by showing mixed result types in the dropdown, with each type labeled and recent statistics rendered inline for immediate inspection.
A more advanced variation is the tap-ahead autocomplete pattern. Here, the dropdown is designed to do double duty: users can jump directly to a suggestion, but they can also construct a query from the suggestions. Selecting a suggested term replaces or extends the text in the input field, letting users refine their intent across multiple iterations before ever hitting search.
Mediamarkt.de implements this cleanly. A user can click a suggestion to navigate immediately, or click the arrow beside it to populate the query in the input field for further refinement. The effort saved on typing compounds across each iteration, keeping users confident they're converging on the right results.
Stripe takes the same idea in a different direction. Filters are specified inline with the query, freeing users from post-search filter controls and delivering previews of results as they type.
Expert users with advanced needs can benefit as well. Stack Overflow combines inline querying with specialized filters that let power users construct precise searches straight from the search box — without leaving the input field or reaching for external tools.
Navigation That Reduces Friction
Reliable navigation patterns are the backbone of most sites precisely because they are predictable and familiar. On their own, however, they assume users follow hierarchical paths that complicate finding content scattered across vast or aging websites.
Navigation queries fill that gap by meeting users where they are. An A-Z index helps users who don't know where to begin; tap-ahead autocomplete gives those with partial intent a faster path to a well-formed query; inline feedback pushes them toward better phrasing before search even executes.
These techniques complement proven patterns rather than replace them. They offer alternatives for users coming from search engines, direct links, or content shared via email — those who have limited patience for top-level categories and need a faster route to a specific answer. Reducing typing effort and iteration cycles is worthwhile design work, especially when maintaining a perfect information architecture is years away.
