Articulate AI Labs · radar pass · 30 May 2026

The AI-Native Marketing Landscape, three reads.

152 companies. One landscape. Three buyers reading it from where they actually stand. The research is the page that sits underneath: 100 marketing firms + 52 collaboration platforms — 16 with primary-source intelligence from Firecrawl scrapes, the rest search-confirmed. This page is the read on what that map means depending on whose desk it lands on.

Compiled by SuperSebastian For Anthony Booth Source hype-radar PDF
Why read the landscape at all

It is the room I sell into. Naming it changes the pitch.

Every UAE buyer has already seen Singlegrain, hovered on NoGood’s USD 20K floor, been quoted by Konvergense, or had Hovi’s “3x faster, 60% cheaper” line dropped into a conversation. Knowing the room is not strategy — it is preparation.

The research underneath this page maps seven archetypes globally and three collaboration-platform layers (incumbent + AI, AI-native workspace, embedded coworker). What changes per reader is not the map. It is what the map means depending on whose desk it lands on. A UAE client reads it for trust signals. Articulate reads it for positioning. just-add.ai reads it for category permission. Same 152 companies; three different stories.

Read One · A UAE client

Trust, channel, price, exit.

A regional GM, a Dubai family-business owner, a senior IC running matrix authority into Saudi. The buyer who actually signs the cheque.

What the landscape looks like to a UAE buyer

A long list of US agencies they will never call. A short list of regional names they have heard of (Push MENA, Konvergense, Hovi, iAdsClick, AI Marketing Pro, GCC Marketing). One or two global brands — NoGood, Singlegrain — quoted by global procurement at a price they will not match. Glean and Dust on the IT side, both incomprehensible to anyone not in the buying-committee’s engineering chair.

What they are actually looking for

Four questions, in order. Will it work in Arabic, on WhatsApp, with my team? Most US-headquartered firms fail this immediately. Regional firms pass on channel and language but read as outsourced execution shops. What does it cost? NoGood’s public USD 20K+/mo floor is a hostile signal — UAE buyers ask first. Konvergense’s AED-transparent pricing is comfort. Will it still run if you disappear? Family businesses survived three generations of consultants leaving; they will not buy something locked to one person. Have you done this here before? Named regional case studies, in their sector.

What the landscape does not answer

None of the regional firms publish a clear delivery model. None of the global firms publish a regional case study. The buyer’s comparison set is incomplete by design — everyone’s page is a claim and nobody’s is a finished install. Whoever ships the first publicly-named, regionally-relevant case study (a Roche, a Christies, a GIG) takes the conversation. That is the gap on the buyer’s side, not just Articulate’s.

Read Two · For me (Anthony)

This is my skill set. I do not need to reinvent if I can use.

Twenty-five years of consultative marketing, intersected with agent-mesh + AEO + data-sovereign substrate. The radar is my room, not my obstacle.

Where I sit on the map

Structurally adjacent to RZLT, Graphite, Demand Curve. Not to Monks or WPP. Closer to an AI-native growth shop with a proprietary discipline than to a holdco agentic platform. Closer to a senior-IC mentor product than to a SaaS tool. The seven-archetype map clarifies the engagement shapes that fit my two-day-a-week posture and the ones that do not. Stops me discounting against the wrong comp.

What I can sell against

Named prices, named claims, named clients. Most regional comps sell AI-augmented services at AED 15-30k/month. NoGood charges the global equivalent at USD 20K+. We sell a built-and-handed-over engine at AED 55k flat, then leave. That is a positioning sentence I could not write before the research existed. Each comp on the map becomes a sentence in the pitch room. The radar is a glossary of competitor primary-source claims, kept honest.

What I already use, will not rebuild

Higgsfield Soul Cinema ships image discipline. ElevenLabs ships voice. Dust ships multiplayer-AI workspaces. Cloudflare Workers ships automation. Hermes Agent ships the daemon. Goodie AI ships AEO tracking. Anything I rebuild that the radar shows shipped at scale is a tax on my margin. The work is assembling these into a UAE buyer’s context with regional discipline — PDPL-clean, Arabic-capable, WhatsApp-first. That assembly is what nobody on the map is selling.

Read Three · For a Leo + Takis collaboration

Workshop and labs sit in a category of one.

Anthony + Leonidas + Takis as a three-person craft team. Distinct register from articulate-ai.work. Manara linkage. Workshop subpage.

Where the trio sits on the map

Not in the agency category at all. The honest read is that just-add.ai intersects the embedded-coworker layer (Viktor, Lindy, Taskade) and the agency-as-OS sliver (MindStudio, Relevance AI, Dust implementation partners). The Slack-channel mockup hero is the most honest framing on any Articulate property — the product is a conversation, not a deck.

What the radar shows is missing

Globally, almost no one sells the discipline to ship the agent. Dust has implementation partners. MindStudio targets agencies. None of them runs an Arabic-language workshop on shipping production AI agents inside a regional WhatsApp + family-business context. Connecting that gap to Manara’s Arabic-technical-talent pipeline gives the trio category-of-one permission. The radar shows it because nobody else is there.

What collaboration brings that solo would not

Two things. Team-of-three structure. No regional comp on the map operates as a small named team-of-three labs-shape. Konvergense has 40 specialists, Hovi has 50, both read as agencies. A craft team with publicly named partners is structurally distinct. Cohort revenue + craft IP, separated. Workshop revenue at AED 4-8k per seat builds the talent pipeline; bespoke installs at AED 40-80k feed the craft IP. The radar shows Demand Curve running this exact two-tier (agency + program). The pattern works; the regional variant is unclaimed.

Definitions

What the labels mean.

Vocabulary used on this page

AI-native
A firm whose unit economics, delivery model, or proprietary IP depend on agent mesh, model orchestration, or LLM-aware workflow — not a firm that bolted ChatGPT onto a 2019 service line.
AEO / GEO
Answer Engine Optimisation / Generative Engine Optimisation. Showing up inside ChatGPT, Perplexity, Claude, Google AI Overviews. The new SEO. Goodie AI is the category-defining tool.
Claygency
An outbound agency built on the Clay + Smartlead + HeyReach trio with LLM personalisation. New 2026 category, named publicly by Clay’s blog. Replaces traditional SDR-led outbound.
Multiplayer AI
Dust’s framing: humans and agents working in the same shared workspace, with the same context, conversations, and notifications. The opposite of “single-player AI assistant.”
Embedded coworker
The Viktor / Lindy / Stamp shape: an AI agent that lives inside the tool teams already use (Slack, Teams, Gmail) and operates as a colleague, not a separate app.
Data-sovereign substrate
The stack a regulated UAE/KSA buyer needs — UAE PDPL Federal Decree-Law 45/2021 clean, regional inference (Hostinger UAE PoP), client-owned at handover. Articulate’s structural wedge in MENA.
The dataset

What sits underneath these three reads.

100marketing firms 52collab platforms 16primary-source Firecrawl scrapes 7marketing archetypes 3collab layers

The dataset is canonical at hype-radar-ai-marketing-collab-2026-05-30.pdf. A live sortable / filterable version of the same data sits in the Cowork artifact “Hype Radar — AI-native marketing + AI-collaboration companies.” This page is the read; that one is the spreadsheet.

Marketing entries are flagged: p = primary-source Firecrawl scrape (16 entries, includes Singlegrain, NoGood, RZLT, Graphite, Animalz, Monks, Omneky, Dust, Glean, Karrot, Konvergense, Push MENA, AI Marketing Pro, Hovi, GCC Marketing, iAdsClick, Demand Curve, Directive). s = search-confirmed. i = inferred from training knowledge. Re-radar trigger: offer.md updates, a new MENA AI-agency with productised IP enters, or the Dust / Glean / Sana category produces a category winner.