Your Biggest Client Might Be Losing You Money

Revenue Is Not Profit

Every contractor knows their top clients by revenue. The property management company that sends 200 service calls per year. The GC that pulls you into every project. The building owner with 40 locations on a master service agreement.

The broader Level founding-team research covers 2,242 contractors, $13.25B in job revenue and 2.5M invoices across operating, private-equity and CFO roles. Each metric uses its own smaller sample. The research count is not a count of current Level clients. Individual company observations stay qualitative. An operator lesson from those reviews is: the biggest customer by revenue is often not the most profitable, and sometimes isn't profitable at all.

Revenue per customer across the dataset ranges from $4,000 to $132,000 (company-level averages; the sample size and period behind this range are not stated on this page). That older range spans about 33x, but the comparison does not tell us how much comes from customer type, company size, accounting methods or work mix. Use your own customer contribution analysis to explain differences before assigning a cause.

The Composite Scorecard Problem

A customer scorecard compares revenue, margin and collection timing. The table below uses independent fictional worksheet inputs, not observed company results, to show why revenue alone cannot rank customer profitability:

The five profiles are independent fictional inputs for a worksheet, not disguised company records. Individual client scorecards stay qualitative because combinations of revenue, customer counts and margins can identify a firm. Each revenue-per-customer figure divides the stated annual revenue by its customer count.

The volume-play row earns $432K of gross profit ($2.4M x 18%), while the deep-account row earns $1.44M ($3.6M x 40%). That is about 3.3x the gross profit on 1.5x the revenue. It illustrates the effects of the chosen revenue and cost assumptions, not a measured causal effect of customer quality. Review pricing, scope and delivery cost before explaining the difference. Reconcile collection ratios separately against aging, credits and write-offs.

Three Types of Unprofitable Customers

Unprofitable customer relationships tend to come from three distinct mechanisms:

Type 1: The Slow Payer

This customer generates decent revenue at decent margins, on paper. But they take 90+ days to pay. Your invoiced revenue shows up as margin, but the cash doesn't arrive for 3 months. Meanwhile, you've paid your techs, your material suppliers, and your subcontractors.

Fictional example: a $500K/year customer that pays in 95 days carries about $130,137 of average AR ($500K x 95 / 365). At 30 days it would be about $41,096. The roughly $89,041 difference is cash you are lending that customer. At an assumed 10% annual cost of capital that is about $8,904/year of carrying cost. Collecting faster frees the $89K once; it does not add $89K of profit every year. Not all of a gap between billed and collected is lost (normal AR, retainage and recent billing are in it), and only true write-offs reduce profit directly.

The fix: track collection rate by customer, not just in aggregate. Your average DSO might be 55 days, but if Customer A pays in 20 days and Customer B pays in 95 days, the average is meaningless. DSO by customer shows which customers tie up the most cash.

Type 2: The Scope Creep Specialist

This customer approves jobs, then adds requests during execution. "While you're here, can you also look at..." The tech does the work because it's faster than refusing. No change order. No additional billing. The job closes at lower margin than quoted, and nobody connects the dots.

Operator check: compare approved change orders to jobs where tech notes mention added work. A labor-hours budget variance does not establish whether scope changed or whether extra revenue was captured. Verify authorization, contract terms, billing and delivery cost before classifying the variance.

If a customer consistently generates jobs that come in over budget, the customer isn't the problem. Your scope management process with that customer is the problem.

Type 3: The Relationship Trap

You've worked with this customer for 15 years. They're "loyal." They send steady work. But their pricing hasn't been updated in 5 years. Their service agreement was priced when your labor costs were 25% lower. Their service calls are further away than your average, so drive time eats into technician utilization.

This is the hardest type to fix because it feels disloyal to reprice a long-standing relationship. But loyalty that costs you money isn't loyalty, it's a subsidy.

In operator reviews, a legacy agreement can look valuable by revenue while its delivery cost consumes that revenue. Keep the individual company figures qualitative. Independent fictional illustration: a $1.2M annual agreement book costs $1.44M to deliver, producing a $240K annual gross loss and a -20% margin. Match revenue and cost by customer before deciding whether to reprice, limit scope or end the agreement.

How to Calculate Customer Profitability

Most accounting systems track revenue by customer. Few track profitability. Here's how to build a customer profitability analysis:

Step 1: Revenue by customer. Pull from your invoicing system. Include all revenue: service calls, projects, service agreements, T&M.

Step 2: Direct costs by customer. This requires job-level cost tracking. If your jobs are properly costed, aggregate labor, materials, and subcontractor costs by customer.

Step 3: Collection by customer. Track two things separately: write-offs (invoices you have given up on) and payment speed. Fictional example: two customers are each billed $200K/year. Customer A pays in 20 days, so average AR is about $10,959 ($200K x 20 / 365). Customer B pays in 95 days, about $52,055 of average AR. At an assumed 10% annual cost of capital, B costs about $5,206/year to carry vs about $1,096 for A, a $4,110 difference. If B also has $6,000 of write-offs, subtract that from B's contribution too. Getting B to 20 days frees about $41,096 of cash once; it is a balance change, not $41K of recurring profit.

Step 4: Service intensity. How many service calls per dollar of revenue? How many callbacks? How much drive time? These are operational drivers whose labor, vehicle and materials costs may already be in direct costs. Use them to explain the ranking, not to charge the same cost a second time.

The calculation:

Customer Contribution = Revenue - Direct Costs - Bad-Debt Write-Offs - Allocated Overhead

Annual AR Carrying Cost = Customer's Average AR Balance x Annual Cost of Capital (average AR is roughly annual billings x DSO / 365)

Keep them separate. Slow payment costs you financing on the balance; it does not erase the revenue unless the invoice is actually written off. Count each dollar once: a written-off invoice belongs in bad debt, not also in the carrying cost.

When you run this for your top 20 customers, expect some rankings to change versus the revenue list. How many is an empirical question for your own books.

The 80/20 That Matters

In the SA (service agreement) data, the power law is extreme:

In this pooled SA data, 1.4% of service agreements generate 49.9% of revenue (sample size and period are not stated on this page; your own share depends on your commercial mix). The $520K average per agreement for that top tier among companies with large enterprise contracts comes from a subset whose size is not stated here, so treat it as directional. Those are the contracts worth protecting, growing, and measuring meticulously. The long tail of $500-$2,000 SAs is necessary for pipeline and market presence, but the economic engine in that distribution is the top 1-2% of agreements, not necessarily the top 1-2% of customer accounts.

This means customer profitability isn't just an interesting exercise, it's existential. If your top 5 customers represent 40% of revenue and one is unprofitable, quantify the loss and its capacity use. New work can help if it contributes enough and does not reinforce the same underpricing; repricing or changing scope may address the loss more directly.

Customer Concentration: The Number Your Bank Cares About

The SA revenue data tells part of the story. But the customer concentration data tells the part your bank, bonding company, and insurer care about most.

Here's how revenue concentration breaks down. The top-1 row is the Level Index measure of share of total revenue from the single largest customer, across 959 companies with at least $100K in annual revenue. The top-5 and top-10 rows have a source sample and period not stated on this page, so treat them as directional:

The median contractor gets 31% of revenue from a single customer and 69% from their top 5. At P75, the top customer alone accounts for 55% of revenue. At P90, one customer is 82% of the business.

Your bank and bonding company care about this number. High customer concentration is a lending risk. If your biggest customer leaves, downsizes, or goes bankrupt, your revenue cliff is immediate and severe. Surety underwriters look at customer concentration when sizing bond lines. Lenders flag it when reviewing credit facilities. If you're above P75 (more than 54.6% from one customer), be ready to explain it in your next financial review.

The fix isn't necessarily finding new customers (though that helps). It's growing existing accounts to rebalance the distribution. If one customer is 55% of revenue, hold that account constant and solve for the other revenue needed. On a fictional $1M total, the top account is $550K and other accounts are $450K. Getting to 30% requires total revenue of about $1.833M and about $1.283M from other accounts, 2.85x their starting revenue. Merely doubling the other accounts gets concentration to 37.9% ($550K / $1.45M). Multi-property expansion and service agreement renewals with your next-tier customers are the fastest route to a healthier concentration profile.

Multi-Property Customers: The Deepest Relationships

The data also reveals that the most valuable customer relationships are multi-property, a single customer with service agreements across multiple locations. Some contractors service 13+ properties per SA.

These relationships are:

• More predictable (recurring revenue across many locations)
• More defensible (switching costs are high when 40 locations are under contract)
• Potential operating efficiency (measure repeat-visit labor, travel and site-specific parts costs before claiming higher margin)
• Less sales cost (one relationship manager, many properties)

If you have customers with single-property SAs, the growth strategy is account penetration, expanding to their other properties, not finding new customers. This is especially true for property management companies that manage 50-500+ properties.

What To Do With Unprofitable Customers

Option 1: Reprice. Most unprofitable customers don't know they're being subsidized. A transparent conversation ("Our costs have increased 20% since we set this pricing, and we need to adjust") can be received better than contractors fear, though some customers will shop the work, so decide your walk-away price first. The alternative is losing you as a vendor, which is expensive for them too.

Option 2: Change the scope. If a customer is unprofitable because of service intensity (too many callbacks, excessive scope creep), restructure the relationship. Tighter SLAs, change order requirements, or a different service tier.

Option 3: Fire the customer. This is the hardest one. But if a customer is consistently unprofitable, consumes disproportionate resources, and won't accept repricing, the resources you free up can be deployed to profitable customers or new business development. When crews are at capacity, every hour a tech spends on an unprofitable account is an hour not spent on a profitable one. If crews have idle time, dropping the account can remove revenue without freeing anything useful, so check utilization first.

Option 4: Reclassify. Some "unprofitable" customers generate indirect value: referrals, market credibility, a flagship project for your portfolio. Quantify that value. Count the margin on referred work, not the revenue. Fictional example: if a $50K customer refers $200K of work at a 30% gross margin, that is about $60K of contribution to weigh against any loss on the account itself.

When Customer Profitability Analysis Doesn't Apply

A small customer base can make the first review simple, but it does not make margin intuitive. Even below $2M, a short list of repeat customers can contain a slow payer or underpriced agreement. Start with those accounts; customer count and cost completeness matter more than a revenue cutoff.

Project-based GCs with one-off jobs don't have "customer profitability" in the recurring sense. Their equivalent is project profitability by type of work and type of client.

If you don't have job-level cost tracking, you can't calculate customer profitability accurately. Fix job costing first, then layer on customer analysis.

---

The Bottom Line

Revenue tells you who your biggest customers are. Customer profitability tells you who your best customers are. They're often not the same.

Run the numbers. Rank your top 20 customers by profitability, not revenue. Identify the slow payers, the scope creepers, and the underpriced legacy relationships. Then have the conversations: reprice, restructure, or redirect your resources to the customers who actually generate wealth for your business.

Q: How does Level help with customer profitability analysis? A: We build a customer profitability dashboard using your QuickBooks and field service data: revenue, direct costs, collection rate, and service intensity per customer. We identify unprofitable relationships, track margin trends by customer, and help you build repricing strategies backed by data. The first audit is free.

Q: How often should I review customer profitability? A: Quarterly for your top 20 customers. Annually for the full customer base. Repricing conversations normally happen at SA renewal. A margin deterioration alone does not authorize a mid-contract increase: follow the signed pricing, change-order and notice provisions.

Q: What if my biggest customer is also my least profitable? A: That's more common than you'd think. The conversation is straightforward: show them the data, explain what's changed (labor costs, material costs, scope), and present a revised pricing structure. A customer can reasonably reject a new price, so evaluate renewal risk, contribution and available replacement work before the conversation. Keep signed pricing and notice terms in force until a permitted change takes effect.