Paid Advertising Case Study
Real data. Real campaigns. Real findings.
Every paid advertising case study on this page comes from a live CDAI Engine deployment against real client data. No simulations. No projections. What you see is what the engine found — on the first run, automatically, overnight. See the full methodology or review pricing when ready. According to industry research, the gap between reported and true ad spend costs is one of the most persistent problems in paid acquisition.
What the engine found — day one
The findings that platforms never flagged.
In both paid advertising case studies below, CDAI found problems on the first live run that had been invisible to the ad platform, the CRM, and the client. No human triggered the finding. No manual analysis was required. The engine ran overnight and surfaced the results automatically.
Day 1
First finding on a live client account — first run
2
Independent systems — same finding, same night
0
Manual work required to surface either finding
0
Platforms or CRMs that flagged either issue first
Case Studies
Two paid advertising case studies. Both verified in production.
Apex Care Solutions — Senior Care — North Carolina
Read Full Case Study
A campaign had been running for months with zero trackable customers. CDAI caught it on the first night.
CDAI connected to Apex Care Solutions’ live HubSpot account automatically. It pulled in verified Meta ad spend, matched it to their customer database to the exact dollar, and surfaced one active campaign that had been spending with zero trackable customer outcomes over the lookback period. The CDAI health monitor flagged it. The directive engine flagged it independently. Both systems reached the same conclusion overnight — without anyone doing anything. Meta had not flagged it. HubSpot had not flagged it. The client did not know.
2Independent systems — same finding
$0Trackable customers from flagged campaign
Day 1First run — overnight, automatically
Performance Marketing — Lead Generation — Multi-Channel
Read Full Case Study
The reported CPL looked healthy. The true CPL after all costs was materially higher. The team had been scaling the wrong campaign for weeks.
A multi-channel lead generation operation was running Meta and Google simultaneously with broker distribution through Boberdoo. Platform dashboards showed a clean CPL and improving ROAS. CDAI reconciled the actual cost stack — media spend plus broker margins, refund leakage, chargeback exposure, and compliance holds — and found the true cost per lead was materially higher than what the platforms reported. The top-spend Meta campaign received a CUT directive on day one. The team had been scaling it based on platform data for weeks. The data engine got there first.
3×True CPL vs. platform reported
Day 1CUT directive issued on first run
$0Additional analyst headcount required
Technical Validation
Production-verified. Not a demo environment.
Both paid advertising case studies were run against live production systems — not staging environments, not simulated data. The following technical validations were confirmed during these deployments.
✓ Multi-Tenant Isolation
Both tests ran against the same database with Row-Level Security policies active. No cross-organization data leakage was detected at any point during either deployment.
✓ Dual-System Verification
The health monitor and directive engine operate independently. In both cases they reached identical conclusions without sharing state — confirming the finding was not a single-point error.
✓ Zero Consumer PII
Consumer identity data was stripped before processing in both deployments. No names, emails, or phone numbers entered the system at any stage. Verified by the built-in PII scrubber architecture.
✓ Automated Nightly Run
All findings were surfaced by the automated nightly engine run — not triggered by manual input. The system ran on its own schedule and delivered findings before the client team started their day.
Common Questions
What people ask about the proof.
Are these paid advertising case studies based on real client data?
Yes. Both case studies were run against live production client accounts — not simulated or projected data. The findings were produced automatically by the CDAI engine on the first deployment, overnight, with no manual intervention.
Why didn't Meta or HubSpot flag the problem in the Apex case?
Ad platforms report only the costs they control — media spend. They do not reconcile customer outcomes against spend at the campaign level, and they have no visibility into broker fees, refunds, or chargebacks. HubSpot shows CRM records but does not connect them to campaign-level spend. CDAI does both.
How do I know the findings are accurate?
The directive engine and health monitor operate independently and reached identical conclusions in both cases. Spend reconciliation was verified to the exact dollar against the client's live database. Full technical validation documentation is available in the case study pages linked above.
Can CDAI run this kind of audit on my campaigns?
Yes. The free distortion audit runs on one campaign with a 30-day lookback. It costs nothing and requires no commitment. Fill out the form below and we will be in touch within 24 hours.
Client Intake
Find out what your campaigns are hiding.
One campaign. 30-day lookback. No consumer PII. The same engine that found the Apex finding on day one runs on your data. Free.
01
Submit this formCampaign-level information only. No consumer PII required at this stage.
02
Sign NDA and Service AgreementProtects your data before anything is shared. Required before any data access begins.
03
Provide read access or CSV exportsGoogle Ads, Meta, refund history, chargeback data — 90 days minimum. Consumer identity data is stripped before processing and cannot be stored by design.
04
Receive your Directive Sheet in 5–14 daysTrue cost per lead per campaign, distortion signals identified, and enforceable capital directives.
Request Received
We'll review your submission and reach out within 24 hours.