Internal QA appendix · not for client distribution

What the deck says that the website doesn't — and every flag on the site.

A line-by-line comparison of the website proposal against the current source-of-truth deck, PSL RFP 2026 working - Updated MASTER.pptx (67 slides + speaker notes). This master supersedes the 54-slide working deck the original audit ran against; the site and this appendix were re-synced to it on 7/29. Two outputs: deck content that never made it onto the site (below), and an index of the 21 claim flags placed throughout the site — red for fabrication risk, yellow for unverified or embellished.

Re-compared 7/29/2026 · deck: "PSL RFP 2026 working - Updated MASTER.pptx" (67 slides + notes) · site: proposal v3 · flags: 2 red / 19 yellow

← Back to the proposal

Part 1 · In the deck, not on the website

Substantive content in the updated master that is absent from the site. The updated deck is far more complete than its predecessor — it closed most of the old gap list (disclosure, references, bios, pricing, a named PR case study), and the site has now absorbed those answers. What remains is below.

Slide 65 · Because We're Already Partners

The itemized "already partners" evidence lists

The deck's strongest strategic slide itemizes what the current engagement has produced: learned — website, communities, Yardi, Further, investor groups; already improved — reporting, attribution, integrations; next — AEO, PR, AI, storytelling, automation. The site carries the narrative ("Why now, not before") but not these lists — arguably the most persuasive evidence in the deck.

Slide 54 · Phoenix in 2028

The flywheel framing

A family asks ChatGPT → AI recommends Phoenix → website answers → tour booked → qualified lead in Yardi → move-in → resident story → AI cites Phoenix again: "every resident story becomes tomorrow's AI citation." The site makes each piece of this argument separately but never draws the loop as one picture.

Slide 60 · Resource model

The headcount-level team structure

The deck commits to a specific resource model: a core creative team (designer, copywriter, production, creative director, QA, traffic specialist, senior web developer, web coordinator), a four-person named account team, a media team (senior planner, paid, programmatic, analyst) and extended specialists that scale as needed. The site describes the bench generically.

Slide 64 · References

Three additional references outside senior living

Beyond the three senior-living references now on the site, the deck carries full contact detail for Wasteology (Diane Gahafer, CRO), Fat Brands / Fazoli's (Stephanie Holthaus, Brand Marketing Manager) and Lexington Co-Op / Arby's. The site names them without contact detail. Note: the reference slide's layout made some name/email pairings ambiguous in extraction — have Wingspan confirm the pairings before anything client-facing uses them.

Slide 5 · Proven AI Innovation

What the deck actually claims about AI and agents

The deck's supported AI claim is: "project management tools that utilize agentic workflows and agents" (e.g. proofreading agents in the PM tool) and investment in "proprietary AI tools" — with the reporting AI agent framed as "the future of reporting." Relevant context for the red flag below: the deck deliberately stops short of claiming a current agentic reporting beta.

Consistency check · Re-passed against the updated master

The site's "3 finalizing" list matches the updated deck's remaining gaps

The updated deck answers Q27/28, Q32–34 (including the Vivid Impact disclosure), Q43/44/46 (with the "Sing Us a Song" case study), Q50, Q54, Q56 (named team) and Q62 (references) — and the site now carries those answers with ANSWERED status. Still genuinely open: Q2 (firm financial detail, under separate cover) and Q51 (a written case of a piloted technology and its outcome). The site's status tracker (59 answered / 3 finalizing) reflects exactly that.

Part 2 · Flag index

Every flag placed on the proposal, in page order. ⚑ RED = fabrication risk: presents invented data as fact or contradicts the deck's deliberate posture. ⚑ YELLOW = unverified or embellished: not supported by the updated master; verify or soften before submission. Hover any flag on the site for this same context in place.

LocationClaimIssue
Hero"All sixty-two RFP questions answered in place"Site's own status list says 59 answered, 3 finalizing.
Ch 1 · Q2"Debt-conservative"Financial characterization not in deck.
Ch 1 · Q3Named partner examples (production crews, PR networks, Further)Deck never specifies which services are partner-delivered.
Ch 1 · Q47"Agency reviews run on three-year cycles"Unsourced industry statistic.
Ch 1 · Q47"Financial services and higher education" verticalsAdded beyond the deck's four verticals.
Ch 1 · Q48Both client listsNot in deck text; confirm names + logo clearance.
Ch 1 · Q49"Almost nobody in this category is measuring it yet"Sweeping competitive claim, unsupported.
Ch 2 · governance$127K+/mo gross spend, 65+ active campaignsWas in the old deck; REMOVED from the updated master. Reconfirm.
Ch 3 · schema demoThe Retreat at Lady's Island — 4.8★, 126 reviewsInvented example data for a real community.
Ch 3 · Q32 band"Every platform here is one we operate in production today"Updated deck names only Google + Meta certifications.
Ch 3 · Q23Mailchimp and Klaviyo implementationsPlatforms added beyond deck's "HubSpot, Salesforce, Yardi and more."
Ch 3 · Q30Wiki field detail (credential custody, rollback…)Deck now has wiki + test cadence; custody/rollback still site-side.
Ch 3 · Q31"Provisioned in PSL's name from day one"Commitment the deck never makes.
Ch 5 · resultsPSL: 8,766 conversions / 11% / $2,500Case-study numbers, not in deck — PSL will check them.
Ch 5 · resultsSt. George Village: 910 leads vs 22, $17 CPLCase-study numbers, not in deck.
Ch 5 · resultsMorning Pointe / The Blake / Wickshire figuresCase-study numbers, not in deck.
Ch 5 · Q22/24Synthetic submissions + weekly Yardi reconciliationQA regimen not described in deck.
Ch 6 · Q27GBP management "across all 55 communities"Deck claims "at scale," not full-portfolio coverage today.
Ch 7 · Q52/53"Actively developing and beta-testing an agentic layer"Updated deck walked this back to "the future of reporting" — a vision, not a current beta.
Ch 9 · Q58/5930-day account transfer, per-market cutover, weekly statusUpdated deck's Days 1–30 are discovery + audit; no transfer commitments.
Ch 9 · Q36/376/12-month outcome projectionsUpdated deck (slide 57) explicitly sets benchmarks before engagement instead of predicting outcomes.

Part 3 · Resolved since the original audit

Flags removed on 7/29 because the updated master now supports the claim — kept here so the trail is auditable.

Removed flags · 5

Further/Yardi integration scoping (was red) — the updated deck affirms built integrations with Further, including the Google Offline Conversions sync. · HubSpot MQL test tense — MQL/SQL scoring is now presented as the working system and is included in the monthly fee. · "Disclosed in writing" — the deck now contains the written Vivid Impact disclosure. · Premium-storytelling experience — the deck now names "Sing Us a Song" with results. · Physician-template campaign framework — the framework slide was rebuilt for PSL; no OBGYN/physician remnants remain.

Verified against the updated master rather than flagged: the 10 sub-brands and 55 communities, 350 combined years, ~500 communities served, 50 employees, 800+ businesses in 40+ states, seven-year average tenure, the GA/HubSpot/Gumshoe reporting stack, the heat map (correctly labeled illustrative), the 90-day plan, the named account team, the fee structure ($1,500/community/month; $2,000 + $400/yr website builds; 20% media commission), the "Sing Us a Song" results, and the BBC StoryWorks continuity plan. Key dates and Justin Harden as contact were verified against the RFP document itself.

← Back to the proposal