Skip to main content
llm-verifiedNumbers it can't prove never ship.

A verified daily analyst for your ad spend.

AdRes connects to your ad accounts and, every day, explains what changed and why, recommends the one move that matters — with dollars attached — and grades its own track record. Every number is computed and checked, so it can't make them up.

Book a demoWhy it's different
Read-only by principle · Connects to Google, Meta & TikTok · Works inside the AI you already use

Sits on your own ad data — and the AI you already use

Google AdsMetaTikTok·Claude & ChatGPT (read-only connector)

The architecture

Every other AI for ads makes numbers up. AdRes structurally can't.

Most tools bolt an LLM onto a dashboard and let it generate prose — so it invents plausible numbers that aren't real. We inverted the pipeline.

01 — compute

Deterministic detectors

Every fact, dollar, and comparison — trends, anomalies, pacing, baselines, peers, targets — is computed directly from your own synced data. No model in the loop yet.

02 — narrate

The LLM only narrates

The model writes the brief at low temperature using only facts that have already been verified. It explains the why — it never sources the numbers.

03 — verify

A verifier gates output

Any claim that doesn't match a computed fact is rejected, and the brief falls back to the deterministic story. A hallucination cannot ship.

✓ llm-verified — narration is clean⤵ deterministic-fallback — narration drifted, facts still ship

The product

One product. Four shipped modules.

The verified daily analyst — in production and proven end-to-end.

01 · The core

The Grounded Daily Brief

The report a competent media planner would write — at two levels, per-campaign and portfolio, in decision order: the verdict and the one at-stake dollar, what changed since yesterday, the recommended plays, the numbers and pacing, peer standing, and a provenance footer.

  • Fact-gated clause by clause — no fact, no claim
  • CPM / CTR / CVR mechanism, rule-outs, baselines, peers
  • One source of truth: dashboard, daily email + CSV, connector
02 · Nobody else ships this

The Self-Grading Advisor Loop

AdRes grades its own past recommendations against what actually happened — and publishes the cumulative hit rate, misses included. Observational, never causal: the accountable track record no dashboard, agency, or AI tool produces.

  • Forecast calibration — projections graded, median error published
  • Plan-vs-actual on target CPA & budget
  • Cross-campaign marginal-dollar reallocation + a learnings ledger
03 · Read-only · live

Bring-Your-Own-AI Connector

A remote MCP server. Add AdRes as a connector in Claude or ChatGPT, sign in, and your own AI gets grounded, read-only tools over your ad data — every payload carrying verified provenance. Not a proprietary chatbot; it puts the AI you already use on top of the verified layer.

  • Works in Claude & ChatGPT — bring your own AI
  • Per-user OAuth, read-only — it recommends, you apply
  • list_campaigns · get_campaign_report · get_portfolio_brief
04 · The demo moment

The Warehouse Network Graph

A live 3D map of your entire ad-data warehouse — brand → channels → campaigns → creatives, audiences, media plan, sync feeds — where every edge explains why it exists. Today: 49.2M rows across Google Ads, Meta and TikTok, with history back to Jan 2023.

  • Every edge carries the reason it exists — no unexplained links
  • Brand → channels → campaigns → creatives, audiences, media plan, sync feeds
  • Sync status and freshness per source, with history back to Jan 2023

How it works

From connected account to trusted brief

01

Connect your ad accounts

Read-only, per-user OAuth. Google, Meta, TikTok and more sync into your own siloed warehouse.

02

Get a grounded daily brief

Every day: what changed, why, and the one move that matters — with dollars attached and a provenance footer.

03

It grades its own calls

Each recommendation is scored against what happened next. The hit rate — and the misses — are published.

04

Ask your own AI

Add the read-only connector in Claude and query the verified layer directly — answers that can't hallucinate.

See the full verified stack →

How we compare

A connector is table-stakes. Verified analysis isn't.

Shipped capabilities, as of June 2026. Researched across 22 sources and adversarially verified.

Feature comparison of AdRes against Triple Whale, Supermetrics, ads-mcp / Pipeboard, and Smartly.
Triple WhaleSupermetricsads-mcp / PipeboardSmartlyAdRes
BYO-AI / connector over ad data✓yes✓yes✓yes—no✓yes
Verifier-gated — no hallucinated numbers—no—no—no—no✓yes
Automated daily briefing✓yes—no—no—no✓yes
Self-grading track record—no—no—no—no✓yes
Cross-campaign reallocation—no—nopartial✓yes✓yes
Read-only / safe by design·unknown✓yes—no·unknown✓yes

Vendor tool-counts and “no-hallucination” claims are self-reported. Triple Whale's own knowledge base concedes its assistant can invent metric definitions and make math errors — the exact failure mode the verifier prevents.

Pricing

Analyst-grade insight, at a fraction of an agency.

Annual plans that scale with how much you run. Read-only by principle, every plan.

Annual commitment. Quarterly payment available.Paid 60-day pilot available, credited toward your first year.
Starter

For a single brand getting its daily brief in order.

$6,000/yr
+ $1,500 one-time setup
+ 3% of measured ad spend
Managed add-on: +5% of measured spend, $1,000/mo minimum
up to $25,000/mo ad spend
up to 10 campaigns per month
GEO & AEO add-on: +$1,000/mo
Book a demo
  • The grounded daily brief — campaign & portfolio
  • Dashboard + daily email with CSV
  • Google Ads, Meta & TikTok integrations
  • Approve & apply recommended changes (coming soon)
  • Email support
Most popular
Growth

For scaling brands that want the full advisor loop.

$24,000/yr
+ $5,000 one-time setup
+ 1.5% of measured ad spend
Managed add-on: +5% of measured spend, $1,000/mo minimum
$25,000–$150,000/mo ad spend
Book a demo
  • Everything in Starter
  • GEO & AEO optimization: get named in ChatGPT, Perplexity, Gemini & AI Overviews
  • MCP connector for Claude & ChatGPT (read-only)
  • Self-grading track record & forecast calibration
  • Cross-campaign marginal-dollar reallocation
  • Warehouse network graph
  • Priority support
Enterprise

For large portfolios and agencies: white-label the verified analyst across your brands.

from $120,000/yr
+ from $15,000 one-time setup
+ 0.5% of measured ad spend
Managed add-on: from +5% of measured spend (scoped)
$150,000+/mo ad spend · talk to us for exact terms
Book a demo
  • Everything in Growth
  • White-label briefs & client-facing track record
  • Unlimited brands, accounts & seats
  • Custom integrations & dedicated success manager
  • SLA & security review

The % applies to the ad spend AdRes measures · you pay Google and Meta directly; AdRes invoices only its own fee · on the managed add-on our team makes the changes, the software stays read-only · final scoping confirmed on your demo call. See full pricing

FAQ

Questions, answered honestly

What is AdRes?
A verified daily analyst for your ad spend. It connects to your advertising accounts and, every day, explains what changed and why, recommends the one move that matters with dollars attached, and grades its own past recommendations. It augments or replaces the daily reporting layer you currently wait on an agency for — it doesn't replace your marketing team.
How is it different from other AI ad tools?
Most tools bolt an LLM onto a dashboard and let it generate prose — which means it invents plausible numbers that aren't real. AdRes inverts that: deterministic detectors compute every fact from your own data, the LLM only narrates already-verified facts, and a verifier rejects anything it can't prove (falling back to the deterministic story). A hallucination structurally cannot ship. On top of that, we grade our own recommendations and publish the hit rate — something no competitor does.
Does AdRes make changes to my campaigns?
The software doesn't. AdRes is read-only by principle. It computes, explains, and recommends — humans apply the moves. That's a deliberate trust position: the connector and the analysis are safe to give your own AI because nothing in the software can write back to your ad accounts. If you'd rather not apply the moves yourself, the optional managed add-on has the AdRes team make the changes in your ad platforms; the software itself stays read-only.
What platforms does it connect to?
Your own ad accounts — Google Ads, Meta and TikTok today; Microsoft, LinkedIn and Amazon Ads on the roadmap — sync read-only into a siloed, per-brand warehouse via official APIs. The same verified report then renders on the dashboard, in a daily email (with the data attached as CSV), and through the connector.
Can I use my own AI?
Yes, on the Growth and Enterprise plans. AdRes ships a remote MCP server. You add it as a connector in Claude (Desktop, Code, or web), sign in with your AdRes account via per-user OAuth, and your own AI gets read-only, grounded tools over your ad data — every answer carrying verified provenance. The connector is the closer, not the point: connectors are table-stakes; the verified data layer behind it is the differentiator.
How do I know the recommendations are any good?
Because we publish our own graded track record — including the misses. Every recommendation and forecast is stored and scored against what actually happened next ("CTR 1.56% → 1.72% in the 7 days since — moved the right way"), and the cumulative hit rate and median forecast error are shown. It's observational, never causal — but no dashboard, agency, or AI tool produces this artifact.
Is my data secure?
Each brand's data is siloed and multi-tenant by permission. Authentication is per-user (Clerk OAuth), and the connector is read-only with per-user scope. Nothing can write back to your ad platforms.
What stage is the product at?
In production and proven end-to-end — but pre-pilot. It runs today on demo data plus one real synced campaign, and everything compounds on real daily spend. The best way to see it is to book a demo — we'll run the verified daily brief on one of your own accounts, read-only.

See it run on your data.

Book a 30-minute demo. We'll run the verified daily brief on one of your accounts — what changed, why, the move that matters, and our graded track record, live on your data.

Book a demoSee the daily brief
In production and proven end-to-end · 30-minute demo, on your data