Harvard Agrees. We Already Built It. MedusAIPM Enters Enterprise Beta.
- Patrick Duggan
- Mar 12
- 4 min read
# Harvard Agrees. We Already Built It. MedusAIPM Enters Enterprise Beta.
**Published:** March 12, 2026
**Author:** Patrick Duggan, DugganUSA LLC
**Category:** Product / AI / Enterprise
Harvard Business Review published a piece this month titled *"LLMs Are Overtaking Search. Here's How to Adjust Your Online Presence."*
We read it. We agreed. Then we realized we'd shipped the solution twelve hours before they published.
The Problem HBR Identified
The article lays it out cleanly: the way people find companies is changing. Instead of typing a query into Google and scanning ten blue links, they're asking ChatGPT. They're asking Perplexity. They're asking Gemini. And those models are answering — with opinions.
The model doesn't give you ten options. It gives you one answer. Maybe two. If your company isn't in that answer, you don't exist.
Search Engine Land is calling it "LLM perception drift" — the idea that how AI models perceive your brand is now a measurable, trackable, optimizable metric. Right next to keyword rankings and share of voice.
Search Engine Journal says Answer Engine Optimization is one of the five key enterprise SEO trends for 2026.
HBR says adjust your online presence.
We say: audit it first.
What MedusAIPM Does
MedusAIPM — AI Presence Management — is the newest head of the MEDUSA product suite. Here's what it does in one API call:
**Phase 1: AI Model Survey**
We query GPT-4o, Gemini 2.5 Flash, and Perplexity Sonar directly. We ask them about your company. Not "what's your sentiment about Brand X" — we ask them the same questions your next customer would ask:
- *What is this company? What do they do?*
- *Would you recommend them?*
- *Who founded them? Where are they based?*
- *What's their reputation?*
- *Rate your confidence from 0-10.*
Then we score the responses across five dimensions: **Awareness, Accuracy, Sentiment, Recommendation,** and a metric we invented called **AIPM-NPS** — Net Promoter Score, but the respondents are AI models instead of humans.
NPS has been the gold standard for customer sentiment since Fred Reichheld published it in 2003. Nobody thought to ask the machines. We did.
**Phase 2: Domain Structure Analysis**
Your AI perception is only half the story. The other half is whether AI crawlers can even read your website.
MedusAIPM crawls your domain and scores:
- **robots.txt** — are you blocking GPTBot, ClaudeBot, PerplexityBot? Most companies are and don't know it.
- **Schema.org / LD-JSON** — structured data is how AI models understand what your business actually is
- **Semantic HTML** — article tags, section hierarchy, accessibility labels
- **Sitemap** — can crawlers discover your content?
- **Meta / OpenGraph tags** — the summary layer
We score each dimension, weight them, and combine with Phase 1 into a single **Combined AIPM Score** (60% AI perception, 40% structural readiness). Capped at 95 — because we guarantee 5% bullshit exists and anyone claiming 100% is lying.
**Phase 3: The Fix**
Most tools tell you what's wrong. We generate the fix.
Bad robots.txt? Here's an AI-optimized one. Missing Schema.org? Here's your Organization and FAQPage LD-JSON, pre-filled with your domain. Priority-ranked recommendations with impact scores.
The audit is the diagnosis. The fix is the prescription. One API call.
What Makes This Different
We've looked at every tool in the space. Conductor AI does sentiment monitoring. LLM Pulse does weekly brand tracking. AIScore Report does website readiness audits. HubSpot has an AEO grader.
None of them do all of it. And none of them do this:
**Brand Threat Enrichment.** MedusAIPM cross-references your audit with our threat intelligence. If someone is impersonating your brand at the ASN level — weaponizing your company name on bulletproof infrastructure — we'll find it during the audit. Because sometimes negative AI perception isn't organic. Sometimes it's adversarial.
We're a cybersecurity company that built a marketing tool. Nobody else in this space has a STIX feed backing their brand audit.
First Results
We ran MedusAIPM on our partner's domains before announcing it.
- **cleansheet.life**: 64/95 combined score. Missing LD-JSON. Robots.txt partially blocking AI crawlers. Four priority fixes identified.
- **cleansheet.info**: 61/95. Similar gaps.
We delivered the audit reports and fix templates the same night. That's the product: audit, score, fix, track.
Enterprise Beta
MedusAIPM is entering beta today for enterprise customers.
**What you get:**
- Full AIPM audit via API (`POST /api/v1/aipm/audit`)
- AIPM-NPS scoring across GPT-4o, Gemini, and Perplexity
- Domain structure analysis with fix generation
- Brand threat enrichment
- Audit history and longitudinal tracking
- Leaderboard ranking vs. other audited domains
**Pricing:**
- Enterprise beta: $499/mo — unlimited audits, full API access, SLA
- Professional (coming Q2): $99/mo — 50 audits/day
- Free tier available for single audits (register at epstein.dugganusa.com/pricing)
**API endpoint:** `analytics.dugganusa.com/api/v1/aipm/audit`
The Timing
HBR told the world the problem exists. Search Engine Land gave it a name. Search Engine Journal listed it as a top-5 trend.
We shipped the solution.
The threat intelligence market is projected to hit $43.3 billion by 2033. The AI SEO market is forming right now — listicles, agency roundups, tool comparisons are appearing weekly. The window for first-mover advantage on a combined perception + structure + threat enrichment platform is measured in months, not years.
MedusAIPM is Patent #102 in the DugganUSA portfolio. The claims are filed. The code is deployed. The first audits are stored.
Harvard agrees. We already built it.
*MedusAIPM is part of the MEDUSA product suite by DugganUSA LLC. Butterbot, our AI chatbot, can run AIPM audits directly — just ask him "how does AI see my brand?"*
*DugganUSA LLC | Minnesota | dugganusa.com*
*D-U-N-S: 14-363-3562 | SAM.gov UEI: TP9FY7262K87*
**Sources:**
- Harvard Business Review, March 2026: "LLMs Are Overtaking Search. Here's How to Adjust Your Online Presence."
- Search Engine Land: "Why LLM perception drift will be 2026's key SEO metric"
- Search Engine Journal: "5 Key Enterprise SEO and AI Trends for 2026"
- World Economic Forum: "Global Cybersecurity Outlook 2026"
*Her name was Renee Nicole Good.*
*His name was Alex Jeffery Pretti.*
Comments