Summary — 2026-06-03
Thirteenth pass — 2026-06-03 (Pomelli — Google's free AI marketing tool for SMBs)
One item this pass — and a different shape of item from everything above. Surfaced by Anthony 2026-06-03. Pomelli — a Google Labs × DeepMind AI marketing tool, launched 28 Oct 2025, free public beta. Verified via firecrawl_search + the official blog.google launch post. Three-step flow: (1) Business DNA — enter a URL, Pomelli scrapes the site + images and auto-extracts brand identity (tone of voice, fonts, colour palette, imagery); (2) tailored campaign ideas grounded in that DNA; (3) edit + export on-brand social/site/ad creatives, fully editable in-tool. Feb 2026 added Photoshoot (basic product photo → studio-quality shot). Most items on this radar are tools to adopt or steal. Pomelli is the first that is primarily a free incumbent commoditising a slice of Articulate's own offer — so it scores low on adoption and routes to competitive intelligence, not the install queue.
| Item | Source | P | Q | S | Total | Move |
|---|---|---|---|---|---|---|
| Pomelli — Google Labs × DeepMind AI marketing tool. URL → "Business DNA" brand extraction → on-brand social campaign generation → edit/export. Free public beta. US · Canada · Australia · New Zealand only, English only. | Anthony's forward 2026-06-03; verified via firecrawl_search + blog.google launch post |
1 | 1 | 1 | 3/9 | Skip as a tool to adopt — escalate as competitive intelligence + a positioning hook. Geo-locked out of the UAE/MENA (not available where Anthony operates), English-only, SMB self-serve, commodity-tier output, no strategy/sovereignty/consultative layer. It does not raise what Articulate ships and is not resellable surface. But it is the most strategically significant item to cross the radar in weeks: Google has put a free tool on the exact "generate on-brand SMB social content" task the MEP content layer sells. The correct response is positioning, not adoption — and the geo-gap + missing layers actually sharpen Articulate's whitespace. Flag to Articulate positioning + a Belinda/Will Stack Watch piece. |
Why P=1 (skip-tier adoption)
Could Anthony run it himself for own marketing? Marginally, and only behind a VPN — it's geo-locked to US/CA/AU/NZ, English-only, and it overlaps tooling he already runs at far higher control (the brand-voice plugin for voice, discover-brand + guideline-generation for brand extraction, the ai-image Higgsfield stack for creatives, all Belinda/Dylan-gated). It removes no hours from a real Articulate/GIG/Villa workflow that the existing stack doesn't already cover better. No daily lever.
Why Q=1
Pomelli's output is templated, SMB-grade social content — the commodity tier. It does not beat the brand-anchor + Soul Cinema + Belinda-gate pipeline on a single deliverable Anthony ships. The one genuinely interesting pattern is Business DNA: one-pass auto-extraction of tone/fonts/colour/imagery from just a URL. That's the same job as discover-brand + guideline-generation, compressed to a single automated step — worth a reference look as a front-of-funnel speed-up for the brand-discovery workflow (does Pomelli's auto-DNA beat our multi-step discovery on a real client URL? cheap to test where it's reachable). But the tool itself lifts nothing in flight. Q stays 1 on the reference value of the pattern, not the output.
Why S=1 (and why the threat axis isn't the S axis)
S measures "sellable inside an Articulate workflow." Pomelli is free, geo-blocked, and reseller-margin-less — it cannot be a billable line item. So S=1. The reason this item still matters is competitive, and threat ≠ sellability, so it does not inflate S. Three structural gaps keep Pomelli off Articulate's turf and widen the whitespace already named in the 2026-05-16 competitive scan:
- Geography. US/CA/AU/NZ only. The entire UAE/KSA mid-market Articulate targets cannot use it. (And a US-cloud Google tool scraping a regulated client's site is the opposite of the §3g data-sovereign pitch.)
- No strategy / no human consultative layer. Pomelli is self-serve asset generation. Articulate sells GTM strategy + brand + delivery under one operator — the "solo operator + AI leverage + strategy" whitespace nobody in the GCC scan owns.
- Commodity content only. It automates the cheapest layer (social creatives), not the engine (MEP), the diagnostics (Connie/IC), or the sovereign stack.
The positioning hook to steal: "Google now gives away the on-brand-social-content layer — so why does a UAE SMB still pay an agency? Because the content was never the moat. Strategy, sovereignty, and a named operator are." That's on-brand Stack Watch material and a sharpener for the MEP one-pager.
New generalisable note — the "incumbent-commoditiser" class
Sibling to the RAG-widget-SaaS skip filter (fifth pass) and the crawler/scraper skip filter (eleventh), but pointed the other way. When a hyperscaler ships a free tool that automates a slice of Articulate's own offer, it scores low on P/Q/S adoption — but it does not get filed and forgotten. It routes to competitive intelligence + positioning content. The output is a Stack Watch / MEP-positioning note and a "what's now commoditised vs what's still the moat" read, not an install decision. Adoption-skip and strategic-significance are orthogonal; this class is the case where they diverge hardest.
Class-match against failures.md and prior passes
- agnt.one (fifth pass, 1/9) — closest category match (AI marketing/content automation) but inverse handling: agnt.one was a no-name US SaaS → auto-skip-and-forget; Pomelli is Google → skip-as-tool but escalate-as-intelligence. The vendor weight is the difference.
- Corey Haines / 10-prompt pack (second + sixth passes) — "build the superior Articulate version, don't install the artefact." Same instinct, harder: here you can't even build on Pomelli (closed, geo-blocked) — the only move is to out-position it.
- Vendor-doc spot-check (ninth/twelfth passes) — applied: confirmed region lock and free-beta status against the official blog.google post rather than the hype framing ("builds your entire marketing campaign"). The "entire campaign" claim is social-creative generation, not strategy or distribution.
- No
failures.mdnegative class-match — net-new competitive-intelligence item.
Re-score triggers
- (a) Pomelli opens in UAE/MENA → re-score P (Anthony can use the Business DNA extraction as a front-of-funnel tool) and re-evaluate the competitive threat to the MEP content layer with real local availability.
- (b) Business DNA auto-extraction beats
discover-brand+guideline-generationin a head-to-head on a real client URL (where reachable) → Q promotes; consider lifting the one-pass-from-URL pattern into the discovery workflow. - (c) Google ships an API / Workspace / Ads integration that makes Pomelli embeddable or reseller-accessible → re-score S as a possible delivery component rather than a pure competitor.
- (d) A prospect names Pomelli as the reason they don't need an agency → that conversation is the trigger to ship the positioning piece; log the objection + the answer.
Where it lives (the placement decision)
Filed to canon this pass: this radar entry + a one-liner in SuperSebastian/radar.md inbox. The actionable forks are surfaced to Anthony, not assumed — both sit in other personas' lanes: (1) a Stack Watch / LinkedIn piece on Google-commoditises-SMB-content (Will + Belinda brand-fit call), and (2) a sharpening line in the MEP one-pager on what's now commoditised vs what's still the moat (Articulate positioning, Anthony's call). Sebastian proposes; he does not write to Articulate/ canon.