Problem

As Salesforce’s Einstein Search grew in adoption, a new challenge emerged: Admins and end-users were overwhelmed by irrelevant search results. Admins lacked control over which objects and fields appeared in results across different user profiles and search experiences. I led UX efforts to build Search Manager from the ground up—a no-code configuration tool enabling Admins to customize the search experience across Salesforce products, aligning with SLDS and core platform mental models.

Business Goals & UX Strategy

The UX strategy for Search Manager was aligned with Salesforce’s V2MOM framework and overall product strategy, ensuring a strong connection between product vision, platform scalability, and real-world admin needs.
 The business goal was to strengthen Einstein Search by improving result relevance, increasing Admin control, and supporting adoption across Salesforce orgs—without adding complexity.

Our UX approach focused on:

• Unifying configuration: Replace fragmented setup tools with a centralized, trusted entry point
• Empowering Admins: Deliver a no-code experience that gives control without requiring technical support
• Driving relevance and trust: Help Admins fine-tune what’s indexed for different profiles and use cases
• Designing for scale: Support both org-wide and profile-level customization with minimal friction
• Reducing setup burden: Optimize for low-frequency, high-impact tasks through smart defaults and clear UI patterns

We adopted a progressive UX strategy, launching with a focused MVP that delivered immediate value. From there, we planned to expand functionality iteratively, guided by direct Admin feedback and our product vision. With each release, our UX decisions were grounded in the specific goals of that phase—ensuring new features were intuitive, scalable, and aligned with evolving user needs.

Discover

Early in the project, we heard consistent complaints from Admins and end users alike: too much irrelevant data appearing in search results, limited control over what gets indexed, and an overall cluttered experience. These weren’t isolated issues—they were recurring pain points that directly impacted productivity and trust in the platform.

To dig deeper, I partnered with the PM to structure discovery around real-world use cases. We prioritized open-ended conversations with long-time Salesforce researchers and experienced Admins to hear their pain in their own words, without leading the discussion. One of the key questions we explored was: “How much control should be given to Admins versus individual employees when customizing search?” The answer was consistent: Admins wanted more control—but not just more settings.

They needed profile-based, channel-specific ways to manage which objects and fields were indexed. And critically, they wanted to do this within the boundaries of Salesforce’s trusted UI paradigms. Employees, on the other hand, preferred that Admins manage this complexity behind the scenes—they didn’t want yet another layer of customization to deal with.

Key Technical Insights

Define

Before exploring solutions, I designed a tailored approach to clarify the problem space. For every project, I prioritize strategic alignment—choosing design tasks that will bring the most clarity and momentum, whether that’s mapping jobs-to-be-done, facilitating team workshops, or translating research into visual frameworks.
With the insights in hand, I collaborated with the PM to distill the findings into a clear direction for the team.

My contributions included:

• Synthesizing insights from interviews and technical deep dives
• Crafting problem statements and “How Might We” questions
• Mapping out user jobs-to-be-done
• Creating visuals and frameworks to communicate core challenges
• Helping shape the PRD with PM
• Responding to early engineering questions to shape feasibility
• Aligning cross-functional teams through clear planning artifacts

I used storytelling and visuals to bring the problem space to life—making the PRD sharper and the engineering plan more focused. When technical questions arose, I partnered closely with PMs to break them down and align on solutions.

Key user goals emerged:

• I want to see the org-wide index
• I want to update the org-wide index
• I want to see the consumption of the index
• I want to configure search experience for specific profiles

The search experience in Salesforce isn’t generic—it must be tightly integrated with workflows. Our approach demanded a native, trusted, and frictionless experience that respected how agents worked while enhancing it with intelligent suggestions.

During this phase, another insight emerged: while we initially focused on improving the case investigation step for agents, similar friction existed for end customers trying to solve issues on self-service portals. To address this, I expanded the problem definition to consider opportunities for early resolution before a case is even logged. This included experimenting with ways to integrate AI-generated answers directly into customer-facing search and ticket submission flows—potentially deflecting cases altogether.

By the end of Define, we had a clear and expanded problem statement, aligned stakeholders, and a refined strategy that included internal and external user touchpoints.

Unlike generic enterprise search tools, this challenge demanded a Salesforce-native solution—deeply integrated into how Admins already work. That understanding shaped our direction.

By the end of this phase, we had a solid grasp of the problem, a shared vocabulary across teams, and a strong foundation for ideation.

Ideate

Design began with a core question: How can we support Admins in their flow of work without slowing them down? I grounded my interaction design in real user moments, using a "before → during → after" framework to evaluate needs across the configuration journey. After mapping detailed interaction flows for the two core paths—org-wide index and profile-based configurations—I created a JTBD chart divided into before, during, and after phases, outlining the specific information, interactions, and experiences Admins require at each stage. This chart became a central tool during ideation, helping me focus screen design efforts on solving real user needs.

I translated these JTBD insights into low-fidelity screens using rapid iterations in Figma, creating tailored interfaces for each phase:

• Before Configuration: Screens that provide visibility into the current index and its size, helping Admins prepare effectively.
• During Configuration: Interfaces to locate standard and custom objects, toggle field searchability, and confirm or save changes with clear UI signals.
• After Configuration: Feedback screens communicating success, errors, or processing status to close the interaction loop.

Using FigJam, I captured Admin needs at every step to ensure ideation remained aligned with real pain points. Leveraging SLDS components and Figma libraries, I focused on functionality and avoided UI distractions early on.
This JTBD-driven, phase-based approach helped us create an Admin experience that reduced friction and optimized workflows. Early testing refined these ideas before progressing to a closed pilot in line with Salesforce’s internal process, with learnings deferred to post-GA.

Design & Validate

FAfter several rounds of ideation and feedback, I developed two distinct design directions:

• One followed a guided step-by-step approach aimed at clarity and ease for newer or occasional users.
• The other emphasized speed and flexibility, optimized for experienced Admins performing repeat configurations.

We shared high-fidelity SLDS-based prototypes of both versions with seasoned Salesforce Admins in moderated usability sessions. Participants completed tasks like reviewing index health and configuring search for user profiles.

While the guided version reduced cognitive load, most Admins preferred the faster, streamlined approach—despite requiring more upfront understanding—because it involved fewer clicks and better suited their routine workflows. This highlighted an important insight: personas doing repeat tasks often favor efficiency over onboarding ease, even if there's a slight friction the first time.

We moved forward with the speed-first version and shared the selected design with the broader engineering team. Key technical stakeholders—including platform architects and the UI development team—had been involved throughout the process. As backend spiking began to explore feasibility, I worked closely with engineers to iterate based on platform constraints, component availability, and performance considerations.

In parallel, I hosted office hours with the SLDS and Accessibility teams to ensure everything aligned with Salesforce’s design standards. While Slack channels were helpful for quick validation, incorporating formal design reviews with complete screens into the process proved more effective—especially in streamlining collaboration with both backend and UI development teams.

We also ran a closed internal pilot, gathering usage data and qualitative feedback. The results reinforced the need for clearer messaging around system limitations and inspired enhancements like inline help and smarter notifications.

We also ran a closed internal pilot, collecting usage data and qualitative feedback. This validated the need for transparent messaging around system limitations and led to further improvements like inline help and more effective notifications.

North Star

We’re evolving Search Manager into a centralized, intuitive hub where Admins can control everything from query logic to result presentation. Our strategy focuses on:

• AI-Powered Dashboard
  Surfaces proactive insights like relevance gaps and indexing issues—so Admins can optimize search without manual effort.
• Contextual Guidance

  Add contextual guidance with text, images and preferably videos.

• Control Over Logic & Presentation

  Lets Admins define both what shows up and how it’s displayed, ensuring results match business goals and user needs.

Impact

Since its general availability, Search Manager has been adopted by over 10,000 Salesforce organizations—and the number continues to grow. It has become a foundational tool in Einstein Search’s AI-driven evolution, enabling Admins to reduce irrelevant results, fine-tune relevance, and personalize the search experience across user profiles—all without code.

By centralizing search configuration, Search Manager has boosted Admin confidence and accelerated adoption of Einstein Search across orgs, supporting faster, smarter, and more scalable workflows. Its widespread usage reflects the increasing demand for proactive, insight-driven configuration in modern enterprise search.

Future Vision: Scaling Through Configurable Modular Design

The initial release of Search Manager successfully addressed the immediate challenge of giving admins greater control over search data configuration. As an ongoing product, Search Manager will continue evolving in that space — for example, subsequent releases have already introduced rules, filter settings, and sorting configuration to refine how search results are managed.

But my vision extends beyond data-related configuration into configuring the entire search experience. I see Search Manager becoming a one-stop shop for all aspects of search, where admins can assemble their own search interfaces from modular building blocks — search bar, results layout, filters, and other interactive components. These could be added, removed, rearranged, or broken into smaller subcomponents to create highly tailored, context-aware search workflows.

This modular configuration approach would:

• Scale the product across industries, regions, and use cases.
• Unlock new SaaS business opportunities through advanced configuration tiers.
• Empower users to create and adapt search experiences without custom development.

The diagrams illustrate the approach:

• Components and Patterns: Core Einstein Search elements combine into scalable, reusable patterns.
• Fish-to-Shark Analogy: Rearranging the same components results in entirely new, purposeful solutions — just as admins could adapt search to their unique needs.

By expanding configuration from “managing search data” to “designing search experiences,” Search Manager can evolve from a specialized admin tool into a flexible, future-proof platform capability that aligns business growth with user empowerment.