Walmart — Deal Risk Assessment

Account / Deal

Walmart — Walmart Opal 2026

Owner

Rachel Gray (AE) · Conor Leary (overlay)

Stage / Target Close

Discovery TBD

Size / Strategic Importance

TBD · High (Fortune 1, enterprise CMS, multi-agent vision)

Key Champions

Viknesh Salivahanan (lead PM, strong pull) · Samarth Sharma (CMS platform lead, champion-adjacent)

Economic Buyer

Not identified

Competition

In-house build (fine-tuned open-source models, multi-agent router)

Current Status (as of 2026-03-16)

Two meetings completed: discovery (03-10), demo/SVC with Salesforce case study (03-14). Two new engineering stakeholders joined Session 1: Suresh Rao (architect) and Karthik Cherukuri (engineering director). Samarth admitted "false starts" on internal builds — strongest buying signal. Critical unanswered question: can external agents access full Opal chat history? Next step: deep technical dive with Imran and Walmart engineering (90 min preferred). Follow-up email with architecture request pending.

ID	Category	Risk	Evidence	Impact	Owner & Mitigation
R1	Political	No identified economic buyer or executive sponsor Known	Viknesh and Samarth are evaluators, not budget holders. "Proof of value for leadership" — leadership undefined. No exec-to-exec engagement path exists.	Cannot advance past discovery without knowing who holds the budget. Deal stalls if exec sponsor is not surfaced during or after technical validation.	Rachel / Conor — surface through Samarth's POV and next technical session. Ask: "Who makes the funding decision for platform investments like this?"
R2	Technical	Build-vs-buy tension — Walmart building agentic capabilities internally Known	Fine-tuning open-source models within existing ecosystem. Multi-agent router system already built. Samarth admitted "false starts" — internal timeline not meeting GTM goals. Samarth: "we know where the roadblocks are, where we actually need help."	Could become "we'll build it ourselves" objection. Mitigated by "false starts" admission — the opening is acceleration, not replacement.	Conor — position Opal as acceleration partner. Request specifics on internal build gaps during POV exchange. Samarth already offered to share.
R3	Technical	Shared memory / chat history question left unanswered Known	Suresh asked if external agents can access full Opal chat history. James acknowledged he didn't know the answer. Architecture requires shared memory — each agent summarizes context differently from same history. Users must not repeat context when transitioning between workflow stages.	Make-or-break requirement. If Opal cannot support shared context across agents, the deal cannot progress. Credibility gap from unanswered question at Session 1.	James Stout — get definitive answer before Imran technical deep dive. Non-negotiable. If answer is no, prepare alternative architecture approach.
R4	Technical	"Yet another tool" risk — Opal must be embeddable, not standalone Known	Viknesh: 2026 workstream to reduce from 20 tools to unified experience. Expects agents to bridge integrations. "Play to your strength" — they want proven embeddable capabilities.	Opal positioned as another standalone UI will be rejected. Must lead with headless/API-first composable positioning.	Conor / James — lead with embeddability at next session. Demo API-first composability, Agentspace integration, CMS-agnostic publishing.
R5	Timing	Scheduling momentum loss — next session deferred to email Known	Despite on-call attempt to schedule, deferred to email. Momentum lost once before during Matt → Rachel handoff. Samarth offered "all the time you need" — willingness exists.	Each day of delay reduces momentum from strong Session 1 signals. Risk of evaluation going dormant.	Rachel — follow up within 48 hours. Conor — send architecture request email as forcing function.
R6	Technical	Technical depth exceeded team's ability to answer Known	Questions on shared context across multi-session workflows, observability schema consistency, and federated context management require engineering depth. Session shifted from planned broad walkthrough to deeply technical conversation.	Credibility risk if next session cannot address technical depth. Imran's participation is non-negotiable.	Conor — loop in Imran and brief thoroughly on Walmart's specific requirements before next session.
R7	Value	No budget or deal value — purely exploratory Known	No ROI, budget, or timeline discussion in either meeting. Deal value field is blank. Unknown whether this is a funded initiative or unfunded exploration.	Cannot size the deal or forecast. Appropriate for discovery stage but must be addressed post-technical validation.	Conor — introduce value/ROI framing once technical fit is confirmed. Scale (hundreds of operators, daily content changes) provides strong efficiency story.
R8	Paper	Unknown procurement / paper process Known	No discussion of legal, security, architecture review, or procurement. Walmart is Fortune 1 enterprise — significant gates expected.	Enterprise procurement at this scale could add months. Zero visibility into process or timeline. Must begin discovery once technical validation is positive.	Rachel — begin parallel workstream discovery post-technical validation. Ask about prior vendor onboarding experience.
R9	Value	"Play to your strength" credibility risk — overpromising on agent maturity Known	Samarth: "play to your strength" — wants proven capabilities, not roadmap. Walmart is technically sophisticated; vaporware positioning will be detected immediately.	Credibility loss if we overstate current capabilities. Trust erosion would kill the deal.	Conor / Imran — be precise about what Opal does today vs. roadmap. Lead with production-ready capabilities. Let Imran speak to technical depth credibly.

Metrics

Red — No metrics identified

No budget, ROI, or quantified value discussion. Action: Introduce value framing post-technical validation. Use scale (hundreds of operators, daily changes) as efficiency multiplier.

Economic Buyer

Red — Not identified

"Leadership" referenced but no name, title, or reporting chain. Action: Ask Samarth directly during POV exchange or next session.

Decision Criteria

Yellow — Technical criteria clear, political unknown

Four criteria surfaced: shared context/memory, embeddability, observability, vertical scalability. Action: Address at Imran session. Surface political/commercial criteria as exec engagement develops.

Decision Process

Yellow — Next step clear, post-validation process unknown

Two-phase evaluation: broad case study (done), technical deep dive (next). Action: Ask after next session: "If this validates technically, what does your internal process look like from here?"

Paper Process

Red — Complete gap

No discussion of legal, security, procurement, or vendor approval. Action: Begin discovery once technical validation is positive.

Pain

Green — Well validated

Manual content lifecycle at scale, no decision support, no feedback loops, pre-execution bottlenecks. Validated across both meetings.

Champion

Yellow — Strong pull, not yet validated

Viknesh: initiated contact, wants speed. Samarth: escalating to deal shaper, "false starts" admission. Gap: Neither has demonstrated ability to sell internally to leadership. True champion test: do they coach us on how to win internally?

Competition

Yellow — Primary threat is internal build

In-house build with fine-tuned open-source models and multi-agent router. Writer failed at Salesforce (contrast point). Gap: Don't know full scope of internal investment, team size, or exec commitment to build.

Loop in Imran and brief on Walmart requirements — shared memory, multi-session context, observability schema, vertical scalability, embeddability.
Owner: Conor · When: immediately after scheduling confirmed
Determine existing Optimizely footprint at Walmart — smaller business unit uses testing platform. Co-term or expansion opportunity may exist.
Owner: Rachel · When: this week
Prepare Opal positioning that leads with production-ready capabilities — agent builder, workflows, observability, instructions system. No roadmap promises.
Owner: Conor · When: before technical deep dive
Encourage Walmart to share their POV on challenges — Samarth offered. Builds partnership framing and reveals internal decision dynamics. Also surfaces internal build scope and gaps.
Owner: Conor (in follow-up email) · When: 2026-03-17
Prepare Gemini Enterprise / Agentspace integration talking points — Opal as first commercial tool on Agentspace. Walmart uses Gemini Enterprise and this was well-received.
Owner: Conor / James · When: before technical deep dive