Profound vs OtterlyAI vs Sight AI vs DIY: a 2026 AEO tooling teardown for agencies

If you run an SEO/AEO agency in 2026, you have four real options for tracking how AI engines cite your clients' brands. They are not interchangeable. Each makes different trade-offs on price, coverage, integrations, and how much engineering you have to own.

This is a grounded teardown of the four. By the end you should know which option fits which agency size, what each really costs at scale, and the specific gotchas that matter when you're committing client retainer to one of them.

TL;DR

Tier	Tool	Best for	Floor
Enterprise	Profound	1-3 enterprise clients with $50K+ retainers; non-technical buyer experience	$399/mo basic; $5K+/mo for full LLM coverage
Mid-market	OtterlyAI	Teams who want a polished dashboard without enterprise spend	$29/mo Lite; mid-tier unspecified
Bundled	Sight AI	Teams who also want content generation + AI optimization in one product	$49/mo and up
DIY data layer	Apify Actor (AEO Citation Monitor)	Agencies with 5+ clients who want to own the data, build their own dashboard	~$0.36/prompt across 6 engines

The single biggest decision is whether you want a SaaS dashboard you rent or a data layer you own. Agencies with under 3 clients usually rent. Agencies with 5+ usually own. This post explains why.

Profound (the category leader)

Funding: $155M total raised across A/B/C; $96M Series C in February 2026 led by Sequoia.

Pricing tiers:

$99/mo (ChatGPT only)
$399/mo (+Perplexity, +AI Overviews)
$2,000–$5,000+/mo enterprise (Grok, DeepSeek, full multi-engine)

Wins for: non-technical end-users in mid-market or enterprise who want a turnkey dashboard. SOC 2, SSO, support included. ChatGPT Shopping support is a recent differentiator.

Loses for: agencies running multiple clients. The pricing is per-brand, so 10 clients on the $399/mo tier = $3,990/mo against you, before your retainer math even works. The basic tier is also ChatGPT-only — for full 6-engine coverage you're at enterprise pricing.

What you actually get: a polished dashboard with brand mention tracking, sentiment, competitor share-of-voice. The data shape is proprietary; you can't easily port it elsewhere if you outgrow them.

Unit economics check:

Their floor at full coverage is ~$5K/mo for 1 brand.
Effective per-resolution cost: at 100 prompts × 6 engines × weekly = 2,400 records/mo, that's ~$2.08 per resolution.
For comparison, the Apify Actor's resolution cost averages ~$0.06 wholesale.

Profound is 30× more expensive per resolution at the basic SMB tier — but they're paying for product polish, dashboards, support, and the SOC 2 / SSO compliance their enterprise buyers expect. A solo agency without those needs is overpaying.

Scrunch (CDN-level integration)

Different positioning: Scrunch is integrating at the CDN layer to modify content delivery to AI agents at the edge. They're not just tracking — they're optimizing in real-time.

Funding: $26M raised.

Pricing: $79/mo basic, $250/mo standard, custom enterprise.

Engines: ChatGPT, Gemini, Perplexity, AI Overviews. Notably no Claude or Grok in their public tier.

Wins for: brands serving content directly (publishers, ecommerce, knowledge bases) where CDN-level tweaking has measurable lift. Their pitch is "Agent Experience Platform" — different category from pure tracking.

Loses for: agencies. The CDN integration assumes the brand owns its CDN config. For multi-tenant agency setups, the integration model adds friction.

Unit economics: at ~$250/mo for moderate coverage, this is actually competitive on a per-brand basis if their CDN-level optimization is delivering measurable lift. We haven't validated that claim independently.

OtterlyAI (mid-market)

Pricing: $29/mo Lite. Mid-tier pricing not publicly disclosed (a recurring pattern in this space).

Engines: ChatGPT, Perplexity, Google AIO. No Claude, no Grok in public tier.

Wins for: marketing teams who want a polished dashboard without enterprise spend. They claim 20K+ marketing pros on the platform; clearly the most-known mid-market option.

Loses for: agencies who need full 6-engine coverage. You'll outgrow Otterly's Lite tier within 2 months of multi-client work.

The hidden cost: Otterly aggregates by brand and locks in to their data shape. You don't get raw records you can pipe into BigQuery or your agency's reporting stack. For 1-2 clients, this is fine — for 10 clients it's a workflow ceiling.

Sight AI (bundled)

Pricing: $49/mo and up.

Engines: ChatGPT, Claude, Perplexity, others (varies by tier).

Differentiator: bundles monitoring + content generation + indexing in one product. Worth considering if you also want a "we'll help you write the content that AI engines will cite" workflow.

Wins for: teams who want a one-stop solution and don't mind the trade-off of being locked into Sight AI's content quality.

Loses for: teams whose content is already strong; you don't need their generation features and you're paying for them anyway.

Khadinakbar (the only existing Apify Actor in this space)

Pricing: $0.080 per result, pay-per-event.

Engines: ChatGPT, Claude, Gemini, Perplexity.

Stats: ~35 users on Apify Store. Functional but minimal documentation, no public schema, no vertical templates.

Wins for: developers who want a quick spike for one brand and don't need polish.

Loses for: agencies. No locale targeting, no sentiment, no rank position, no public schema. The data is fine for spot checks but not for ongoing client work.

At $0.080/result it's still 3.2× cheaper per record than Profound's effective rate, but missing material features for any agency workflow.

DIY data layer: AEO Citation Monitor (Apify)

Pricing (per record):

$0.010 (Perplexity)
$0.015 (Google AIO)
$0.020 (Anthropic, Grok)
$0.025 (Gemini grounded)
$0.075 (OpenAI base) + $0.05–$0.50 (grounding bracket — typical case +$0.20 for medium)

Engines: ChatGPT, Claude, Gemini, Perplexity, xAI Grok, Google AI Overviews. Full 6-engine coverage at the same per-record price across the board.

Differentiators:

Public Zod schema published on npm — you own the data shape; portable.
Vertical templates — saas-b2b, ecommerce-d2c, local-services, agency, media-publisher, fintech, with prompts pre-grouped by intent.
Locale targeting — native country/language routing for Perplexity, OpenAI, AI Overviews; system-prompt instruction for Anthropic, Gemini, Grok.
Bracketed pricing — OpenAI grounding cost varies dramatically; the brackets keep margin positive across light/medium/heavy queries.
Automatic transport fallback — direct API → Vercel AI Gateway → OpenRouter on rate-limit or 5xx.
HTML report + Slack/email digest — buyers can download a self-contained HTML report from each run, or wire up Apify's built-in Slack/email integrations.
Lite (wizard) variant — 3-field form for non-technical buyers; same engine, simpler input.

Wins for: agencies running 5+ clients who want to own their data layer and build their own dashboards. The unit economics scale linearly.

Loses for: non-technical buyers who don't want to write any JSON or run scheduled Apify Actors. (For those, the Lite variant and Profound serve better — at very different price points.)

Unit economics at scale

Let's run the numbers for an agency with 10 clients × 50 prompts × 6 engines × weekly:

Tool	Per-month cost	Notes
Profound enterprise	$50,000+	$5K/mo × 10 clients minimum
OtterlyAI mid-tier	$3,000–$8,000	10 clients × $300–800/mo per brand
Sight AI	$2,000–$6,000	10 clients × $200–600/mo per brand
AEO Citation Monitor (Apify)	~$430	10 × 50 × 6 × 4 = 12,000 records × ~$0.036 weighted avg

The DIY data layer is 5–115× cheaper than the SaaS alternatives at this scale. The trade-off is you build your own reporting; the SaaS tools build it for you.

The break-even point is roughly:

Under 3 clients: SaaS wins. The dashboard polish is worth more than the unit-cost premium.
3–8 clients: mixed. Depends on whether you have engineering capacity to build a reporting layer.
8+ clients: the data layer wins on every dimension. The reporting tax is amortized; the data ownership matters increasingly.

What to choose if you're starting fresh

Your decision tree:

Are you a solo agency or under 3 clients in AEO right now? → Profound. Just pay for the polish.
Are you 3-8 clients with light engineering capacity? → Otterly or Sight AI for now. Plan to migrate to a DIY data layer when you hit 10 clients.
Are you 8+ clients with at least some engineering capacity (or willing to outsource a small dashboard build)? → AEO Citation Monitor on Apify. The data layer scales linearly with your agency.
Are you mostly non-technical but want low-cost AEO baselines for clients? → AEO Lite (the wizard variant of the Apify Actor) — same backend, 3-field form.

What this teardown won't tell you

We've focused on price and coverage. We haven't covered:

Sales-team handoff quality — Profound has account managers; the DIY route has Apify support tickets.
Specific feature parity for niches — vertical-specific tools may exist for legal/healthcare/finance compliance.
Specific content-generation quality for Sight AI — we haven't audited their output.
Drift over the next 12 months — Profound's $96M cash buys feature velocity. The competitive picture in May 2027 will look different.

Get a baseline you actually own

If you're going the DIY data-layer route, start with the AEO Citation Monitor. Pre-built vertical templates, public Zod schema on npm, transport fallback, parallel runner, HTML report.

A 1-prompt × 1-engine sample run costs under $0.01. A full 25-prompt × 6-engine production sweep costs ~$3.65 with default OpenAI grounding. Multiply by your client count and weekly cadence to see your real monthly spend.

If you have 10+ clients and you're paying $5K/mo+ to a SaaS dashboard today, the cost of a one-week DIY pilot is rounding error. The data shape you take home is portable forever.