Results
One platform. Booked jobs counted before and after.
One engagement is measured end to end in dollars, laid out the way a deal team would want it: baseline, source, limitations, and all. No composite cases, no borrowed logos.
- ≈120booked jobs a month recovered
- ≈$39Ka month, the recovered jobs at the platform’s average ticket
- ≈$470Ka year
Measured One three-location PE-backed plumbing platform. An anonymized client case. A walkthrough of the before-and-after data is available on request, so you can check the counting yourself.
The evidence
I counted booked jobs, before and after.
- Context
- A private-equity-backed plumbing platform, three locations, strong inbound demand, conversion not keeping pace with spend.
- The leak
- The online booking path added friction, intake and follow-up ran differently at each location, and calls slipped through at peak and after hours.
- Intervention
- Instant text-back on missed calls, a rebuilt booking flow, one intake and follow-up model across all three sites, and an IVR for overflow, ranked in dollars before anything was built.
- Baseline
- The platform’s own booking counts before the rebuild. Absolute counts are held for the anonymized walkthrough rather than published here.
- Measurement source
- Booked-job counts from the platform’s booking system, before and after, across channels.
- What is measured
- The job count: roughly 120 additional booked jobs a month. Total bookings rose across channels, which rules out phone volume shifting to the web. It does not rule out seasonality or a change in marketing spend. See the limitations below, which are the reason this number is worth arguing with.
- What is modeled
- The dollars: the job count priced at the platform’s average ticket (≈$39K a month, ≈$470K a year), rather than a separately measured revenue line. And anything at platform scale: one brand times twelve is arithmetic, not measurement.
- Limitations
- Two confounders sit on this number and I did not control for either. The before-and-after windows were not season-matched against the same months a year earlier, and marketing spend was not held flat as a condition of the count. If demand rose on its own or the platform spent into it, some of the 120 jobs belong to that rather than to the rebuild. The walkthrough shows the windows and the channel splits, so you can judge the size of that risk yourself. Beyond the counting: one client, one trade, three locations, and the client is anonymized, so you cannot call them.
A walkthrough of the anonymized before-and-after data is available on request: email Zaha.
What I built
Four fixes, ranked by dollars first.
- 01
Closed the after-hours gap.
Missed and after-hours calls now get an instant text-back that routes to online booking, capturing demand that used to go to voicemail.
- 02
Redesigned the online booking flow
so the demand that starts a booking finishes it.
- 03
One intake and follow-up model across all three locations.
Every site runs the same steps from first call to booked job.
- 04
Set up an IVR
a phone menu that catches overflow calls and routes them fast.
The fixes themselves were commodity: a text-back tool, a booking-flow rebuild, an IVR, one intake standard. Any platform can buy that stack in a week; most already have, and most see nothing for it. I ranked the leaks in dollars before spending on any of it.
The field record
Same layer, other industries.
Four engagements on the same layer as the case above: the distance between an inquiry and booked revenue. Clients are blinded and eras attributed. Nothing here carries a measured dollar claim, because nothing here was counted the way the flagship was. Where a figure appears, it is what the client found or decided, and the claim-label system explains why I hold that line.
- An 800,000-customer utility, rebuilt from the call flow up.Agents toggled six or more applications to handle one interaction. I shadowed the agents, mapped the call flows, and redesigned how an inquiry moves across the phone menu, self-service, and the agent’s desktop. The work surfaced more than 50 service improvements, and an early self-service release cut routine call volume in its first quarter. Deloitte Green Dot Award. Delivered as a workstream lead on a Deloitte team.
- A top-five North American bank, after virtual-agent deployments that failed on integration.I mapped the top 20 call drivers, sized what an agent could genuinely handle, and designed the integration architecture. The bank approved the vendor and piloted. Delivered as a workstream lead on a Deloitte team.
- A retained executive search firm’s intake-to-booked-match layer.The path from client brief to booked shortlist ran on re-keying and memory. I automated the workflow end to end. Different industry, same layer: the distance between an inquiry and booked revenue.
- A PE-backed healthcare technology company, 18 months post-close and off thesis.My diagnostic ran across commercial strategy, operations, and technology, and traced roughly 60% of the margin compression to three operational bottlenecks fixable in six months. The sponsor used the 15-initiative plan to restructure incentives and reallocate capital. The 60% is the diagnostic’s finding, not a measured recovery.
The rest of the record sits off this layer: AI adoption and operating-model work, board investment cases, a 100-day plan, an LLM content platform. It is in the evidence pack rather than on this page, because breadth is not what you are buying.
Start here
Want this read on your platform?
Thirty minutes on one platform you own or operate. I’ll tell you straight where I’d look first for the demand that isn’t booking, what to measure to size it, and whether ZTS is the right help, or who is. No deck, no pitch.