← All services
For small businesses running Meta & Google ads

AI that optimizes
your ads. For real.

hanjo Ads AI feeds Meta and Google the conversion data they normally never see, so delivery concentrates on the customers who actually buy, book, return, and spend. Built for small businesses, from online stores to booking-based services.

Talk to us See how it works
Google Ads Meta
3.5×
ROAS improvementReturn on ad spend
−60%
CPA reductionCost per acquisition
−75%
Less manual opsHands-on adjustment time
15+
IntegrationsPlatforms & tools connected

Representative results across live deployments of the underlying optimization engine. Results vary by vertical and data volume.

The problem

Most advertisers never unlock the AI already inside their ad platforms.

Every ad platform runs on machine learning, and that machine learning runs on the conversion data you feed it. Most “optimization” stops at creative tweaks, keyword edits, and bid adjustments: indirect moves inside the dashboard. The most fundamental lever, improving the quality of the training data itself, goes untouched. Three problems get left on the table:

Not enough data

  • Signal loss from cookie deprecation and ATT
  • Mismatches from missing parameters
  • Too few conversions for the algorithm to learn from

Low-quality data

  • Every conversion sent with identical value
  • No signal to tell a great customer from a one-time buyer
  • Deep conversion points never reach the platform

Real outcomes never make it back

  • Platform data drifts from actual business results
  • Learning stays a black box
  • No mechanism to correct what the algorithm got wrong
What it is

A second brain between your ad platforms and your conversion points.

hanjo Ads AI sits between the ad platform and the places where conversions actually happen: your site, your forms, your chat, your CRM. It takes control of the training data (conversion data) on your behalf and keeps pushing ROAS up.

↗

CPA falls. High-value customers grow. Revenue follows.

The platform's AI finally learns from the conversions that matter to your business, not just the ones that are easy to measure.

A specialty: booking businesses

Especially strong for salons, spas, and clinics, where booking platforms hide your best customers from your ads.

Any business advertising on Meta or Google benefits from better training data. It shines brightest for booking businesses like the ones we serve at growth.hanjo.ai: if you take bookings on Square or another platform that does not allow custom pixels, your ads optimize on link clicks, not completed bookings, so the platform never learns who actually became a customer.

Clicks aren't customers

  • Square and similar platforms don't allow custom pixels or GTM
  • The deepest signal your ads see is a tap toward the booking page
  • Meta optimizes for people who click, not people who book, return, and spend

Booking forms that learn

  • We optimize your booking flow to capture each customer's preferences and likings
  • Treatments they want, how often they visit, what they care about
  • Your form becomes a source of high-quality signal your platform can't collect on its own

LTV flows back to your ads

  • We track each customer's lifetime value after the first visit
  • That LTV feeds back to the ad platform as training data
  • Your ads learn who to approach and, just as importantly, who not to

Service businesses are our home turf. See how we grow bookings for med spas and clinics at growth.hanjo.ai.

How it works

A self-learning flywheel that keeps improving your ad performance.

01

Track: deeper than pixels can go.

Cross-app and cross-domain journeys are usually where measurement dies. Server-side tracking identifies and follows users even as they move between the web and other apps, so no conversion path goes dark.

↓
02

Structure: turn everything into signal.

Chat transcripts, on-site behavior, form inputs, first-party and zero-party data. These messy, unstructured sources are automatically processed and converted into usable, structured signals.

↓
03

Score: predict the revenue behind each conversion.

The prediction model at the core of the engine estimates pLTV, the future expected revenue of each conversion, and scores every conversion in a form ad platforms can actually use.

↓
04

Sync: tailor the data to each platform's algorithm.

Predicted pLTV scores are enriched and formatted into the exact signal shape each platform's AI learns from best, maximizing user match rates inside Meta, Google, and beyond.

↓
05

Verify: close the loop with real outcomes.

Actual downstream results flow back to the platform automatically, so predicted scores get checked against reality. The platform learns from the difference, correcting its own dataset with every cycle.

↓
06

Compound: collective intelligence raises accuracy for everyone.

Anonymized, statistically processed conversion data feeds the shared prediction model, so accuracy keeps climbing. The model gets better with every account it learns from.

What changes

Fiddly ad ops disappear.
Then it runs itself.

Targeting

Delivery locks onto your best customers.

The platform's AI learns from the high-value conversions we score for it. Targeting self-optimizes toward high-pLTV customers, and ROAS compounds.

Your workload

No more micro-tuning.

AI handles the targeting, so painstaking keyword and audience settings become unnecessary. Set creative broadly, ship more variations, iterate faster.

Over time

Operations automate themselves.

As the platform's learning model matures, manual intervention fades out. AI decides who to reach. Your team focuses on what to say.

Case studies

Working across industries where the real conversion happens late.

Education

Survey answers predicted enrollment: ROAS 180% in two months.

This advertiser measured success at chat-app registration. We added a short survey and used the answers to weight each conversion by its likelihood of becoming a paid enrollment, sending those weighted signals to the platform. ROAS rose week over week for two months as delivery concentrated on high-intent users only.

Healthcare staffing

Moving the learning point cut interview CPA by 41%.

The original conversion point was a landing-page link tap. We switched platform learning to survey respondents who matched the real hiring profile: licensed, aged 20–40, actively switching jobs within six months. Interview CPA dropped 41% within one month.

Luxury real estate

Found the cluster that books meetings: meeting CPA down 33%.

Offline conversions from the CRM weren't producing sales meetings. From form fields such as income, demographics, and location, we identified the feature cluster that actually converts to meetings and sent that signal to the platform, cutting meeting CPA by a third.

Getting started

Plan for about a month from kickoff to launch.

This isn't a plug-in pixel. The integration is built around your existing setup: your booking flow, your CRM, your ad accounts, your conversion points. That adaptation work is extensive, and it's exactly what makes the data trustworthy once you're live. Most setups take around a month.

Step 1: Discovery

Map your goals and gaps

We review your current ad operations and conversion measurement, and identify the highest-leverage setup together.

Step 2: Proposal

Design the right conversion points

We propose the optimal conversion points and integration architecture for your product and industry.

Step 3: Agreement

Confirm plan and schedule

Once the proposal fits, we finalize the plan, the build timeline, and your support setup.

Step 4: Implementation

We handle the technical setup

Meta CAPI connections, CRM and chat integrations, conversion-point wiring. Our side handles the initial build.

Integrations

Connects to your stack through standard webhooks.

Meta CAPI Google Yahoo LINE Ads HubSpot Klaviyo Pardot KARTE Tealium Treasure Data Chat CRMs Form tools iPaaS / ELT

Some tools are currently in limited availability. Ask us about your specific stack.

Pricing

One fee. A percentage of your ad spend.

No retainers, no seat fees, no setup line items. You pay a single service fee as a percentage of your monthly ad spend. The rate is agreed before we start and slides down as your budget grows.

Our incentive stays aligned with yours: the fee only grows when you've decided the results justify spending more.

How billing works:
One invoice, calculated as the agreed % of that month's actual spend
Engine, implementation, and optimization included
Spend less, pay less. Automatically
Plan Monthly ad spend Service fee
Starter Under $10k % of spend
Growth $10k – $50k % of spend
Scale $50k – $100k % of spend
Premium $100k+ Custom
Common questions

What advertisers ask us first.

Which industries does this work for?

Beauty salons, spas, aesthetic clinics, real estate, staffing and recruiting agencies, SaaS, and education: businesses where the real conversion happens off-platform or long after the click see the biggest gains, because that's where default platform learning is weakest.

Our bookings run on Square and we can't add pixels. Does this still work?

Yes, that's exactly the case we built for. When your booking platform doesn't allow custom pixels or GTM, we optimize your booking flow to capture the signals the platform can't, then send qualified conversions and lifetime value back to the ad platform server-side. Your ads learn from real customers instead of link clicks.

What does "curated conversions" actually mean?

Only genuinely valuable actions, such as form submissions, booked visits, and closed deals, get sent to the ad platform as conversions. Training the platform exclusively on high-LTV users steers its auto-optimization in the direction you actually want.

Can we change the conversion point later?

Yes. A common pattern is starting with a shallower conversion to accelerate learning, then switching to a deeper one as data accumulates: a staged tuning approach you can iterate on a daily cadence.

What can it connect to?

The CAPIs of Meta, Google, Yahoo, and LINE Ads, plus HubSpot, Pardot, Klaviyo, KARTE, Tealium, Treasure Data, and chat-based CRMs. Some integrations are in beta, so ask us about your stack.

We don't have engineers in-house. Is that a problem?

Not at all. A dedicated lead handles discovery, conversion-point design, and implementation end to end. Because the integration is adapted to your existing booking flow and tools, expect setup to take about a month. We handle the technical work throughout.

Can it handle multiple accounts or campaigns at once?

Yes. You can design unlimited conversion points per initiative, and data can be unified across accounts.

See what your conversion data
could be doing.

Tell us a little about your ads and we'll follow up with details and next steps. No cost, no commitment.

We only use your information to respond to this inquiry.