✦ Simple setup

From zero to autonomous
in under an hour

Connect your ad accounts across Google, Facebook, TikTok, LinkedIn, and Bing — configure your goals, and Gnos AI handles the rest. Here's exactly what happens.

1
2 minutes
Connect your ad accounts

Click "Connect account" and authenticate with Google OAuth. Gnos AI requests read/write access to your Google Ads account via the official API. No credentials are ever stored — only encrypted OAuth tokens.

If you manage multiple accounts through a Manager Account (MCC), Gnos AI imports all of them at once. You pick which accounts to activate.

Ad Platforms
Google Ads
Facebook Ads
TikTok Ads
LinkedIn Ads
Bing Ads
Also connect
Google Analytics 4
Merchant Center
HubSpot
Salesforce
2
5–10 minutes
Gnos AI imports and analyses your campaigns

Gnos AI pulls 90 days of historical data — campaigns, ad groups, keywords, bids, Quality Scores, search terms, and auction insights. It builds a baseline model of your account's normal performance patterns.

You'll see your first anomaly detections and wasted spend opportunities within minutes of connecting.

90d
History imported
~5min
Time to first insight
100%
API-based, no scraping
3
10 minutes
Set your goals and autonomy level

Tell Gnos AI what success looks like for each campaign: a target CPA, a minimum ROAS, a maximum daily spend cap. Then choose how autonomous you want Gnos AI to be — per campaign, per module.

Manual
Insights only — you do everything
Advisory
AI recommends, you approve each action
Co-Pilot
Daily batch approval — most popular
Autopilot
Fully autonomous within your guardrails
4
Ongoing
Gnos AI optimises. You stay in control.

From this point, Gnos AI works around the clock — adjusting bids, testing copy, mining negatives, reallocating budget, and monitoring competitors. Every action is logged with a plain-English explanation and is fully reversible.

You get a morning briefing each day: what changed, what Gnos AI did, and what needs your attention. Override anything at any time from the dashboard or via AI chat.

☀️ Your morning briefing
Overnight Gnos AI adjusted 43 keyword bids, paused 2 underperforming ads, and added 7 new negative keywords. Your estimated spend saving today: $240. One item needs your review →

Frequently asked

Does Gnos AI replace my PPC team? +
No — Gnos AI eliminates the repetitive, manual work (bid adjustments, negative keywords, performance monitoring) so your team can focus on strategy, client relationships, and creative direction. Think of it as a tireless junior analyst that works 24/7.
What permissions does Gnos AI need on my ad accounts? +
Gnos AI requests Standard access via Google OAuth — the same level as an agency you'd invite to manage your account. It can read and write to campaigns, keywords, ads, and budgets. It cannot access billing or payment details, and you can revoke access at any time from your Google account settings.
Can I undo something Gnos AI did? +
Yes — every action Gnos AI takes is logged in the audit trail with the exact before/after values. You can roll back any individual change with one click. Nothing Gnos AI does is permanent or irreversible.
How long does it take to see results? +
Most accounts see meaningful improvements within the first 14 days — typically in the form of wasted spend reduction and CPC decreases. ROAS improvements generally compound over 30–90 days as Gnos AI's models learn your account patterns and the self-improving engine accumulates data.
Does Gnos AI work with small accounts or only large spends? +
Gnos AI works best with accounts spending at least $2,000/month — below this, there isn't enough conversion data to train meaningful ML models. The Starter plan is designed for accounts up to $20k/month managed spend; Growth and Agency scale from there.
Is my data used to train models for other accounts? +
Anonymised, aggregated patterns (e.g. "accounts in e-commerce with >$10k/month tend to convert better at X bid modifier") are used for cross-account benchmarking. Your specific campaign data, keywords, and creative content are never shared or used to train models for other customers.