Your Biggest Account Just Left — Don't Replace It Yet
Your Biggest Account Just Left — Don't Replace It Yet
In October 2008, I lost two builder accounts in the same week.
Not one. Two. Same week.
My first move was to get on the phone. Every GC I'd ever pulled a permit with, every property manager who'd ever left a message, every builder who'd been "interested" in the last three years. I was going to replace that volume. I had twelve guys, six trucks, and I was not going to be the shop that shrank.
That instinct cost me weeks I didn't have. By the time I stopped dialing and started reading my own numbers, I was already behind on decisions I should have made ten days earlier.
This piece is for the version of you sitting in your truck right now, account gone, phone in your hand. Your gut is telling you to start selling. That's wrong.
Stop the Bleeding Before You Chase New Blood
The replacement instinct is biological. Revenue hole appears, you fill it.
But the first question after a major account walks isn't "who can replace this?" It's the 90-day question: how long can you pay your guys, your truck notes, your insurance, and your rent with zero new revenue?
Write that number down. If it's less than 90 days, that's the emergency. The lost account is the symptom.
I didn't ask that question in October 2008. I was too busy trying to rebuild the revenue line before I'd even looked at my cash position. Weeks went by. I was pricing jobs I shouldn't have been pricing, meeting with a builder in Fitchburg who wanted four houses on a handshake and a promise to pay "when closings came through." That's how scared I was.
If I'd run the 90-day math on day one, I'd have seen what I eventually saw anyway — just sooner, with more options still on the table.
The first 48 hours are diagnostic. Not sales. Not prospecting. Diagnostic.
The Dependency Was the Problem — Not the Client Who Left
When one client represents 35, 40, 50 percent of your revenue, you haven't built a business with a big client. You've built an extension of their operation. You're running their overflow. When they stop calling, you find out what you actually own.
The client didn't create that problem by leaving. They revealed it by staying as long as they did.
The account leaving didn't break your shop. It showed you the shop was already broken in a way you'd stopped seeing because the checks were coming.
The Whitman Builders situation — 2011, builder out of Marlborough — is a version of this I lived from the receivables side. Whitman wasn't my biggest account by then, but they were my most persistent slow payer. Sixty-one cents of every dollar I was owed came back. The rest disappeared into a bankruptcy I saw coming and didn't act on because I needed the volume.
Big accounts that pay slow feel like revenue. They feel like a relationship. What they actually are, in many cases, is a lien waiting to happen and a concentrated bet you're pretending is a partnership. The check keeps coming so you keep going. Meanwhile you've stopped quoting smaller jobs as aggressively, stopped developing your service base, stopped building anything that doesn't run through their project schedule.
When they go quiet, you're not just down revenue. You're down the revenue and the capability you let atrophy around it.
Replacing Volume Fast Is How You Re-Create the Same Problem
The frantic replacement chase doesn't fix the dependency. It copies it — usually with a worse client on worse terms, because you're scared and the GCs across the table from you can see it.
GCs who pay at 60-plus days are already using your labor as a free credit line. That's a business model, not a cash flow accident. Pay-when-paid clauses, retainage withheld past substantial completion, punch lists weaponized at collection time — those aren't edge cases among builders. That's standard operating procedure for the kind of GC who's actively looking for a hungry sub. A shop coming off a major account loss, behind on payroll planning, is exactly who they're looking for.
You will sign a pay-when-paid clause in month two of your panic that you'd have laughed at in a normal year. I damn near signed one in November 2008 with that Fitchburg builder. Didn't, but only because I finally stopped and ran my numbers first.
After the crash I didn't replace the lost builder volume with new builder volume. I went backward, deliberately — residential service, small commercial, jobs I could control, jobs where the homeowner paid on completion. Smaller jobs, yes. But jobs where I knew what I was owed and when I'd see it.
It took longer than I wanted. I hated parts of it. But it's why I still had a shop in 2011 to fire Whitman from, and why there was something left to sell to Reliant in 2018.
What the Numbers Actually Tell You Right Now
A few weeks into October 2008, I finally stopped dialing and pulled my numbers.
My receivables had stretched from 45 days to 110 days without me fully registering it. The accounts hadn't just left — they'd been bleeding my cash position for months before they walked out. The volume kept me from seeing it clearly. Big invoices going out, payments coming in eventually, and the gap between those two things had become wallpaper.
Pull your AR aging report right now. Anything over 60 days. Then anything over 90. Add those numbers up. That money is technically yours. It is not actually yours.
The layoff decision in October 2008 was the hardest thing I did. Four guys, hired personally. I knew their situations. But when I stopped chasing replacement volume and read my cash position honestly, the math was plain: I couldn't carry twelve men on a line of credit I couldn't service. I kept three plus myself. I hated it. I still think it was right.
If I'd kept scrambling for new builder volume instead of facing those numbers, I'd have carried those extra men another 60 to 90 days on borrowed money. The collapse would have been worse, not smaller, and I'd have run out of options entirely.
What Your Business Actually Needs vs. What It Feels Like It Needs
You want to fill the hole. What you need is to shrink to fit your actual cash position and grow again when you're ready — not when the calendar is pressuring you.
First thing I'd look at is your foreman situation.
A working foreman — a guy who runs three guys and still gets dirty, still pulls tools, still knows which end of a wrench does what — that person survives a downturn. He's billing hours while he's managing. If you've built out a layer where your foreman schedules, takes calls, and coordinates but isn't touching tools, that's overhead with a title. Good times cover it. A 30% revenue drop doesn't.
Make the staffing decision from the runway number. Not from how long the guy's been with you.
When I sold Smith Mechanical to Reliant in 2018, the business was leaner than it had been at the six-truck peak in 2008. Worth more per truck than it had been then. The rebuild after the crash didn't recover by returning to what I'd built before — it recovered by becoming something different. More service-heavy, tighter receivables, no single account over 20% of revenue.
Losing those two builder accounts in October 2008 is what forced that. I wouldn't have done it any other way. I was measuring success in trucks. Nobody redesigns that voluntarily. The crash redesigned it for me, and I spent a long time being angry about that before I could see it for what it was: the thing that kept me solvent.
Monday Morning: What to Actually Do With the Next Two Weeks
Day one. No sales calls. Open your bank account. Pull your outstanding invoices sorted by age. Write down your fixed monthly number — truck notes, insurance, rent, payroll at current headcount. Divide cash on hand by that monthly number. That's your runway in months. Write it on paper. That number runs every decision for the next 60 days.
Days two through five. Call your best current clients — not to sell them something, to collect anything collectible. Completed work sitting on invoices over 30 days: call and ask for payment this week. Not next Friday. This week. Slow-pay accounts get a written request with a deadline. Certified mail if the dollar amount justifies it.
Week two. Look at your cost structure before you look at your pipeline. What's fixed, what's variable, what can you cut without cutting production? The foreman question lives here. So does any equipment lease on gear that isn't billing right now.
Then — only then — think about what new work you actually want. Not what volume you need. What work you want: payment terms, client type, job size you can control from walkthrough to final invoice.
Don't sign anything with a pay-when-paid clause for at least 90 days. That's the minimum time it takes to stop making decisions from fear. You're not thinking straight yet. You're not supposed to be — you just took a real hit. Give it 90 days before you hand a GC that kind of leverage over your cash flow.
FAQ
I just lost my biggest account and I have payroll in ten days — what's the actual first call I make?
Your bank. Not a potential client, not a new GC — your bank. If you have a line of credit, confirm it's available and what the draw process looks like. If you don't, know that now rather than day nine. After the bank, call anyone with an invoice over 30 days and ask for payment this week. Not next week. Ask them directly. Most people pay when you actually ask. Then make payroll. Then start reading your numbers. In that order.
How do I know if I was genuinely dependent on this client or just in a rough patch?
Pull the last 12 months of revenue by client. Any single account over 25% of total billings — that's dependency, not a rough patch. If that same account also represented most of your receivables growth — billing going up while cash came in slower — that's the double exposure I had with my 2008 builder accounts. Dependency and a slow bleed at the same time. The rough patch question is a way of avoiding the answer. Pull the numbers and look at them.
Is there a right size for any single client as a percentage of revenue?
Nothing universal, but in my own shop I got uncomfortable anywhere over 20% from a single source. The risk isn't just the loss. It's what concentration does to how you handle them. At 35% of your revenue, you stop pushing back on slow payment. You stop walking jobs you should walk. You stop being their sub and start being their dependent. Twenty percent is a number where losing them hurts without killing payroll.
My lost account was a commercial property manager, not a GC — does any of this apply?
Yes. Property managers often pay slower than GCs because their approval chains run through ownership groups, out-of-state accounts payable, management software. The dependency trap is identical — concentrated revenue, a relationship that makes hard payment conversations feel dangerous. Lien rights on commercial property run differently than new construction and vary by state, so look that up before you're past your deadline. The cash position question, the runway math, the panic warning on replacement volume — all of it applies exactly.
What if the account didn't leave — they just went quiet?
Treat it as gone for cash planning purposes right now. Don't wait for confirmation. A clean break you can plan around. Silence is a cash flow leak with no closing date. Pull the runway number, stabilize your AR, look at your cost structure. If they come back, fine — you made good decisions in the meantime. If they don't, you didn't burn three months waiting to find out.
Silence is the answer. Act like it.
Enjoyed this article?
Get articles like this in your inbox every Monday. Free, no spam.