Case study
LED City: what AI call intelligence actually delivered.
Measurement window · May 23 – July 15, 2026
Every sales team makes promises on the phone - callbacks, quotes, follow-ups - and every sales team occasionally forgets one until the customer calls back angry. We built a system to catch that at LED City, our own commercial lighting distributor, before selling it to anyone else.
01The result
Measured against real invoices
$11,000–$37,000
in real, invoice-matched revenue tied to kept callback promises the system tracked and enforced.
Measured across 2,296 analyzed calls · No new infrastructure - runs on hardware already in the building
Composite call quality score
66.5 → 76.6
0–100 scale
Kept-callback rate
32.3% → 57.5%
Promised callbacks that actually happened
Bad-tier calls
30% → 16%
Lower is better▼ improved
Calls with red flags
27% → 13%
Lower is better▼ improved
02Methodology
Why you can trust this number
We didn't pick the flattering number. The same data supports headline figures as high as $335,000 (estimated pipeline value) - we rejected that because it's a model guess, not real cash.
We used the most conservative defensible method: comparing kept vs. broken callback promises against actual invoices in the weeks that followed - an observed comparison from one measured window, not a modeled projection. Full methodology available on request.
What we won't claim
No single ROI multiple - the data supports a range, not a headline number, and we say so rather than round up. This is a measured correlation, not a controlled experiment - and we say that too. It isn't the whole business transformed; it's one system, one measured window, real receipts.
03Go deeper
Request the full methodology
Want this measured in your building?
Book a consultation. We'll walk through your operation and show you exactly where the gaps are.
Book a Consultation