When the founding partners of Harmon Structural Engineering walked into their first strategy conversation with Boost, they didn't think they had a revenue problem. They had a "bandwidth problem." That was the phrase Greg Harmon used — a civil engineer by training, precise with language, uncomfortable with ambiguity.

Harmon Structural was a 47-person engineering consultancy based in Charlotte, North Carolina, specializing in structural assessments, retrofit design, and construction oversight for commercial properties. They'd been in business for eleven years. Revenue had hovered between $11M and $13M for the last three. The work was good. The team was strong. The reputation in the Carolinas was excellent. But they couldn't grow.

Greg and his co-founder, Anita Harmon, had tried. They'd hired a marketing agency ($8,500/month for 18 months — total spend: $153,000). They'd invested in a CRM that the sales team used primarily as a contact list. They'd brought in a management consultant for a two-day strategic planning session that produced a 40-page deck and zero operational change. They'd added two business development reps who were productive for six months, then plateaued because the pipeline infrastructure couldn't support them.

The "bandwidth problem" was real, but it was a symptom. The disease was architectural. Harmon Structural didn't need more effort. They needed infrastructure.

Fourteen months later, the firm crossed $36M in annual revenue. Their sales cycle had shortened by 60%. Their close rate had more than tripled. They'd expanded from regional to national without adding a single business development rep. And the systems running their growth would keep working whether Greg and Anita were in the office or on a beach in Portugal.

This is the story of how that happened, told in enough detail that any mid-market operator can see the mechanics behind the numbers.

The Diagnosis: Seven Vendors and Zero Integration

The first thirty days of every Boost engagement follow the same discipline: audit before architecture. Before we build anything, we need to understand what exists, what works, what doesn't, and where the highest-leverage interventions live.

At Harmon Structural, the audit revealed a pattern we see in roughly 70% of mid-market companies between $5M and $30M. The individual pieces weren't terrible. Some were quite good. But nothing was connected, and the disconnection was silently destroying value at every handoff.

Here's what the landscape looked like.

Seven vendors, zero shared data. Harmon was managing a marketing agency (handling Google Ads and SEO), a web design firm (maintaining the website), a CRM platform (HubSpot, underutilized), an accounting firm (QuickBooks management), an IT managed services provider (handling infrastructure), a recruiting firm (sourcing engineers), and a part-time fractional CFO (quarterly financial reviews). Each vendor was competent in isolation. None of them shared data, coordinated strategy, or even knew what the others were doing. Greg estimated he spent eight to ten hours per week on vendor coordination — status calls, email threads, reconciling conflicting reports.

A CRM that was technically active but functionally dead. HubSpot had been purchased two years prior. It contained roughly 4,000 contacts, maybe 60% with accurate information. The sales team used it to store phone numbers. Pipeline stages hadn't been updated since initial setup. Deal values were entered sporadically. The "reporting" consisted of a monthly export to Excel where someone manually tallied proposals sent and deals closed. When we asked the two business development reps to describe their sales process, they described two completely different workflows. Both lived in their heads.

Marketing that generated activity but not attribution. The marketing agency was spending $6,200/month on Google Ads and had achieved first-page rankings for several local keywords. Monthly reports showed impressions, clicks, and "leads" (form submissions). What nobody could answer: which of those leads became qualified opportunities? Which became clients? What was the actual cost per acquired client? The answer, we discovered, was that the $153,000 spent over 18 months had generated approximately $340,000 in attributable revenue — a positive ROI, but barely, and far below what the spend should have delivered with proper infrastructure.

A sales process dependent on personal relationships. Harmon Structural's best revenue generator was Greg himself. He had spent eleven years building relationships with general contractors, property developers, and municipal engineers across the Carolinas. Roughly 55% of new business came through Greg's personal network. The two business development reps handled inbound inquiries and cold outreach, but their combined pipeline was less than half of what Greg generated from a single golf outing. If Greg stepped back from business development — which he needed to do in order to focus on running a growing firm — the pipeline would collapse.

Manual operations consuming skilled labor. Proposal generation took three to five days because engineers were manually compiling project scopes, fee calculations, and reference materials. Scheduling was done through email threads. Follow-up was inconsistent — some proposals got a check-in call after a week, some never got one at all. Client onboarding was a 14-step process that lived in a shared Google Doc and required an operations coordinator to manually trigger each step.

The total cost of this fragmented infrastructure wasn't visible in any line item. It was distributed across wasted hours, missed opportunities, slow response times, and the invisible tax of coordination overhead. Our estimate: Harmon Structural was losing between $1.8M and $2.4M annually in unrealized revenue and operational waste. Not from bad work or bad people — from bad architecture.

The Architecture: Building All Five Layers Connected

Based on the audit, we designed a 90-day transformation plan that addressed all five layers of growth infrastructure simultaneously. This is critical to understand: we didn't fix sales first, then marketing, then automation. We designed the entire system as a connected architecture and built it in parallel, so the integrations and feedback loops were active from day one.

Here's what we built, layer by layer.

Layer 1 — Strategy Architecture. We started with a full strategic reset. Harmon's ICP had never been formally defined. They took every project that came through the door, from $15,000 residential assessments to $2M commercial retrofits. We analyzed three years of project data and discovered that commercial retrofit projects between $150K and $800K delivered the highest margins, the shortest sales cycles, and the strongest referral rates. We defined the ICP with precision: property management firms and commercial developers with portfolios of 10+ buildings in metro areas with aging commercial infrastructure. We identified twelve target metros beyond Charlotte where this ICP was concentrated and underserved. This strategic clarity changed everything downstream — marketing could target with precision, sales could qualify with confidence, and pricing could be optimized for the highest-value segment.

The 90-day sprint was designed with three OKRs: (1) Build and deploy integrated sales infrastructure, (2) Launch targeted marketing in three expansion metros, (3) Automate the five highest-time-cost operational workflows.

Layer 2 — Revenue Engine. This was the highest-leverage intervention. We rebuilt Harmon's sales infrastructure from the ground up. CRM was reconfigured with pipeline stages that matched their actual sales process — not the generic template that HubSpot installed by default. Every stage had entry criteria, exit criteria, and automated next actions. Lead response was transformed: we deployed AI lead response that engaged every inbound inquiry within 30 seconds, 24 hours a day. The AI was trained on Harmon's service offerings, common questions, and qualification criteria. It could have an intelligent conversation with a prospect, assess their project scope and timeline, and route qualified leads to the right team member — all before a human touched it.

We introduced commission-only closers trained in technical services sales. These weren't generic sales reps. They understood engineering proposals, project timelines, and the language of commercial construction. Their compensation was 100% commission — they earned when Harmon earned. Incentives perfectly aligned. No base salary risk for the firm.

We designed the quoting system to generate proposals in hours rather than days. Standard project types were templated with configurable scopes, automated fee calculations, and pre-loaded reference materials. Engineers still reviewed every proposal, but instead of building from scratch each time, they were reviewing and customizing a 90%-complete document.

Layer 3 — Growth Amplification. Marketing was redesigned around the newly defined ICP. Instead of broad local SEO and generic Google Ads, we built targeted campaigns for the twelve expansion metros, focused specifically on commercial property managers and developers searching for structural assessment and retrofit services. Content strategy shifted from generic "about us" blog posts to technical content that demonstrated expertise: case studies of successful retrofits, regulatory compliance guides, and building assessment frameworks. Reputation management was systematized — every completed project triggered an automated review request sequence that built Harmon's Google Business Profile to 4.9 stars across locations.

The key integration: every marketing lead flowed directly into the reconfigured CRM, was immediately engaged by AI lead response, and was tracked from first click to signed contract. For the first time, Harmon could see exactly which campaigns generated qualified pipeline and closed revenue.

Layer 4 — Operational Intelligence. We automated the five workflows consuming the most skilled labor. Proposal generation (templated and semi-automated — engineer review time dropped from 4 hours to 45 minutes per proposal). Client onboarding (the 14-step Google Doc became a triggered workflow — each step launched automatically based on the previous step's completion). Follow-up sequences (every open proposal received a structured follow-up cadence — no proposal went cold unintentionally). Scheduling (AI-assisted scheduling eliminated the email thread ping-pong). Reporting (real-time dashboards replaced the monthly Excel export).

All of this ran on $1/action pricing. Harmon's automation volume started at roughly 1,800 actions per month and grew to 4,200 by month six as we expanded workflows. Total automation cost: under $4,500/month, replacing what had previously required two full-time administrative positions.

Layer 5 — Compound Infrastructure. Every system was connected from day one. When a deal closed, the CRM triggered the onboarding workflow. Marketing received the conversion data and adjusted campaign optimization. The financial dashboard updated revenue projections. The operations team received their project brief. The retention engine scheduled the first check-in at 30 days post-engagement. This wasn't bolted on later — it was designed into the architecture from the beginning.

Quarter by Quarter: The Compounding Trajectory

Month 1–3: Foundation and Quick Wins

The first 90 days focused on getting the infrastructure live and generating early wins to build confidence and fund continued investment.

By the end of month one, AI lead response was active. The impact was immediate and startling. Harmon had been averaging a 26-hour response time to new inquiries. Some leads waited three days for a callback. With AI response, every lead was engaged in under 30 seconds. The qualification rate — the percentage of leads that converted to a booked meeting — jumped from 8% to 23% in the first month alone. Not because the leads were better. Because the speed and quality of initial engagement changed dramatically.

By month two, the CRM overhaul was complete and the commission-only closers were active. Close rate began climbing. Pipeline visibility transformed — Greg could see every active opportunity, its value, its stage, and its probability for the first time in the company's history.

By month three, the first marketing campaigns in the expansion metros were live. Proposal automation was deployed. The operations coordinator who had been manually managing the onboarding process was reassigned to client success — a higher-value role that directly impacted retention.

Quarter 1 results: Revenue pace increased from $12M to a $14.5M annual run rate. Close rate moved from 11% to 24%. Average lead response time dropped from 26 hours to 28 seconds. Greg's personal business development time dropped from 20 hours/week to 6 hours/week.

Month 4–6: The Compound Effect Activates

This is where the architecture started paying compound dividends. The feedback loops between layers began producing results that no single layer could have generated alone.

Marketing data showed that the expansion metro campaigns in Atlanta and Nashville were generating significantly higher-quality leads than Charlotte. Why? The competitive landscape was thinner — fewer engineering consultancies had strong digital presence in those markets. This insight (only possible because marketing was connected to sales data through the CRM) triggered a reallocation of marketing spend. Atlanta and Nashville budgets doubled. Two additional metros — Tampa and Dallas — were added.

The proposal automation freed up enough engineering bandwidth that Harmon could respond to larger, more complex RFPs that they'd previously passed on due to capacity constraints. Two of these larger projects closed in month five — one at $420K and another at $680K — projects that would not have been pursued under the old system.

The retention engine triggered its first wave of 30-day check-ins and quarterly review outreach. Three existing clients expanded their engagements based on these touchpoints. One property management firm that had used Harmon for a single building assessment signed a portfolio-wide agreement covering 14 properties — a $340K annual contract that came from an automated check-in email.

Quarter 2 results: Revenue pace hit $18M annualized. Close rate reached 34%. Pipeline grew 2.5x as expansion metros began contributing meaningfully. Automation volume reached 3,100 actions/month. Operational costs dropped 28% relative to revenue.

Month 7–10: Scaling National

With the infrastructure proven, scaling became a matter of inputs, not reinvention. The system worked. Now it needed more fuel.

Marketing expanded to eight metros. Each new market launch followed the same playbook — targeted campaigns, local SEO, reputation building — and each launch was faster than the previous one because the templates, workflows, and integrations were already built. A new metro could go from zero to generating qualified pipeline in under three weeks.

The commission-only closing team expanded from two to four. Because the sales process was systemized (not dependent on individual tribal knowledge), new closers became productive within two weeks of onboarding. The onboarding itself was partially automated — training materials, CRM access, pipeline management protocols, and call scripts were delivered through a structured sequence.

Greg stepped back from active business development entirely. For the first time in eleven years, he was running the company rather than selling for it. Anita took over strategic client relationships while the system handled pipeline generation and closing.

A pattern emerged that we see consistently in successful engagements: as the infrastructure matured, the quality of opportunities improved. Harmon was no longer chasing every project. The system attracted and qualified the right projects — higher value, better fit, stronger margins. The average project size increased from $85K to $142K, not because they raised prices, but because the ICP targeting and qualification infrastructure filtered for better-fit opportunities.

Month 7–10 results: Revenue pace hit $28M annualized. Close rate stabilized at 41%. Twelve metros active. Pipeline was 4x what it had been at the start of the engagement. Average project size up 67%.

Month 11–14: Compounding at Scale

The final phase of the first fourteen months was characterized by compounding — each improvement amplifying previous improvements in ways that were difficult to predict but unmistakable in the data.

Marketing cost per qualified lead dropped by 40% as campaigns accumulated data and optimized. The system knew which keywords, which ad creative, and which landing pages generated not just leads but qualified leads that closed at high rates. This is the kind of optimization that's impossible when marketing and sales data live in separate systems.

The retention engine produced its most significant result: referral revenue. Harmon's systematic follow-up, quality delivery, and review generation created a referral network that began generating inbound opportunities without marketing spend. By month 14, referral and repeat business accounted for 30% of new pipeline — up from 12% at the start.

The automation volume reached 4,200 actions per month at a cost of approximately $4,200/month. For context, the two administrative positions that previously handled these tasks manually had cost roughly $8,500/month in salary and benefits. But the real savings weren't in the direct cost replacement — they were in the speed, consistency, and error reduction that automation provided. Proposals went out the same day instead of four days later. Follow-ups happened consistently instead of sporadically. Onboarding was smooth instead of chaotic.

Month 14 results: Annual revenue run rate crossed $36M. Close rate: 43%. Active in fourteen metros across the Southeast, Mid-Atlantic, and Texas. Operational costs as a percentage of revenue had dropped from 34% to 21%. Employee headcount had grown from 47 to 62 — but revenue per employee had nearly doubled.

The Numbers, Summarized

Metric: Annual Revenue | Before Boost: $12M | After 14 Months: $36M+ | Change: 3x

Metric: Close Rate | Before Boost: 11% | After 14 Months: 43% | Change: 3.9x

Metric: Lead Response Time | Before Boost: 26 hours | After 14 Months: 28 seconds | Change: 99.97% faster

Metric: Active Markets | Before Boost: 1 (Charlotte) | After 14 Months: 14 metros | Change: 14x

Metric: Avg. Project Size | Before Boost: $85K | After 14 Months: $142K | Change: +67%

Metric: Pipeline Volume | Before Boost: Baseline | After 14 Months: 4x baseline | Change: 4x

Metric: Operational Cost (% of revenue) | Before Boost: 34% | After 14 Months: 21% | Change: -38%

Metric: Vendor Count | Before Boost: 7 | After 14 Months: 1 (Boost) | Change: -86%

Metric: CEO BD Time | Before Boost: 20 hrs/week | After 14 Months: 0 hrs/week | Change: Eliminated

Metric: Automated Actions/Month | Before Boost: 0 | After 14 Months: 4,200 | Change: —

Metric: Proposal Turnaround | Before Boost: 3–5 days | After 14 Months: Same day | Change: -80%

What Any Operator Can Learn from This

You don't need to be an engineering consultancy to apply the principles behind Harmon's transformation. The structural lessons are universal.

Lesson 1: Your "bandwidth problem" is probably an architecture problem. When growth stalls despite effort, the bottleneck is almost never a lack of hard work. It's a lack of infrastructure. Harmon's team was working hard. They were working on the wrong things, in the wrong sequence, with the wrong tools.

Lesson 2: Integration beats optimization. Harmon could have gotten a better marketing agency, a better CRM consultant, and a better sales trainer — and they still would have been stuck. The problem wasn't the quality of any individual function. It was the complete absence of connection between functions. A 10% improvement to seven disconnected systems produces less value than connecting those systems at their current capability.

Lesson 3: Speed compounds. Responding to leads in 28 seconds instead of 26 hours isn't a 3,300x improvement in response time. It's a fundamental change in the economics of lead conversion. The same principle applies to proposal turnaround, onboarding, and follow-up. Speed at every stage of the customer journey creates a compounding advantage that slow competitors can't match with effort alone.

Lesson 4: Systems free leaders to lead. Greg Harmon spent the first eleven years of his company as the primary revenue generator. He was too valuable selling to focus on running the business, and too busy running the business to sell effectively. The infrastructure freed him from both traps. By month ten, he was doing neither — the system generated pipeline and the team closed deals — and the company was growing three times faster than when he was doing everything himself.

Lesson 5: The compound effect is real, but it requires all five layers. Harmon's growth didn't come from any single intervention. It came from the interaction between interventions. Marketing fed sales. Sales data improved marketing. Automation freed capacity. Capacity enabled expansion. Expansion generated data. Data improved everything. Remove any one layer from that system and the compound effect breaks.

Fourteen months. One integrated system. From $12M to $36M.

The infrastructure is still running. The compound effect is still accelerating. And Greg Harmon finally has the "bandwidth" he was looking for — not because he found more hours in the day, but because the system doesn't need his hours to grow.

About Boost

Boost is the growth infrastructure company for ambitious mid-market businesses. We integrate AI-powered sales, marketing, automation, and strategic consulting into one compounding ecosystem. Founded by operators. Powered by AI. Learn more at Boost.com.

For more information, visit Boost.com.