UX Case Study · tenXengage

The Workflow That Was Losing Enterprise Deals

A broken promotion-creation workflow was causing enterprise prospects to disengage at demo. As the sole client-facing UX leader, I conducted ground field research with partners, diagnosed the cognitive load gap in the "AI" flow, and designed the structured UX architecture and wireframes that Cisco and Google now use to run partner programs at scale.

B2B SaaS AI + Manual Workflow User Research Conflict Resolution Cisco · Google Wireframing

My Role Sole Client-Facing UX Leader
Led field research, friction diagnostics, and system UX architecture.

My Ownership Strategy & Wireframes
Designed and validated structural blueprints with Cisco & Google partners.

Team Hand-off Visual Design & Code
Final UI styling and front-end coding delivered by tenXengage product team.

Business Impact

The numbers tenXengage leads with

40–50%

Partner engagement increase

20–25%

Partner-led bookings growth

50%

Reduction in support costs

99.5%

Partner support SLAs met

How UX drove the 40–50% Engagement Lift:

By replacing the opaque "black-box" conversational setup flow with a dual-mode workflow (side-by-side parameters and live visualization), we eliminated visual hesitation. Partner admins could validate their setup criteria dynamically, reducing abandonment rates mid-flow.

How UX drove the 20–25% Bookings Growth:

Training partner sellers became effortless because the interface was framed as a guided conversational engagement builder. Since the UI demanded minimal technical knowledge, sales reps could instantly launch campaign promotions with zero configuration learning curves.

How UX drove the 50% Support Cost Reduction:

The original conversational UI was an ambiguous black box, forcing partner administrators to constantly submit support tickets just to understand how the system worked. The new self-explanatory layout contextually "spoon-fed" parameters, cutting out "how-to" and system-understanding tickets.

Source: tenXengage Customer Success Metrics · Cisco & Google deployments

The Problem

An endless form disguised as a conversation.

The old promotion builder created an illusion of simplicity — a Q&A panel on the right, a wizard on the left. But the right panel was never real AI. It was a passive progress log — users still had to fill every field manually, screen by screen, with no sense of how many steps remained. The result was a form that felt infinite. Senior channel executives at Cisco and VMware abandoned mid-flow. Prospects disengaged at the first demo screen.

⬤ Before — The illusion of a conversation

Promotion setup — question-by-question wizard with Back/Next steps

Wizard + passive Q&A sidebar The right panel tracked answers but was never actionable — users still filled every field manually. The "conversation" was just a log.

Q&A sidebar showing cascading questions — region, country, company, certification

No end in sight Region → Country → Company → Certification → Partners. Every answer revealed another question. Users had no map of what remained.

File upload step alongside Q&A — cognitive overload

Mixed tasks, broken flow File uploads, acronym fields, Q&A — unstructured and unrelated in sequence. 98% shown but the user still had no certainty of what was left.

No progress visibility No ability to jump between steps Questions without context Mixed task types in one flow Demo drop-off at first screen

The Internal Conflict

Backend predictive power vs. user cognitive load

The system was promoted as having a state-of-the-art backend AI prediction engine. However, the business focus was entirely on optimizing the backend calculations rather than reducing the user's cognitive load during campaign creation. The product team wanted to hide the complexity behind an automated chat interface. My research proved that users needed structured system architecture, clear logic flow, and full visibility over financial configurations to trust the AI's predictions.

Product Team Demand

Focus on Backend AI & Chat Automation

Hiding the campaign configuration behind a conversational UI to showcase backend predictive capabilities.

UX Architecture Position

Structured Alignment + Real-time Guidance

Reducing cognitive load by presenting structured visual logic, letting users control configurations, and using AI for real-time validation and forecast previews.

Research

Ground research with partner admins and sellers

Daily field calls

Ongoing sessions with partner managers, admins, and sellers — not a one-time interview pass.

Demo video behavioural review

Non-verbal signals at the Q&A flow — where users closed off — more honest than any survey answer.

Survey — Cisco partner admins

Targeted the exact users in dispute. Focused on AI trust in financial decision-making.

A/B test: AI-only vs. dual-mode

Data settled what debate couldn't.

Wireframe testing with real users

With Cisco partner admins and Google channel managers — before any visual design.

The A/B result confirmed the cognitive load gap

The testing validated the friction diagnosis: senior partners didn't reject the AI backend; they rejected the lack of transparent structure. Handing over financial parameters to a black-box conversational UI created massive hesitation. When we provided a dual-mode workflow (structured manual tracking and guided AI assistance), partner engagement surged. The key was aligning backend AI prediction capabilities with real-time user validation.

The Solution

Structure, context, and real AI assistance

The endless Q&A was replaced with a 5-step accordion — all steps visible upfront, jump to any section, full progress clarity. Critically, the AI Copilot is now genuinely actionable: it guides the user through intent, pre-fills fields, and generates live ROI forecasts — not a passive log of answers. Manual Mode preserves full user control for executives who need it. Both paths converge at the same AI Prediction screen before any commitment.

⬤ After — Structured flow with real AI assistance

	Before — Wizard + Passive Log	After — Structured + Real AI
Flow structure	One screen per question, no overview	5-step accordion, all steps visible upfront
Right-panel role	Passive log — tracked answers, not actionable	Real AI Copilot — guides, pre-fills, forecasts
Progress visibility	None — users couldn't see the finish line	Always visible, % complete at every step
User control	Forced linear sequence, no skipping	AI Mode or Manual Mode — switch anytime
ROI visibility	Only visible after full completion	AI Prediction before any dollar committed
Demo conversion	Drop-off at first screen in live demos	Structured entry builds immediate confidence

New dual-mode — AI Copilot alongside 5-step accordion

AI Mode + Manual toggle · All 5 steps visible · Jump to any section

AI Forecasting · 3.5x ROI predicted · Live parameter sandbox

Budget Information · Step 4 of 5 · Progress always visible

Approval Process · Step 5 of 5 · 100% complete

Outcome

A decision backed by data — adopted at enterprise scale.

The restructured flow — validated by Cisco partner admin survey data and A/B testing — resolved the internal conflict and directly contributed to the platform outcomes tenXengage now leads with in every enterprise conversation. Partners completed incentive creation. Prospects converted from demo. The workflow stopped losing deals.

Pratap and his team have been instrumental in creating a predictable, positive ROI for our Seller Rewards Program through real-time rewards, simple to use platform, and automation of forecasting and reporting.

Senior Partner Strategy Leader, Cisco

"Simple to use platform" — a Cisco senior partner leader, describing a platform that manages incentive programs for 25,000+ partners.

What I Learned

Name the real problem before solving it

The surface problem was a bad workflow. The real problem I diagnosed was that the AI was highly optimized for business predictions but completely failed to reduce cognitive load for the user. Once we restructured the UX architecture around user validation and wireframes, we aligned the product with real partner behavior. The A/B test and Cisco partner survey didn't create the insight — they confirmed it with enough evidence to move the business.

A note on confidentiality & project scope

The UX methodology, strategic research, and core informational architecture are documented here to show my design process. Direct business metrics (40—50% engagement lift) are sourced from public customer decks, and client names (Cisco, Google) reflect marketing releases.

All interface screens represented in this case study are early UX architecture wireframes I designed for stakeholder validation and testing.
The final high-fidelity visual design and front-end development were implemented by tenXengage's internal product design and engineering teams.
No proprietary backend logic, unreleased platform capabilities, or contract specifics are disclosed.

Happy to discuss the full scope of disclosure in any interview.