Access Required

Enter your access code to view this research document.

Krepling Intel
Profound AI — Deep Research
1 / 12
Overview

Profound AI

The market-leading platform for AI search visibility — tracking brand mentions across ChatGPT, Claude, Perplexity, Gemini and 7 other engines. Here's everything we found.

$1B
Valuation (Feb 2026)
$155M+
Total Raised
700+
Enterprise Customers
10%
Fortune 500 Coverage
6M+
Prompts/Day
400M+
Real User Convos
Product Mechanics

Three Data Layers — How It Actually Works

① Answer Engine Insights

  • 6M+ synthetic prompts fired per day
  • Queries sent to 10 AI engines via API
  • Full AI response captured & parsed
  • Brand mention, position, sentiment scored
  • Competitor share-of-voice tracked
  • Updates on configurable cadence

② Prompt Volumes

  • 400M+ real user conversations
  • Growing 150M conversations/month
  • Source: browser extension clickstream panels
  • Supplier: Datos (Semrush subsidiary)
  • Refreshed weekly, <7 day latency
  • Stack: dbt, Modal, Snowflake, ClickHouse

③ Agent Analytics

  • Server-side CDN log analysis
  • Tracks AI bot crawls on your site
  • Integrates: Cloudflare, Vercel, CloudFront, Fastly
  • 4-step bot verification (UA → IP → DNS → ASN)
  • Classifies: Indexing vs Citation vs Training bots
  • Shows which pages AI is reading
Key distinction: ChatGPT generates 91 unique "fanout" rewrites from a single user prompt. Profound tracks those rewrites in Answer Engine Insights — but Prompt Volumes only captures the original user query, not the fanouts.
Coverage

10 AI Engines Tracked (+ 1 Coming Soon)

Every engine queried via API with synthetic prompts — Profound captures the full response and parses it for brand signals.

🤖
ChatGPT
Live
🔍
Perplexity
Live
💠
Gemini
Live
🔵
Google AI Mode
Live
✨
AI Overviews
Live
🪟
MS Copilot
Live
🟣
Claude
Live
𝕏
Grok
Live
🐋
DeepSeek
Live
👤
Meta AI
Live
📦
Amazon Rufus
Coming Soon
API & Integrations

Yes — They Have a Public API BetaEnterprise Only

# Python SDK pip install profound from profound import Profound client = Profound(api_key="YOUR_KEY") # Get brand visibility report report = client.reports.get( brand_id="krepling", engines=["chatgpt", "perplexity"] )
# REST API GET https://api.tryprofound.com/v2/reports X-API-Key: YOUR_KEY # MCP Server npx -y @profoundai/client # → mcp.tryprofound.com

✅ What Exists

  • REST API at api.tryprofound.com
  • TypeScript SDK: @profoundai/client (npm)
  • Python SDK: profound (PyPI)
  • MCP Server at mcp.tryprofound.com
  • Webhooks (POST /v1/webhooks)
  • Agent API (run Agents programmatically)
  • Sentiment API — launched June 26, 2026

⚠️ Limitations

  • Enterprise plan only (~$2,000+/mo)
  • Currently in beta — manual approval required
  • Email support@tryprofound.com to request
  • No public OpenAPI/Swagger spec
  • 600 req/hour rate limit per API key
  • Growth plan ($399): CSV/JSON export only
Pricing

Plans & Pricing (2026)

Starter
$99/month
ChatGPT only. Good for small teams testing AI visibility.
  • ChatGPT tracking only
  • 1 brand
  • Basic visibility score
  • Weekly reports
  • No Prompt Volumes
  • No Agent Analytics
  • No API access
  • No data export
Most Popular
Growth
$399/month
All 10 engines. Best for growing brands serious about GEO.
  • All 10 AI engines
  • Multiple brands
  • Prompt Volumes data
  • Agent Analytics (CDN logs)
  • CSV / JSON data export
  • Competitive benchmarking
  • No API access
  • No Profound Agents
Enterprise
Custom pricing
Full platform. API access, custom reporting, dedicated support.
  • Everything in Growth
  • REST API + SDKs (beta)
  • MCP server access
  • Profound Agents (autonomous)
  • Custom engine coverage
  • SSO + team management
  • Dedicated CSM
  • Typically ~$2,000+/month
Annual billing saves ~17% (2 months free). Enterprise contracts typically negotiated annually.
Competitive Landscape

Profound vs. The Field

Tool Engines Real-User Data API Starting Price Best For
Profound 🏆 10 engines ✓ 400M+ convos Enterprise only $99/mo Enterprise brands
Otterly.ai 6 engines ✗ Synthetic only ✓ All plans $Free tier Agencies / SMB
Goodie AI 5 engines ✗ Synthetic only Paid plans ~$49/mo E-commerce merchants
Peec AI 6 engines ✗ Synthetic only ✓ Available ~$79/mo B2B SaaS brands
Semrush AI Toolkit 2-3 engines ✓ (via Datos) ✓ Semrush API Semrush bundle Existing Semrush users
Rankability 4 engines ✗ Synthetic only ✗ ~$99/mo Content teams
Note: Otterly.ai is the strongest Profound alternative for SMB — has API access on all plans, free tier available, and covers 6 engines. Goodie AI is most e-commerce focused.
Due Diligence

⚠️ Data Quality Concerns — Know Before You Buy

The Register Investigation (Sept 2025)

  • Source of Prompt Volumes = browser extension clickstream panels
  • Extensions silently forward users' ChatGPT conversations
  • Upstream supplier: Datos (owned by Semrush)
  • Profound confirmed this to The Register
  • Privacy concern: users opted in via extension ToS, not explicit AI data sharing

4 Structural Accuracy Flaws Identified

  • Panel skews: desktop / male / tech-savvy users
  • Extrapolation amplifies small sample errors
  • Keyword match logic misapplied to conversational prompts
  • Mixed-intent queries inflate category volume

What This Means in Practice

The directional signal (are people asking AI about shoes? is Nike mentioned more than Adidas?) is likely valid. Treat exact volume numbers as estimates ±40-50%, not precise counts.

For enterprise brand decisions, validate Prompt Volumes data against your own Google Search Console queries and customer research before betting the content calendar on it.

Bottom line: Good for trend direction and competitive benchmarking. Weak for precise search demand forecasting. Not a replacement for primary research.
AI Citation Mechanics

How AI Decides What to Recommend

ChatGPT Search (Bing-backed)

  • Product queries trigger Bing Shopping data pull
  • Bing Shopping Ads give merchants paid placement
  • Schema.org Product markup feeds organic results
  • 91 unique fanout rewrites per user prompt
  • Recency signal: recently indexed content ranks higher

Perplexity

  • Merchant Program = preferential product ranking
  • GTIN/UPC required for product matching
  • Google Merchant Center feed format
  • Products with checkout integration ranked higher
  • PayPal Instant Buy merchants get badge boost

Claude & Gemini

  • Claude: primarily training data + web search (Brave API)
  • No native shopping index — cites authoritative editorial sources
  • Gemini: Google Shopping tightly integrated via Product Knowledge Graph
  • Google Merchant Center feed = direct Gemini product visibility
  • Google Ads Shopping campaigns appear in Gemini AI results

What Moves the Needle

  • Schema.org Product + Review + Offer markup
  • Google Merchant Center feed (GMC)
  • Content answering conversational queries (not keywords)
  • Third-party editorial citations (reviews, comparisons)
  • Fast page load for AI crawler re-indexing
Merchant Relevance

Is This Viable for WooCommerce / SMB Merchants?

What Works for SMB Today (Free or Low Cost)

  • Add schema.org Product markup to product pages
  • Submit Google Merchant Center feed (free)
  • Apply to Perplexity Merchant Program (free)
  • Run Bing Shopping Ads (low CPCs, reaches ChatGPT Search)
  • Write conversational FAQ content targeting AI query patterns
  • Use Otterly.ai or Goodie AI for tracking (~$49-79/mo)

Timeline to See Results

Week 1–2

Schema markup + GMC feed indexed by AI crawlers

Week 3–6

Perplexity Merchant Program approved + active; Bing Shopping live

Month 2–3

Measurable citation increase in Perplexity + ChatGPT for brand queries

Profound Specifically — SMB Verdict

  • Starter ($99/mo) is ChatGPT only — too narrow for merchants
  • Growth ($399/mo) makes sense for 7-figure+ revenue stores
  • Enterprise API is out of reach unless Krepling integrates it
  • Better SMB alternatives: Otterly.ai, Goodie AI
Krepling opportunity: Merchants on Krepling's plugin could get AI visibility features bundled in — Krepling calls Profound (or Otterly) API on their behalf, surfaces insights in the merchant portal. Merchants get GEO data without needing their own subscription.
Real Results

Case Studies & Known Customers

Target
Tracks AI brand share-of-voice across all 10 engines. Monitors competitor mentions in AI-generated shopping recommendations.
Walmart
Uses Profound to identify content gaps where competitors appear in AI results but Walmart does not.
Ramp
B2B fintech using Profound to ensure Ramp appears when CFOs ask AI about expense management software.
MongoDB
Tracks developer-intent AI queries ("best NoSQL database") across Claude, ChatGPT, Gemini for developer marketing.
Figma
Design tool tracking AI mentions in creative/design tool comparison queries across all major LLMs.
U.S. Bank
Financial services monitoring brand visibility in AI responses to personal finance queries.

Reported Outcomes

  • Typical improvement: 15-35% increase in AI citation rate over 90 days
  • Content optimized via Profound Agents: ~2-3x faster time-to-citation vs manual
  • G2 reviewers note: strongest ROI for brands with existing content volume (>200 pages)
  • Weak ROI for new/small brands — AI systems still prefer established authorities
Caveat: No independently audited case study numbers have been published. All outcome data comes from Profound's own marketing or G2 self-reported reviews. Treat as directional, not benchmark-grade.
Krepling Strategy

How Krepling Could Use Profound's API

1

Bundled AI Visibility Dashboard in Merchant Portal

Call Profound's Growth API (or Otterly as cheaper alternative) on behalf of merchants. Surface "Your brand appears in X% of AI shoe queries this week" directly in the Krepling merchant portal — zero friction for the merchant, no separate subscription needed.

2

GEO Score as a Sales Tool

Show merchants their current AI visibility score during Krepling onboarding. "Your competitors appear 3x more often in ChatGPT shopping queries. Install Krepling + our AI visibility optimizer to close the gap." Strong hook for merchant acquisition.

3

Auto-Generate Schema Markup via Plugin

The WooCommerce plugin could automatically add schema.org Product + Offer + Review markup to all product pages — the single highest-ROI GEO action for merchants. No Profound subscription needed for this piece, just engineering.

4

GMC Feed Auto-Submission

Krepling plugin generates a Google Merchant Center-compatible product feed (GTIN-keyed JSONL) and handles submission to Perplexity Merchant Program automatically. Merchants get AI-discoverable products without knowing what any of that means.

Action Items

What to Do Next

🚀 Immediate Actions (This Week)

  • Apply to Perplexity Merchant Program for pilot merchant (free)
  • Add schema.org Product markup to WooCommerce plugin
  • Set up Bing Shopping for one test store
  • Trial Otterly.ai ($free tier) before committing to Profound
  • Email support@tryprofound.com to request Enterprise API trial

📋 30-Day Goals

  • Schema markup live on 3+ pilot merchant stores
  • Perplexity Merchant Program approval in hand
  • Profound API access approved (or decision to use Otterly)
  • Merchant portal mockup with AI visibility widget designed
  • First AI-initiated transaction data point collected

💬 API Decision: Profound vs Alternatives

  • Profound Enterprise: Most data, best coverage, $2K+/mo — only makes sense if Krepling bundles it across 50+ merchants
  • Otterly.ai: API on all plans, free tier, 6 engines — best for MVP proof of concept
  • Build thin layer ourselves: Fire queries to AI engines directly, parse responses — cheapest but lowest coverage and maintenance burden
Recommended path: Start with Otterly.ai API (free/cheap) to build the merchant portal integration. Prove the concept works and merchants value it. Then negotiate a Profound Enterprise deal with usage data as leverage.