2025

Scoped Search

Sales reps in large organizations spend more time finding content than selling. Scoped Search gives reps a fast, low-friction way to search within the content that's relevant to them.

Role
Lead Product Designer
Stage
Zero to One
Domain
Sales Enablement
Timeline
2025
Scoped Search Scoped search interaction or final design

Scoped Search — search UI with collection scope selector

+Problem / context visual

Open access at scale — content sprawl in Highspot

Open access to everything means finding nothing fast

In large organizations, Highspot gives sales reps access to almost all content in the platform. Paradoxically, this abundance creates friction — reps spend more time searching for the right content and less time selling.

Open access at scale doesn't serve the rep or the org. The problem isn't a lack of content. It's that relevance is buried inside volume.

+Content collections diagram

From spots to content collections — the organizational layer

Content Collections: grouping spots into a searchable scope

Content in Highspot lives in spots — folder-like constructs with individual access control. Every item added requires a manual decision about which spot it belongs to.

To make scoped search possible, we first introduced content collections — an admin-defined grouping of related spots. Once a collection is configured, item-to-collection association becomes automatic.

Collections map naturally to how large orgs already think: by sub-org, product line, or region. They don't replace spots — they aggregate them into a meaningful scope.

Scoped Search

Early exploration — broad interaction patterns

The more you can scope, the harder it is to scope

The first iteration went wide — exploring every way a user might select a collection scope before or during a search. The exploration surfaced a core tension: more collections meant more complex interaction patterns in the search dropdown.

The search dropdown is not a place for complexity. It needs to be fast and instinctive. Every additional option in that surface is friction.

The Insight

We went back to customers and studied actual usage. Users either have access to 1–3 collections, or they have access to more but realistically only care about 1–3. The design problem collapsed.

Users have access to many. They care about a few.

We went back to customers with a focused question: how many Collections do users actually have access to, and how many do they realistically use?

Users had access to many collections — but would consistently engage with only 1–3. This grounded a core design principle: make frequent tasks easy, and less frequent tasks possible.

Key Finding

Users had access to many collections but functionally used 1–3. Surfacing every collection equally in the search dropdown would optimize for a scenario that rarely happens.

Frequent
Promoted collections in the search dropdown
A small set of relevant collections appear directly in the dropdown — letting users scope before submitting. Fast, contextual, zero friction.
Possible
Collection filter in the results sidebar
For less frequent cross-collection searches, the sidebar filter handles it — exactly like any other facet. The dropdown stays clean. The sidebar scales to edge cases.

Search is a dynamic surface. Design every state, not just the default.

Interacting with search is not a single action — it's a sequence. Every keystroke, pause, and click changes what the dropdown should show. If transitions aren't fluid, the perception of speed breaks down even when results are fast.

01
Optimize for simplicity
Scoping should feel quick, not like a configuration step. Every extra decision in the dropdown is friction.
02
Fluid transitions
The perception of speed matters as much as actual speed. State changes should feel invisible.
03
Dropdown is prime real estate
Keep it clean and fast. Rely on sidebar filters for scalability — don't let edge cases crowd the common case.
04
Bring the collection to the user
Don't make users hunt for their scope. Surface the 1–3 relevant collections — promoted, not buried in a list.

Five states. Each with a distinct job.

We redesigned each state of the search surface from scratch — deciding what's important at each stage and how to arrange it. The transitions between them became as important as the states themselves.

The scoped variants (Scope + Zero, Scope + Auto-suggest) mirror their unscoped counterparts with a collection pill added to the bar. Removing the pill always takes you back to the plain version of that state.

State 01
Zero
User clicks in. Show recent history and promoted collections.
State 02
Scope + Zero
User has scoped, but not typed. Same history, pill in bar.
State 03
Auto-suggest
User is typing. Keyword and entity suggestions, fast and fluid.
State 04
Scope + Auto-suggest
Scoped suggestions — filtered to the selected collection.

Interactive prototype

What this product taught me

Learning 01
The right scope is smaller than you think
Building for the full range of possibilities produces complexity nobody asked for. Users self-limit in predictable ways. The best design honors that — and optimizes hard for the common case rather than softly covering every case.
Learning 02
Transitions are a design deliverable, not an afterthought
Search is a dynamic surface — every action changes what the user sees. If the transition between states isn't deliberate, the interaction feels broken even when the logic is correct. Designing the states and the motion between them are equally important.