Scaffold the monorepo: pnpm workspaces + Turborepo with packages/tokens, packages/ui, packages/schemas, apps/web

A GitHub monorepo where pnpm install && pnpm turbo build succeeds from clean; four workspaces (tokens, ui, schemas, apps/web) with shared tsconfig and eslint; packages/schemas imported by both Sanity Studio and the Next.js runtime; README documenting the layout. Merged to main.

Stand up the CI skeleton in GitHub Actions: typecheck, lint, test placeholders, Vercel preview deploy per PR

A .github/workflows/ci.yml that runs on every PR: pnpm install with cache, turbo typecheck, lint, test (placeholder passing), and a Vercel preview deployment with the URL posted to the PR. Branch protection requires CI green. Secrets (Vercel token, org, project) stored in GitHub, not in the repo.

Confirm Figma Variables are systematised as the token source: colour, type, spacing, radii

Written confirmation from the designer (branding agency) that the Figma file exposes colour, typography, spacing and radii as Variables (collections and modes), not only as styles or hard-coded values on delivered components; a Figma REST API call returns them; naming convention agreed and documented.

Build the GitHub Action that fetches Figma Variables via REST API, pins the API version, and schema-checks the response

A workflow (tokens-sync.yml) on schedule and manual dispatch that calls the Figma Variables endpoint with a pinned API version, validates the JSON against a Zod or JSON Schema, writes packages/tokens/src/figma.json, and opens a PR if it changed. Fails loudly on schema mismatch.

Style Dictionary transform and publish the wallboard/tokens package as CSS variables + Tailwind config, versioned semver with tagged releases

packages/tokens builds from figma.json via Style Dictionary into tokens.css (CSS custom properties) and tailwind.tokens.js (theme extension); published to the GitHub package registry as @wallboard/tokens with semver and a git tag per release; consumed by packages/ui.

Token snapshot test and lint rule banning literal colours, both as merge blockers

A Vitest snapshot of the generated tokens.css that fails CI on unreviewed changes, and an ESLint rule (or Stylelint) that fails on any hex/rgb/hsl literal in packages/ui and apps/web. Both wired into WB-58 CI as required checks.

Stand up Storybook in packages/ui as the docs and visual-regression home

Storybook 8 running in packages/ui, consuming @wallboard/tokens, deployed to a Vercel preview on every PR, with Chromatic or Playwright screenshots configured as the visual regression baseline.

Build the registered section components against tokens, one Storybook story per section as the merge gate

Every component in the section library built in packages/ui with typed props, no raw values, variants as props: NavBar, Footer, CtaButton, HomeHeaderSection, RoleHeaderSection, FeatureHeaderSection, ContentBlockSection, ThumbnailGridSection, GallerySection, FilterSection, FormSection, TestimonialSection, LogoSection, CtaSection. Each has a Storybook story covering its variants; a PR without a story is blocked.

Vitest + Playwright + visual regression on section variants

Unit tests (Vitest) for section prop handling, Playwright component tests rendering each section variant, screenshot baselines committed, CI fails on visual diff above threshold. Closes the preview-accuracy gap.

Provision Sanity project and Studio; define document types Page, Product, CaseStudy, BlogPost

Sanity project with prod and staging datasets (per Infra #3), Studio deployed (sanity.studio or embedded at /studio in apps/web), four document types defined in packages/schemas with slugs, SEO fields, and the pageBuilder array on Page; CORS and tokens configured; team invited.

Author section schemas mirroring component props; configure page-builder for marketing pages vs typed schemas for product, case study, blog

One Sanity object schema per registered section whose fields mirror the React component props exactly; a dual-update PR template requiring schema and component to change together; a contract test comparing the two; the Page type uses an open pageBuilder array of those sections; Product, CaseStudy and BlogPost use fixed typed fields. One named owner of the contract (gating decision 1).

Wire Sanity Presentation preview to the production components

Studio's Presentation tool renders draft content through apps/web using the same section registry as production, with click-to-edit overlays; no separate preview components.

Build the editorial workflow in Sanity including reviewer-to-editor send-back

A workflow status on documents (Draft, In review, Changes requested, Approved, Published), roles for editor and reviewer, a send-back action with a comment, and a Studio view listing what awaits each role. Documented for the team.

Build the Next.js runtime: App Router, RSC, GROQ queries, section-type to component registry, ISR + on-demand revalidation

apps/web renders any Page document by mapping each pageBuilder block _type to its component via a registry; typed GROQ queries with sanity-typegen; ISR with tag-based revalidation; global navigation and footer from singletons; 404 and error routes; deployed to Vercel preview.

Two-path deploy: code path via Vercel, content path via Sanity webhook to /api/revalidate, with delivery logging, alerts and a manual fallback

Content publishes in Sanity trigger a signed webhook to /api/revalidate that revalidates by tag; every delivery is logged (Vercel logs or a small table) and a failed or missing delivery raises an alert (Sentry or email); a manual revalidation route exists for fallback; schema changes still ship via the code path.

FormSection: HubSpot field shape + server-side Forms v3 POST, with build-time form-definition sync and a contract test

FormSection renders a form from a HubSpot form ID, posts server-side to the Forms v3 API with hutk cookie and page context, handles success/error states accessibly; a build step fetches form definitions from HubSpot and a contract test fails the build if field shapes drift from what the component expects.

Wire observability: Vercel Analytics, Sentry, HubSpot tracking code, with a pre-launch baseline

Vercel Analytics and Speed Insights enabled; Sentry SDK in apps/web with source maps and release tagging in CI; HubSpot tracking script loaded with consent handling; a baseline snapshot of Core Web Vitals and error rate recorded before go-live.

Confirm all guardrails live and set the quarterly section audit

A checklist confirming each guardrail is enforced in CI: dual-update PR template, token lint and snapshot, form contract test, webhook alerts, visual regression, preview parity; plus a recurring quarterly Todoist task to audit the section library for sprawl and drift.

Wire GA4 + HubSpot persona tracking: persona_type population, UTM structure, cross-storyline attribution

GA4 events for selector clicks, storyline views, CTA clicks and form submits with persona dimension; HubSpot tracking code populating persona_type on contacts from the persona context (WB-78); UTM parameter convention documented and enforced in all outbound links; attribution report showing first-touch storyline to conversion.

Implement technical SEO on the runtime: schema.org JSON-LD, semantic HTML, meta and OG tags, sitemap.xml, 301 redirect layer, llms.txt, canonicals

Every page emits JSON-LD (Organization, WebPage, Product, Article as applicable), semantic landmarks, title/description/OG/Twitter meta from Sanity SEO fields, canonical URLs; sitemap.xml generated from Sanity; the redirect layer consumes WB-33's map (next.config redirects or middleware for regex); llms.txt updated. Rich Results Test passes on each template.

Execute DNS cutover, verify site live, test redirects + forms + HubSpot in production, 24-hour monitoring

wallboard.us (and www) pointed at Vercel, SSL issued, old Webflow site left reachable at a fallback host for 30 days, redirect map verified in production by script, a live form submission confirmed in HubSpot, GA4 real-time showing traffic, Sentry quiet, and a 24-hour watch with a rollback plan (DNS back to Webflow) documented.

Build the persona context system: URL param or cookie carrying storyline to platform to demo, with dynamic CTA swap

A persona value set from the homepage selector or a ?p= URL param, persisted in a first-party cookie, read by CtaButton and FormSection to swap CTA copy/destination and prefill the hidden persona_type field; works with RSC (cookie read server-side) and respects consent.

Build /security: scannable deep-dive with jump links, RBAC diagram, encryption, compliance, deployment options

/security page built from FeatureHeaderSection + ContentBlockSection + a jump-link nav, rendering WB-43's copy and RBAC diagram, with a security docs download CTA (WB-48), passing WB-51 accessibility checks.

Load test and Core Web Vitals: LCP under 2.5s, INP under 200ms, CLS under 0.1, concurrent-user capacity, mobile speed

Lighthouse CI in the pipeline with budgets (LCP 2.5s, INP 200ms, CLS 0.1, performance score 90+ on mobile); a k6 or Artillery load test against staging at 10x expected peak concurrent users with p95 TTFB recorded; results in the Notion QA table. (Tracker says FID; INP replaced FID as the Core Web Vital in 2024.)

Build /platform: UI, AI and API as three equal interface modes, capability story not feature list

/platform page built from FeatureHeaderSection and three parallel ContentBlockSections (UI, AI, API) plus links to the five feature pages (WB-93), rendering WB-46 copy, with persona-aware CTA (WB-78).

Build the site chrome: header mega-menu nav, utility bar (Login / Contact / Trial), sticky scroll, mobile hamburger, footer

NavBar and Footer components (chrome, rendered outside the page-builder) driven by Sanity navigation and footer singletons; mega-menu grouped per the signed-off sitemap (WB-12), utility bar with Login (app link), Contact, Trial; sticky with transparent-on-hero variant; accessible mobile menu; footer with columns and newsletter toggle.

Build /demo: detect persona context, self-guided demo with persona scenarios, booking flow, video library

/demo reads the persona cookie (WB-78) and shows the matching scenario first; embeds the interactive demo tool chosen in WB-24 (Storylane or Navattic) or the WB-44 videos; a booking flow (HubSpot meetings link or FormSection) and a filterable video library (GallerySection).

Build the storyline CMS template: 4-beat arc, persona-specific content fields, proof points, CTA mapping, rich media

A Sanity document or page-builder preset for storylines with fixed slots for Hook, Change, Practice, AI + CTA, proof-point fields (stat, quote, logo), a CTA mapping reference, and media slots; rendered by RoleHeaderSection + ContentBlockSections; one instance created per storyline for WB-85. Naming of the persona field follows WB-32.

Build the six storyline pages on the template: Operator, Owner, Architect, Guardian, Builder, Reseller, each with its specified proof assets

Six pages built on WB-84 with the copy from WB-36 to WB-41 and these assets: Operator (side-by-side video embed, non-tech user quote, template gallery, AI speed stat); Owner (90s walkthrough video, SMB testimonial, price comparison, hardware compatibility list); Architect (API docs link, architecture diagram, tech case study, MCP demo, data source catalogue); Guardian (security docs download, pre-filled questionnaire, on-prem case study, compliance matrix, RBAC diagram); Builder (partner economics calculator, growth case study, API/SDK docs, platform comparison); Reseller (uptime stats, reseller testimonial, support model overview, margin calculator, retention data). Routes and labels per WB-32.

Set up A/B testing for the homepage selector (label wording, order, visual treatment) and storyline CTAs

An experimentation mechanism (Vercel Edge Config flags or a Sanity-managed variant field) that serves variants of the selector and storyline CTAs, with GA4 experiment dimension and HubSpot property recording the variant, a sample-size calculation, and the first test defined (WB-55).

Build /partners: separate Builder (API / custom builds) and Reseller (reliability / margins) paths with distinct CTAs; portal link only

/partners built from WB-35 wireframe and WB-47 copy: a public partner story, two paths with their own CTAs (Builder to partner programme application, Reseller to partner economics guide), certified sub-partner explanation consistent with WB-09, and a single link to the gated portal (not the portal itself).

Optimise page speed: lazy loading, WebP/AVIF images, deferred JS, CDN caching; target Lighthouse 90+

Lighthouse performance 90+ on mobile for homepage, a storyline, /platform and /pricing; next/image with AVIF/WebP and Sanity image CDN params; below-fold sections lazy; third-party scripts (HubSpot, GA4) deferred or loaded on interaction; cache headers verified; results recorded in WB-80.

Build the homepage: animated selector (A/B testable), social proof, capability overview, CTA, lazy loading

Homepage built from HomeHeaderSection (selector variant), LogoSection, ContentBlockSection x3, TestimonialSection, CtaSection, rendering WB-42 copy; selector writes the persona cookie (WB-78) and supports variants (WB-86); Lighthouse 90+.

Test all six persona conversion paths end to end: storyline to platform to CTA to form to HubSpot field population

A Playwright suite that, for each persona, lands on the homepage, selects the persona, follows the storyline to platform to CTA, submits the form with a test email, then asserts via the HubSpot API that a contact exists with persona_type, source and UTM fields set correctly. Runs against staging in CI and once against production after cutover.

Build interactive margin calculators: Builder (network size to margin) and Reseller (client count to revenue)

Two client-side calculators as a registered section (justified as new: CalculatorSection) with inputs, formula, and result display; formulas and price points supplied by Rod; outputs feed CTAs to the partner economics guide (WB-48); embedded on /partners and the Builder and Reseller storylines.

Verify SEO before launch: Rich Results Test, sitemap, redirects, canonicals, GA4 events, HubSpot attribution, OG tags

A QA pass on staging confirming WB-76 and WB-75 work: Rich Results Test passes on each template; sitemap.xml lists every published page and nothing else; every WB-33 redirect returns 301 to the right target with no chains; canonicals self-reference; GA4 DebugView shows every defined event; a test submission shows correct attribution in HubSpot; OG tags render in a social preview tool. Results in the Notion QA table.

Build the feature page template (headline, scenario, expandable tech detail, CTA) and the five feature pages: CMS, scheduling, devices, data, AI

A reusable feature page composition (FeatureHeaderSection, ContentBlockSection with scenario, an expandable tech-detail block, CtaSection) and five instances rendering WB-46 copy, cross-linked from /platform (WB-81).

Build /about, /customers (filterable case studies), /resources (gated and ungated hub), /blog, and legal pages

/about from WB-47 copy; /customers using FilterSection + ThumbnailGridSection over CaseStudy documents with persona, industry and size facets; /resources listing gated lead magnets (WB-48) and ungated content; /blog listing BlogPost documents with a migration of existing posts (WB-33 covers their redirects); legal pages as Page documents.

Build /pricing: tiered layout with monthly/annual toggle, feature matrix, FAQ accordion, persona-aware CTA

/pricing page rendering WB-47 copy: tier cards with a monthly/annual toggle, a feature matrix, an accessible FAQ accordion, and CTAs that swap by persona (WB-78). Pricing values sourced from Sanity Product documents, not hard-coded.

Integrate the HubSpot Service Hub Knowledge Base with the new Wallboard website

The Service Hub Knowledge Base reachable from the new site (linked from Services & Support, ideally on a Wallboard subdomain such as help.wallboard.com), styled consistently, and the old docs location redirected.

Review Rod's input and create an alternative sitemap in Octopus

An alternative sitemap in Octopus.do that reflects Rod's vision, shared with Rod and signed off, so page copy can start. Hard deadline: a workable version of the new site live by 1 October.

Set up a new GA4 property for the new Wallboard website

A GA4 property created for the new site, tagging installed on the Next.js build, key events defined (form submissions, demo requests, KB visits), HubSpot and Search Console linked, and access granted to Rod and marketing.

Build the 301 redirect map from every old Webflow URL to its new route

A complete spreadsheet (Old URL, New URL, Type: exact/regex, Status) covering every indexed and linked wallboard.us URL, stored in Notion on the SEO & Redirects page and as an xlsx in SharePoint, ready for the Next.js redirect layer (WB-76) to consume. Search rankings preserved: every URL with organic traffic in the last 12 months has a 301 target, none go to the homepage as a catch-all.

Wireframe the homepage: hero, need-state selector, proof bar, platform overview, CTA

A low-fi homepage wireframe in Figma covering hero, the selector (six persona need-state paths per tracker, or the three pillars per WB-32), proof bar (logos, stats), platform overview (UI / AI / API as equals), final CTA band; each block mapped to a registered section from the library (HomeHeaderSection, LogoSection, ContentBlockSection, CtaSection). Reviewed by Rod and Iain.

Wireframe pricing, partners (Builder + Reseller paths), persona-aware demo, about, customers and resources pages

Low-fi wireframes for six page types, each composed only of registered sections (or with a written justification for any new section type), with the partners page split into two paths (Builder: API/custom builds; Reseller: reliability/margins) with distinct CTAs, and the demo page showing how it adapts to visitor context. Reviewed by Rod and Iain.

Write the Owner storyline: templates, update from anywhere, scalability, CTA to 'Start free' in under 2 minutes

4-beat storyline copy for the Owner (SMB) need-state: templates, update from anywhere, scale as you grow, with a CTA to a self-serve start path that takes under two minutes. Composed in Sanity, previewed, blog-review clean.

Write the Architect storyline: API depth, data sources, scheduling complexity, CTA to API docs and sandbox

4-beat storyline copy for the Architect need-state with real technical depth (API surface, data source catalogue, scheduling model, MCP), CTA to API docs and a sandbox. Reviewed by John Kesinger for accuracy. Composed in Sanity, blog-review clean.

Write the Guardian storyline: security-first, RBAC, compliance, deployment options, CTA to security docs download

4-beat storyline copy for the Guardian (security/IT) need-state, CTA to a gated security documentation download (the lead magnet from WB-48). Claims verified against WB-43 (security deep-dive). Composed in Sanity, blog-review clean.

Write the Builder storyline: one platform for every client, API, AI expands market, CTA to partner programme

4-beat storyline copy for the Builder (integrator/developer partner) need-state, CTA to the partner programme (Builder path on /partners). Composed in Sanity, blog-review clean.

Write the Reseller storyline: reliability, support included, margins that work, CTA to partner economics

4-beat storyline copy for the Reseller need-state (uptime, support model, margins), CTA to the partner economics guide (WB-48 lead magnet). Composed in Sanity, blog-review clean.

Write homepage copy: hero headline, selector labels, platform overview, CTA

Final homepage copy: hero headline and subhead, the selector labels (six need-states or three pillars per WB-32), proof bar captions, platform overview (UI / AI / API as equals), final CTA. Strapline agreed: 'Heavy foundations make light work.' Composed in Sanity against HomeHeaderSection etc., blog-review clean.

Write the security deep-dive page: auth, RBAC (6 roles), encryption, Kubernetes, GDPR/SOC 2/ISO 27001, deployment options

A scannable /security page with jump links: authentication, RBAC with the six roles diagrammed, encryption in transit and at rest, Kubernetes deployment architecture, compliance posture (only what is certified or in progress, stated accurately), deployment options (cloud, on-prem, hybrid). Every technical claim verified by John Kesinger or the Hungary engineering team. Composed in Sanity, blog-review clean.

Produce four persona-matched videos: Operator 60s side-by-side, Owner 90s walkthrough, Architect API dive, Guardian security walk

Four videos scripted, recorded, edited and hosted (YouTube unlisted or Vercel-served), embedded on the matching storyline pages, with captions. Scripts approved by Rod; product accuracy checked by John Kesinger.

Source and write six case studies, one per persona, tagged by persona, industry and deployment size

Six CaseStudy documents in Sanity, one per need-state, each with customer name (with permission), problem, solution, results with numbers, quote, and tags for persona, industry, deployment size. Filterable on /customers.

Write the platform overview (UI / AI / API capability story) and five feature pages: CMS, scheduling, devices, data, AI

/platform copy presenting UI, AI and API as three equal interface modes (capability story, not feature list), plus five feature pages on the shared template (headline, scenario, expandable tech detail, CTA). Composed in Sanity against FeatureHeaderSection and ContentBlockSection, blog-review clean, technical detail checked by John Kesinger.

Write copy for pricing tiers, partner paths, about (est. 2012), demo options, resources hub and legal pages

Copy for /pricing (tiers, monthly/annual, feature matrix, FAQ), /partners (Builder and Reseller paths), /about (founded 2012, team, Hungary and US), /demo (options), /resources (hub), and legal pages (privacy, terms, cookies, accessibility statement). Composed in Sanity, blog-review clean, legal pages reviewed by Rod.

Create three gated lead magnets: security whitepaper, AI vision paper, partner economics guide

Three PDFs (btg-doc styled to Wallboard brand), each gated behind a HubSpot form with its own form ID, landing page in Sanity, download delivery email in HubSpot, and lead-scoring rules attached (WB-49). Security whitepaper draws on WB-43; partner economics on WB-09 and WB-91; AI vision on Wallboard 2.1 roadmap.

Configure HubSpot for the new site (marketer-owned): persona_type property, lead scoring per persona, nurture sequences, attribution and UTM model

In HubSpot portal 5122961: a persona property with the agreed values; form field definitions for every site form matching the FormSection contract (WB-72); lead scoring rules per persona; nurture sequences per persona; UTM and cross-storyline attribution model documented; routing and notification workflows. Documented in Notion.

Final proofread across all pages: cross-storyline consistency, no persona disparagement, AI depth appropriate, no dead ends

Every page on staging read end to end with a checklist: terminology consistent across storylines, no page talks down another persona, AI claims match product reality and depth suits the reader, every CTA and link resolves, no placeholder text, house style rules pass (no em dashes, no banned words). Issues logged and fixed in Sanity.

Run a WCAG 2.1 AA audit: keyboard navigation, screen readers, contrast, alt text, ARIA labels, selector accessibility

An audit report per page type with pass/fail against WCAG 2.1 AA, every issue logged with severity and fix, fixes applied in components (WB-64) or content, and a re-test confirming zero critical issues before launch.

Cross-browser and device test every page: Chrome, Safari, Firefox, Edge on desktop, tablet, mobile; storyline flow on small screens

A test matrix (page x browser x device) with results, screenshots of failures, issues logged and fixed, and specific confirmation that the 4-beat storyline narrative reads correctly on a phone.

Pre-launch sign-off: all pages on staging, redirects tested, analytics live, forms working, SSL active, DNS ready

A signed-off launch checklist (Notion QA & Launch page) with evidence for each item: every sitemap page renders on staging; redirect map (WB-33) tested with a script against staging; GA4 and HubSpot tracking firing (WB-75); every form posts to HubSpot and creates a contact with persona_type (WB-72, WB-90); SSL valid; DNS TTL lowered and cutover plan agreed (WB-77). Rod signs off. This gates WB-77.

Post-launch: review storylines with sales, capture missing objections, iterate copy

A session with Rod, Guy, Jodie, Matt and Zoltan walking each storyline, asking whether they recognise the persona in real deals; a list of objections the pages do not answer; copy revisions made in Sanity within two weeks.

Post-launch: monitor 2-4 weeks of persona conversion rates, selector click-through, bounce, and run the first A/B test

A weekly dashboard (GA4 exploration or Looker Studio) for four weeks after launch: selector click-through by option, storyline page bounce and scroll depth, form conversion by persona_type, comparison to the old site's baseline; and the first A/B test (WB-86) launched on selector wording with a documented result.

Post-launch: monitor Search Console for indexing, crawl errors, rankings and organic traffic old vs new

Daily checks for two weeks then weekly for two months: index coverage (no unexpected 404s or noindex), crawl errors zero, redirect chains none, keyword rankings for the top 50 terms tracked against pre-launch baseline, organic sessions old vs new. Any regression triaged within 48 hours.

Partner tiering: decide what blank means

Only 91 of 678 partner-typed companies carry a tier. Since 10 August the report WB Partner Activity and the Active Partners view count only companies with a tier or real revenue, so 587 partners are invisible to partner reporting, and neither property sits in any alert. Backfilling tiers would multiply the report population roughly sevenfold overnight and break every period comparison.

Decision 5 cold triage: run the three approved moves

None of the three approved cold-triage moves has run. archive_reason holds zero Stale values (only No engagement, at 649), SQL stands at 1,466, and the roughly 100 quiet hand-raisers were never routed anywhere. Verified live on 20 August.

Rebuild channel role stamp workflows (association trigger + clearing)

The channel role stamp workflows (1863235489 Partner, 1863236139 Hardware Partner) fire on company Type change only, and they set a value they never clear. So a contact linked to an already-Partner company arrives unstamped, and a contact whose company is retyped away keeps a stale stamp forever. 52 stale stamps stand today, 1 Hardware Partner contact is unstamped, and every auto-associated partner contact since the 18 August auto-create re-enable joins the unstamped side until the trigger changes.

Resolve the unassociated-contact backlog (view 69769717)

Contacts created while company auto-create was switched off have no company association. View 69769717 holds the remainder needing manual resolution, including seven held for human review.

Bucket 30 March connector cohort as webinar-sourced

The 282 contacts loaded through the connector on 30 March 2026 still sit in the unknown-source pool and in the re-opt-in review list (283), skewing source attribution and the re-opt-in audience maths.

Owner cleanup: 711 records on deactivated users

462 contacts and 249 companies sit with owners who no longer work here, verified live today. 30 of the portal's 45 owner records are deactivated. Three owner records carry blank names, and one blank-named owner (97467630) is still active. A record on a departed owner gets no follow-up and appears in nobody's queue.

Redesign the SQL grading pass around activity evidence

The grading pass asks reps "is this contact still live?", and that question cannot be answered, because nobody qualified these contacts in the first place. On every record sampled, createdate and hs_v2_date_entered_salesqualifiedlead are identical to the millisecond: contacts were created AS Sales Qualified Lead by an account default. Asking reps to grade records nobody ever worked produces guesswork dressed as data.

Restore-window register: track every 90-day deadline

Several irreversible deadlines are running and nobody is tracking them. The archived properties (Verticals, Areas of Interest, three persona fields) expire around 6 November. The custom mrr deal field, archived 9 July, expires around 7 October. wb_tracked_partner, archived 10 August, expires around 8 November. Deleted contacts from the June purge and from later rep deletions carry their own windows. Each is a decision that gets taken by default if nobody looks.

Internal-domain containment: confirm the tripwire, decide the end state

Nothing removes internal contacts any more. The auto-delete workflow was retired on 13 August because its survival test was "has an open deal", which has nothing to do with whether someone is internal, and because it ignored the is_internal_user flag the guides tell reps to set. The guard that replaced it only flags. A fortnight's watch was agreed to see whether anything still slipped past Never Log; nobody reported back, and one record already proved the gap.

Churn process: Former Customer anchor and automation

Write a show-and-tell guide for creating deals in HubSpot

Screenshot-led guides (Notion, plus Loom scripts) showing the Wallboard team how to create a deal correctly: under the company, with catalogue SKU line items, correct pipeline and stage, amount fields filled as defined. Delivered before the next themed clinic.

Establish the canonical home and update process for the HubSpot product catalogue and new SKUs (with Rod)

The HubSpot product library is the single source of truth for SKUs, every deal uses catalogue-linked line items, and there is a documented process for requesting new SKUs that routes to Rod or Lisa. The library is cleaned so the sales team can find the right SKU quickly.

Clean deal data and build deal-value reporting (MRR / hardware / services split)

Active pipeline free of stale records and zero-dollar placeholders, deal amount fields defined and documented, credited deal-progression activities agreed, a HubSpot report showing MRR, hardware, services, implementation and total value separately, and a weekly administrative-drag field added. A separate conversation with Rod scheduled on the reporting build.

Freshdesk to HubSpot Service Hub migration: confirm connector limits, trial Help Desk Migration, decide scope before Freshdesk is decommissioned on 9 Sep

A go/no-go on the migration route (HubSpot native connector vs Help Desk Migration tool) with scope agreed (which tickets, whether company records are excluded), a full Freshdesk export taken as backup, and new-ticket handling live in HubSpot before Freshdesk is switched off on 9 September.

EMEA sub-partner certification: settle four edge cases

The EMEA sub-partner certification model is agreed in outline but four edge cases block the build: regional dual-status companies, sub-partners invoiced directly by Chargebee, graduation from sub-partner to a direct agreement, and whether certification applies to direct partners at all.

Sweep the Other bucket (29 companies)

The Other company type holds 29 records, and the contents are wrong: MidAmerica Nazarene University, 24 Hour Fitness, Norwegian Cruise Line and TForce Freight are Direct End Users, and Epiphan Video and MirageVision are hardware manufacturers. Other is quietly inheriting the catch-all role that Decision 1 spent real effort removing.

Backfill partner agreement dates (~30 known)

partner_agreement_date holds zero values across all 9,172 companies. The field exists only as an intention, while roughly thirty known agreement-YES flags sit in the prose of the Decision 1 manifest and in the tracker PDFs.

Implement the distributor / certified sub-partner / end-user structure in HubSpot and Notion

Company-to-company association labels live in HubSpot (Contracting distributor, Certified sub-partner), a flag for sub-partners known but not yet visible, known chains applied, the Notion partner tracker reconciled with CRM, and a confirmation note sent to Zoltan and Dominika.

Evaluate HubSpot Data Hub: use cases, ROI brief, and Pro vs Starter recommendation

A short brief for Rod that (a) defines 3-4 concrete Data Studio use cases (data quality monitoring, enrichment, reduced maintenance), (b) maps ROI across Sales, Service and Marketing Hub, and (c) recommends Pro or Starter with cost. Then a demo booked.

Nexmosphere joint webinar: agree content with Bo Merkx and deliver the invitation + Livestorm plan

Webinar content agreed with Nexmosphere, all six input items from Bo received and reviewed, and a written run-up plan covering email invitations and Livestorm setup, ready to execute.

AutoVision v2: pull Apollo data for Guy's Northern California (14-18 Sep) and Texas (19-23 Oct) trips and run pre-visit email pushes in Instantly

For each trip, an Apollo list of dealership decision-makers in the visit area, deduplicated against HubSpot, loaded into an Instantly campaign with 'I'm in your neighbourhood' copy, sending ahead of the visit so Guy arrives with booked slots.

Set up Jodie with LinkedIn Sales Navigator Pro for municipalities prospecting

Jodie has a working Sales Navigator Pro seat, saved searches and lead lists built for the municipality personas, and a short how-to so she can run prospecting herself.

Build the municipalities target list for Houston, Austin, Tulsa, Oklahoma City and the rest of Oklahoma in Apollo

A deduplicated Apollo list of municipality contacts for the target areas, matching the persona and title filters used for the DFW campaign, checked against HubSpot for existing relationships, and loaded into HubSpot (and Instantly if cold email is the channel).

End-to-end review of partner / reseller sales enablement and resources

A short report listing every partner-facing enablement asset (location, owner, last updated, quality), the partner journey from certification to first deal mapped against it, the gaps, and a prioritised recommendation list for Rod.

Hand over the X (Twitter) account @wallboardcms to the Hungary team

The Hungary team has working access to the @wallboardcms X account, 2FA and recovery email point to them, the password has been rotated after handover, and the old credentials no longer work.