Demandbase launched the AEO Agent on October 20 to track and shape how AI assistants describe brands. Its own assistant-facing file, demandbase.com/llms.txt, does not mention the AEO Agent, Mojo or any agent, and the /ai-agents/ page still lists six older agents without the three launched at GO. The homepage proof numbers also reach the HTML as "+0% Increase in revenue" and "0× Increase in conversions", with the real numbers filled in by a script. An assistant asked "what does Demandbase do for AI search?" is reading pages that predate the launch.
First action
Add the AEO, Campaign and Audience agents to llms.txt and /ai-agents/ today.
How to measure it
Share of 30 tracked AI-search questions where answers mention the new agents
- Now
- –
- Target
- 50%
- By
- 60 days
Steps
Baseline
Ask major assistants 30 buyer questions about Demandbase and AI search; record answers and sources.
Senior Product Manager, AI Adoption & Growth 1 day
Rewrite llms.txt
Current products, all agents, MCP, pricing model, comparisons, case studies.
Product Marketing 1 day
Update /ai-agents/
Three new agent cards and Mojo, each with a use case and a link.
Web team 2 days
Render real numbers
Server-side values for homepage and GO counters; animation optional.
Web team 1 day
Add structured data
SoftwareApplication and Organization markup on agent pages.
Web team 2 days
Dogfood the AEO Agent
Run it on demandbase.com and fix the gaps it finds.
Senior Product Manager, AI Adoption & Growth 1 week
Publish the result
A short "we ran AEO Agent on ourselves" post with before and after answers.
Product Marketing 1 week
Re-measure
Same 30 questions monthly.
Senior Product Manager, AI Adoption & Growth monthly
Why it works
Rewrite llms.txt around today's offer: Demandbase AI, Mojo, the AEO, Campaign and Audience agents, the six existing agents, MCP and the Salesforce Hunter partnership, pricing structure, comparison pages and top case studies. Add the three new agents to /ai-agents/ with one paragraph and a proof point each. Put the real figures (+50%, 3×, +52%, +83%) into the server-rendered HTML and let the script animate from them, not to them; do the same for the GO counters. Then run the AEO Agent on demandbase.com itself and publish the before-and-after.
It is the launch's best demo: a vendor selling AI visibility that measurably fixes its own. The edits are small, and crawlers that do not run scripts stop seeing zeros.
Ready to use
ChecklistAI-readable launch checklist
[ ] llms.txt names Mojo and all nine agents [ ] /ai-agents/ lists AEO, Campaign and Audience agents [ ] Homepage figures in server HTML (+50%, 3x, +52%, +83%) [ ] GO counters in server HTML [ ] Agent pages have structured data [ ] AEO Agent run on demandbase.com [ ] Before/after post published
Risks and how to handle them
Assistants refresh slowly
Also fix pages they cite most, not only llms.txt.
Claims outrun the product
Describe only capabilities that are generally available.
Evidence
- News releaseOctober 20, 2026: Demandbase launched three AI agents at its annual GO conference in New York: an AEO Agent that "tracks and influences how AI assistants portray brands, down to specific members of the buying group" (drawing on "more than 342 million contacts across 13 million companies"), a Campaign Agent and an Audience Agent, building on Mojo, "the B2B marketing agent".Source
- Your websitedemandbase.com/llms.txt (the file written for AI assistants; 39 lines, checked October 22, 2026) describes Demandbase AI, Demandbase One, Advertising, Data, Orchestration, Personalization, solutions and API docs, but never mentions Mojo, the AEO Agent, the Campaign or Audience Agent, the /ai-agents/ page, pricing, comparison pages or customer stories.Source
- Your websiteThe /ai-agents/ page ("Meet the agents at work", checked October 22, 2026) lists six agents: Account Engagement, Intent, Email Personalization, Buying Group Setup, Next Gen Orchestration and Campaign Outcomes. The AEO, Campaign and Audience agents announced on October 20 are not on it, and neither is Mojo.Source
- Your websiteIn the HTML the server sends, the homepage proof figures read "+0% Increase in revenue", "0× Increase in conversions", "+0% Growth in average deal size" and "+0% Increase in pipeline velocity"; the real values (+50%, 3×, +52%, +83%) sit only in data-target attributes that a script counts up to. The GO page's Innovators, Partners, Speakers and Sessions counters are served as "+0" the same way. The homepage quotes the CMO of Zoom, the CMO of IBM Consulting and SAP Concur.Source
- HiringSenior Product Manager, AI Adoption & Growth (remote US, posted September 16, 2026): "Define the product-led growth strategy for Demandbase's AI-native application", own the funnel "across discovery, activation, repeat use, capability composition, AI consumption, and expansion", "Reduce time to first meaningful AI outcome", and help define "trials, entitlements, credit visibility, and usage-based expansion".Source