Case Study: How We Generated $2.4M in Recovered Revenue in 90 Days

The Situation

The client was a $12M residential and light commercial HVAC company operating four locations across a major metropolitan area. The business had been acquired 18 months prior by a lower-middle-market PE firm as part of a home services roll-up strategy. Post-acquisition, the operating partner had focused on management team upgrades, fleet optimization, and vendor consolidation — the standard PE playbook. Revenue had grown 8% year-over-year, but the operating partner suspected the company was leaving significant money on the table.

The suspicion was correct. When FoxTrove was engaged to conduct an operational audit, the findings were stark. The company was investing heavily in lead generation — $45K per month across Google Ads, LSA, and directory listings — but hemorrhaging leads on the back end through operational inefficiency. The marketing machine was filling the funnel, but the operational machine was letting it drain out the bottom.

The company had a competent management team and solid technicians. The brand was well-reviewed — 4.6 stars across 800+ Google reviews. Customer satisfaction was not the problem. The problem was that qualified leads were contacting the company and not getting through. The front door was open, but nobody was standing at the counter.

The CSR team consisted of 4 full-time customer service representatives spread across the four locations — an intentionally lean structure that the PE firm had maintained to keep labor costs controlled. During peak hours, this team could not keep up with call volume. After 5 PM, the phones went to voicemail. On weekends, coverage was spotty at best. The assumption was that this was an acceptable trade-off. The audit proved otherwise.

The Diagnosis: Three Revenue Leaks

FoxTrove's operational audit followed our standard methodology: two weeks of data analysis, call tracking, CRM review, and process mapping. We installed call tracking across all four locations, analyzed 30 days of call data, reviewed the CRM for lead disposition patterns, and interviewed CSRs, dispatchers, and the operations manager. The audit identified three distinct revenue leaks, each quantifiable and each addressable with AI.

The Solution

FoxTrove deployed a three-component AI solution across all four locations over a 6-week implementation window. The approach followed our standard methodology: audit, configure, deploy at a pilot location, validate metrics, then roll out to the remaining sites. No experimental technology was used — every component was a proven system configured for this specific operating environment.

The implementation timeline was deliberate. Week 1 was configuration and training data preparation. Weeks 2 and 3 were pilot deployment at the highest-volume location. Week 4 was optimization based on live call data. Weeks 5 and 6 were rollout to the remaining three locations. By the end of week 6, all four locations were fully operational on the new system. CSRs were not replaced — they were redeployed to handle complex calls, outbound sales, and customer relationship management that the AI escalated to them.

The Results

Results were measured against the 30-day baseline established during the audit. All metrics below reflect 90-day averages post-deployment, verified against the client's CRM data and call tracking records.

The 95% call answer rate was achieved through the AI system's ability to handle unlimited simultaneous calls. During peak hours — previously the worst period for missed calls — the AI routinely handled 15 to 20 concurrent conversations without degradation in quality or booking accuracy. The 5% of calls not answered by the AI were edge cases involving garbled audio, non-English speakers requiring specialized routing, or calls that disconnected within the first 2 seconds.

The response time improvement was the most impactful single metric. Moving from a 4.2-hour average to a 12-second average fundamentally changed the competitive dynamic. When a homeowner searches for HVAC service and calls three companies, the first one to provide a qualified, helpful response and book an appointment wins. At 12 seconds, this company was consistently the first responder — even when the customer called competitors simultaneously.

The $2.4M annualized recovered revenue figure was calculated conservatively. It includes only booked and completed jobs that came through the AI system and would not have been captured under the previous operating model (missed calls, after-hours calls, and follow-up sequence conversions). It does not include downstream value from new customer lifetime value, referrals, or the brand impact of faster response times. The actual long-term revenue impact is likely significantly higher.

Equally important: the company did not increase its marketing spend by a single dollar to achieve these results. The $2.4M was recovered from leads the company was already generating and paying for. The AI system did not create demand — it captured demand that was being wasted by operational inefficiency.

The Methodology: Why This Is Repeatable

This case study is not about HVAC. It is about a pattern that exists in nearly every service business: leads come in, a percentage are lost to operational inefficiency, and the business underperforms its potential. The specific leaks — missed calls, slow response, no after-hours coverage — are present in plumbing, electrical, roofing, landscaping, pest control, med spas, dental practices, and dozens of other service verticals. The leaks are industry-agnostic because the root causes are universal.

FoxTrove's methodology is built around this pattern. Every engagement follows the same structure, adapted for the specific vertical and operating environment. The key insight is that most service businesses do not have a demand problem — they have a capture problem. They are already spending money to generate leads. The highest-ROI investment is not more marketing; it is better operational infrastructure to convert the leads they already have.

What the PE Sponsor Saw

The metrics above tell the operational story. The PE sponsor — the operating partner at the fund — was watching different numbers. Their lens was EBITDA impact, multiple expansion potential, and return on invested capital. Here is what the engagement looked like from their perspective.

The EBITDA impact alone was significant — a $1.25M annualized improvement to a business that was generating approximately $2.2M in EBITDA pre-engagement. That is a 57% EBITDA uplift from a single operational initiative. But the multiple expansion implication was where the operating partner got most interested.

Home services businesses in the lower middle market trade at 6 to 9x EBITDA, depending on growth rate, scale, and operational quality. A company with stronger margins, consistent growth, and AI-powered operational infrastructure commands the upper end of that range. The operating partner estimated that the AI deployment — by improving margins, demonstrating scalable operations, and establishing a platform for rapid integration of future add-on acquisitions — could contribute 0.5 to 1.0 turns of multiple expansion at exit.

On an AI implementation investment of approximately $175K (including audit, deployment, first-year operating cost, and training), the return was not 340%. When measured as enterprise value creation — the metric that actually matters to the fund — the return was 19x to 34x on invested capital. That is the kind of operational value creation that defines top-quartile PE performance.

The operating partner has since initiated the same deployment methodology across two additional portfolio companies — a plumbing platform and a commercial cleaning business. The implementation timelines are shorter (3 to 4 weeks instead of 6) because the playbook is already proven. The fund is building AI infrastructure as a repeatable value creation lever across the portfolio, exactly as outlined in our PE AI Value Creation Playbook.