CUSTOMER SUCCESS SCORECARD · SALESFORCE

Turning customer health data into confident business decisions.

Turning customer health data into confident business decisions.

I developed the multi-method research strategy and partnered with a team of researchers to shape the next generation of Salesforce’s Customer Success Scorecard—connecting customer evidence to product adoption, retention, and critical product decisions.

ROLE — Research Strategy Lead & UX Researcher

SCOPE — Enterprise UX · Customer Success · AI

RESEARCH — Interviews · Surveys · Heuristic Evaluation · Usability Testing · A/B Testing

Some product visuals and details have been simplified or adapted to protect confidential information.

Customer Success Scorecard case study visual

THE CHALLENGE

A health score isn’t useful if customers don’t know what to do with it.

A health score isn’t useful if customers don’t know what to do with it.

Salesforce created the Customer Success Scorecard to help customers understand the health of their implementation and identify opportunities to increase adoption and value. But the experience could tell customers that something had changed without consistently explaining why it changed, what mattered most, or what they should do next. Following an accelerated initial release with significant executive visibility, the next iteration needed to evolve from a reporting experience into a more credible and actionable decision-support tool.

ADOPTION
Help customers identify opportunities to use more Salesforce capabilities.

RETENTION
Make customer-health information useful enough to support long-term customer success.

ACTIONABILITY
Help customers understand what requires attention and what to do next.

MY ROLE

My responsibility extended beyond conducting the studies.

My responsibility extended beyond conducting the studies.

I identified the critical questions the organization needed to answer and developed a multi-method research strategy to address them. I coordinated research activities across a team of researchers, drafted interview guides, surveys, and usability studies, oversaw the heuristic evaluation, contributed directly to research execution, and helped lead synthesis across the program. I translated evidence across methods into prioritized recommendations for Product, Design, Engineering, Sales, and Strategy—connecting individual usability findings to broader questions about adoption, retention, customer value, and product direction.

DEFINE
Identify critical questions

PLAN
Build research strategy

COORDINATE
Align research activities

SYNTHESIZE
Connect evidence

INFLUENCE
Drive product decisions

WHAT WE NEEDED TO LEARN

Before choosing methods, I defined the questions that mattered.

Before choosing methods, I defined the questions that mattered.

01 — UNDERSTANDING
How do customers interpret the Scorecard and understand what is affecting their customer health?

02 — ACTION
Does the experience give customers enough context to determine what they should do next?

03 — DISCOVERABILITY
Can customers find and understand new capabilities, recommendations, alerts, and beta features?

04 — BUSINESS VALUE
Can the experience connect Salesforce product usage to the business outcomes customers actually care about?

RESEARCH STRATEGY

One product. Multiple questions. A coordinated research strategy.

One product. Multiple questions. A coordinated research strategy.

No single method could answer the range of questions facing the product. I structured the research program so that each method reduced a different type of uncertainty, while findings could be synthesized across the broader experience.

DISCOVER
Interviews + Journey Mapping
Understand workflows, expectations, and trust

EVALUATE
Heuristic Evaluation
Identify systemic usability risks

VALIDATE
Usability + A/B Testing
Evaluate emerging product directions

SYNTHESIZE
Cross-study Analysis
Connect evidence across methods

PRIORITIZE
Recommendations
Translate evidence into product direction

Throughout the program, I partnered biweekly with Product Managers, Engineers, Sales Executives, Researchers, and Strategists to align evolving questions, communicate emerging evidence, and connect findings to product and business priorities.

DISCOVERY

The Scorecard supported different decisions across the customer lifecycle.

The Scorecard supported different decisions across the customer lifecycle.

Discovery research showed that the Scorecard wasn’t serving a single workflow. Customer Success, Renewal, and Sales teams approached customer-health information with different responsibilities and different decisions to make.

CUSTOMER SUCCESS MANAGER

Monitor customer health
Investigate technical issues
Build customer success narratives
Add qualitative context to the data

RENEWAL MANAGER

Forecast attrition risk
Audit account footprint
Model renewal scenarios
Identify early warning signals

ACCOUNT EXECUTIVE

Identify expansion opportunities
Build commercial strategy
Track customer sentiment
Identify unused capabilities

THE IMPLICATION

The Scorecard couldn’t simply report customer health. It needed to support different decisions across customer success, renewal, and expansion workflows.

The Scorecard couldn’t simply report customer health. It needed to support different decisions across customer success, renewal, and expansion workflows.

A FOUNDATIONAL INSIGHT

Before customers could act on the data, they had to trust it.

Before customers could act on the data, they had to trust it.

“The biggest pain point is severe data distrust. I manually verify data outside of CSS before every client meeting because a ‘0’ could just be a tracking error.”

“The biggest pain point is severe data distrust. I manually verify data outside of CSS before every client meeting because a ‘0’ could just be a tracking error.”

Senior Customer Success Manager · Discovery Participant

TRUST WAS THE FOUNDATIONAL PROBLEM

When users couldn’t confidently explain where a score came from, they created manual workarounds and verified information outside the Scorecard. This limited the product’s ability to function as a credible customer-facing decision tool.

RESEARCH EVIDENCE — USAGE PATTERNS

EVALUATION · HEURISTIC ANALYSIS

Complex terminology made an already complex system harder to trust.

Complex terminology made an already complex system harder to trust.

The heuristic evaluation reinforced patterns emerging from discovery research. Important signals were often presented using specialized terminology without enough contextual explanation, making it difficult for users to confidently interpret what the data meant.

WHAT WE OBSERVED

Signal names within the dashboard were often complex and lacked immediate contextual definitions. Users could not always determine what a metric represented or why it mattered without seeking information elsewhere.

RECOMMENDATION

Introduce contextual definitions or lightweight explanations for complex signals and use familiar, plain-language terminology where possible.

WHY IT MATTERED

The Scorecard was intended to help customers understand their Salesforce health. Requiring users to leave the experience to interpret its metrics increased cognitive load and made it harder for the product to function as a trusted decision-support tool.

FROM USABILITY ISSUE TO PRODUCT QUESTION

The problem wasn’t simply whether users understood individual labels. It raised a broader question: when should Salesforce-specific terminology be preserved, and when would familiar customer language create a clearer experience?

The problem wasn’t simply whether users understood individual labels. It raised a broader question: when should Salesforce-specific terminology be preserved, and when would familiar customer language create a clearer experience?

This finding contributed to broader cross-functional conversations about branded terminology versus common language across the experience.

VALIDATION · A/B TESTING

Research turned three competing design directions into a clear product decision.

Research turned three competing design directions into a clear product decision.

Additional Feature Recommendations were intended to help customers discover Salesforce capabilities that could improve adoption and business outcomes. The team had three competing approaches for presenting those recommendations, and we needed evidence about which experience made the next step clearest and most actionable.

THE QUESTION

How should additional feature recommendations be surfaced without overwhelming the customer or disconnecting them from the context of the Scorecard?

How should additional feature recommendations be surfaced without overwhelming the customer or disconnecting them from the context of the Scorecard?

A — SIDE PANEL

Preserved visibility of the underlying Scorecard while presenting recommendations alongside the existing experience.

Mixed externally · Rejected internally

Internal participants found the recommendation experience contextually disconnected despite preserving the underlying dashboard.

PREFERRED DIRECTION

B — MODAL WINDOW

Centered recommendations in a focused experience while temporarily reducing competing interface information.

85.7% internal preference
45.5% external preference

Participants valued the modal’s prominent placement and ability to focus attention on recommended actions.

C — EMBEDDED LIST

Placed recommendation content directly within the existing Scorecard interface.

Lower preference

The embedded approach introduced additional information into an already dense interface and increased concerns about cognitive load and clutter.

Internal users showed a clear preference for the focused modal experience.

Internal users showed a clear preference for the focused modal experience.

85.7%
Internal participants preferred Option B

85.7%
Internal participants preferred Option B

45.5%
External participants preferred Option B

45.5%
External participants preferred Option B

Internal users showed strong consensus around the modal. External responses were more distributed, reinforcing the value of testing beyond audiences already familiar with Salesforce conventions.

DECISION

Move forward with the modal approach for Additional Feature Recommendations.

Move forward with the modal approach for Additional Feature Recommendations.

The centralized modal created a clearer moment of focus for recommendations and made the next action easier to understand without permanently adding more information to an already complex dashboard.

WHY IT MATTERED

This study reduced uncertainty around how recommendations should enter the customer workflow and gave the team evidence to move forward with a specific interaction direction.

QUESTION → EVIDENCE → DECISION

FROM EVIDENCE TO PRODUCT DIRECTION

The research didn’t just identify problems. It changed product direction.

The research didn’t just identify problems. It changed product direction.

Evidence across discovery, evaluation, and validation was synthesized into recommendations that shaped decisions around customer value, reporting, information architecture, recommendation design, and AI interaction.

FIVE DECISIONS INFLUENCED BY RESEARCH

01 · BUSINESS VALUE

From product adoption to business objectives

EVIDENCE
Users understood the Scorecard’s value more clearly when Salesforce capabilities connected to business outcomes.

DECISION
Reorient the experience around customer business objectives.

WHY IT MATTERED
Connected Salesforce usage to the outcomes customers were actually trying to achieve.

Business objectives research visual

MONTHLY → WEEKLY
+ HISTORICAL CONTEXT

MONTHLY → WEEKLY
+ HISTORICAL CONTEXT

02 · REPORTING CADENCE

From monthly snapshots to more timely signals

EVIDENCE
Users valued more current indicators while research also surfaced the risk of overreacting to short-term fluctuations.

DECISION
Move toward weekly change tracking while preserving longer-term historical context.

WHY IT MATTERED
Provided earlier signals without forcing customers to interpret a single week in isolation.

03 · RECOMMENDATION DISCOVERABILITY

Give recommendations a clear moment of focus

EVIDENCE
Comparative testing showed the strongest internal preference for the centralized modal experience.

DECISION
Move forward with the modal direction for Additional Feature Recommendations.

WHY IT MATTERED
Made next steps more prominent without permanently adding information to an already dense dashboard.

A — SIDE PANEL B — MODAL ✓ C — EMBEDDED
See validation study above ↑

04 · INFORMATION ARCHITECTURE

Make the connection between features and business outcomes explicit

PRODUCT

FEATURE

BUSINESS OBJECTIVE

EVIDENCE
Users struggled to understand how individual Salesforce capabilities related to broader customer objectives.

DECISION
PRODUCT → FEATURE → BUSINESS OBJECTIVE

WHY IT MATTERED
Created clearer technical and commercial context for customer conversations.

05 · AI INTERACTION

Centralize Agentforce instead of repeating AI entry points

BEFORE
Feature → Explore
Feature → Explore
Feature → Explore
Feature → Explore

AFTER
Features

ASK AGENTFORCE

EVIDENCE
Repeated Explore Feature actions created visual clutter and mismatched expectations about AI interaction.

DECISION
Centralize conversational AI within a dedicated Ask Agentforce experience.

WHY IT MATTERED
Reduced button fatigue and clarified expectations for AI-assisted guidance.