← ~/blog
· 7 min read

The Search Bar is Dying: How AI is Replacing Navigation with Intent

We're moving from navigation (clicking menus) to intention (typing what we want). A deep analysis of why your product design is about to become obsolete, with examples, data, and a redesign framework.

The Search Bar is dying. And the Hamburger Menu is next.

We are moving from “Navigation” (clicking menus) to “Intention” (typing what we want). This isn’t a trend. It’s a fundamental shift in how humans interact with software. And most product teams aren’t ready for it.

The Old World (1995-2023)

Navigation-Based

Users learn your mental model. They click through menus, breadcrumbs, and tabs to find what they need. The interface is a map.

Settings > Account > Privacy > Change Password

The New World (2024+)

Intent-Based

The software understands your intent. You state what you want in natural language. The interface is a concierge.

“Change my password”


The Evolution of Software Interfaces

Every decade, the dominant interface paradigm shifts. And each shift makes the previous one look absurd in hindsight.

EraInterfaceUser BurdenExample
1970sCommand LineMemorize exact commands$ passwd -e username
1990sGUI (Windows/Mac)Learn menu structureFile > Edit > Preferences
2000sWeb (Browsers)Navigate page hierarchyHomepage > Settings > Account
2010sMobile AppsLearn app-specific patternsTab bar, hamburger menu, swipe
2024+Intent-Based AIJust say what you want”Change my password”

The pattern: Each generation reduces the user’s burden. We went from memorizing commands to learning menus to just stating intent. The friction of finding the feature is disappearing. The only friction left is articulating the intent.


Case Study: Perplexity vs. Google

Perplexity is the clearest example of intent-based design eating navigation-based design.

Google (Navigation-Based)

1. Type query in search bar

2. Scan 10 blue links

3. Click a link, read the page

4. Go back, click another link

5. Synthesize the answer yourself

5 steps. User does the synthesis work.

Perplexity (Intent-Based)

1. Type question in natural language

2. Get synthesized answer with sources

2 steps. AI does the synthesis work.

There is no “News” tab or “Images” tab in Perplexity. There is just a box that asks: “What do you want to know?”

The friction of finding the feature is gone. The only friction left is articulating the intent.


The Feature Bloat Problem (and How AI Solves It)

Every mature SaaS product has this problem: they keep adding features, and the UI keeps getting more cluttered.

MetricNavigation-BasedIntent-Based
Adding 10 new features10 new buttons/pages0 new UI elements
UI complexity at 100 featuresOverwhelmingSame as at 10 features
User learning curveIncreases with featuresStays flat
Feature discoveryMenu archaeologyJust ask for it
Onboarding timeGrows linearlyNear-zero

The Scaling Insight

In the old world, you added buttons for every new feature. The UI got cluttered. In the new world, you add capabilities to the model. The UI stays the same. A simple text box can scale to 1,000 features without adding a single pixel of clutter.


The Cmd+K Revolution Was Just the Warm-Up

The “Command Palette” (Cmd+K) that Slack, Notion, Linear, and others popularized was the first step. It let power users bypass navigation and type what they wanted.

But Cmd+K is still search, not intent. You’re searching for a feature by name. The AI version goes further:

InterfaceUser InputSystem Response
MenuClick Settings → Notifications → EmailShows email notification settings
Cmd+KType “email notifications”Jumps to email notification settings
Intent AIType “stop sending me emails on weekends”Changes the setting for you

The difference: Cmd+K helps you find the feature. Intent AI does the thing.


What This Means for Product Teams

The Death of the Sitemap

For 20 years, UX design started with information architecture. You built sitemaps, defined navigation hierarchies, designed breadcrumbs. All of this assumed the user would browse your product.

In an intent-based world, the sitemap is irrelevant. The user never sees your hierarchy. They just state what they want.

The Rise of the “Capability Map”

Instead of sitemaps, product teams need capability maps: a structured list of everything the AI can do, organized by user intent:

User IntentCapabilityRequires
”Schedule a meeting with Sarah”Calendar access, contact lookup, availability checkCalendar API, contacts DB
”Show me last quarter’s revenue”Data query, visualization, summarizationAnalytics DB, chart library
”Refund this customer’s order”Order lookup, payment reversal, notificationOrder system, payment API, email service
”Why is page load slow today?”Performance monitoring, log analysis, root causeAPM data, log aggregator

The New Design Process

Old Process

1. Define features

2. Design information architecture

3. Design screens & flows

4. Build navigation

5. Onboard users to the UI

New Process

1. Map user intents

2. Define capabilities per intent

3. Design the conversation (not the screen)

4. Build tool integrations

5. Let users discover by asking


Where This is Already Happening

ProductWhat ChangedResult
Notion AIAdded “Ask AI” to every page, replacing search, formatting, and 50+ menu itemsUsers perform actions 3x faster via AI than via menus
Shopify SidekickMerchants ask “how were sales last week?” instead of navigating to Analytics40% reduction in support tickets for “how do I find X?”
LinearNatural language project management: “create a bug for the login page crash”Engineers spend less time in UI, more time coding
Arc Browser”Command bar” replaces bookmarks, history, tabs, and settings navigationSingle interface for everything

The Bottom Line

The best interface is the one that isn’t there.

Stop designing screens. Start designing conversations. The “Command K” revolution was just the warm-up. The future isn’t “better navigation”. It’s no navigation. The product that wins is the one where users never have to wonder “where is that feature?”, because they can just ask for it.

The search bar won’t disappear overnight. But the paradigm it represents (forcing users to translate their intent into your product’s language) is dying. The future is products that speak the user’s language. And that future is here.

Enjoyed this? Get more like it.

Weekly on AI product strategy and execution. No fluff.

Unsubscribe anytime.

share: twitter linkedin

Comments

Loading comments...