If a client is not paying construction invoices and keeps sending partial payments with a verbal promise for the rest, you are already in the danger zone. Tye Shumway, a Moab, Utah builder with 25 years in construction, went roughly six months and about $1 million deep with a single client who never refused to pay — he just paid slower, and paid just enough to keep crews on site.
The failure mode is not a client who says no. It is a client who says yes, later, while your payroll clears and theirs doesn’t. This article covers the sequence that got Tye to $1M, the stop-work rule he installed afterward that caught the next bad client before it cost him anything, and how he structured his business so one default couldn’t end it.
Key takeaways
- Partial payments that arrive "just enough to keep us going" are the warning sign, not the absence of payment. A client who pays 60% on time and promises the rest is buying your labor on credit you never extended.
- Set a written stop-work threshold before the job starts: an unpaid invoice means crews come off the site until it clears. Tye’s version of this rule caught a second bad client two or three times and, in his words, saved his bacon.
- One client running five of your simultaneous projects is not a great relationship, it is concentration risk. When that client stopped paying, Tye’s entire production schedule went down at once.
- Working against a client’s revenue deadline — a booked nightly rental, a scheduled closing — erodes your willingness to stop work. Tye finished one remodel at 2am wiping counters because the unit was already rented.
- Contracting revenue is the most fragile leg of a real estate business. Tye’s rental portfolio, not his construction company, is what funded the transition out of corporate in 2020.
From the Investor Fuel Show
This article draws on an interview with Tye Shumway on the Investor Fuel Show, hosted by Mike Hambright. Watch or listen to the full interview.
How a $1M Hole Opens Up Without a Single Refusal to Pay
Nobody hands a contractor a $1 million loss in one conversation. It accumulates in increments small enough that stopping feels like an overreaction each time.
Tye met the client in 2021. By 2022 he had five active projects with him — three remodels and two ground-up builds — plus two more ground-ups queued to start. That is a full production calendar committed to one counterparty.
Then the invoices started arriving slower. Not unpaid. Slower. Tye asked directly whether everything was good, got told yes, and kept crews moving because the work was already scheduled and the momentum was real.
The client kept paying something. That is the mechanism. A partial payment resets your internal clock — it feels like the relationship is functioning, so you extend another two weeks of labor, materials, and subcontractor exposure against the balance. Six months of that and Tye was roughly $1M deep in receivables that never got collected.
The tell to watch is not the size of the arrears. It is the ratio of promises to payments. When you are getting more verbal reassurance than money, you are financing the client’s business with your own working capital, and you did not price that risk into the job.
Practical read on this: track days-outstanding per client, not total dollars owed. Total dollars owed always looks survivable because you are comparing it to the contract value still ahead of you. Days-outstanding tells you the direction of travel, and the direction is what kills you.
The Stop-Work Rule That Caught the Next Bad Client
The policy Tye put in place after the loss is one sentence: an unpaid invoice means crews are off the job until it is paid. No exceptions, no case-by-case negotiation while you are standing on the site.
That last part carries most of the weight. The reason contractors keep working through nonpayment is not that they lack a rule — it is that the decision gets made in the moment, by the person with the most emotional investment in the relationship, usually while a client is explaining that the wire is coming Friday. Written before the project starts, the threshold becomes a term of the agreement instead of an accusation.
Tye’s rule got tested. Another client began repeating the same slow-pay pattern, and because the system existed, he stopped rather than rationalizing his way forward. He describes it plainly as having saved his bacon, and says it triggered two or three times.
Three things make a stop-work threshold enforceable rather than decorative:
- Put the number in the contract. Define what unpaid means — days past due, dollar amount, or both — so the trigger is arithmetic rather than judgment.
- Give the trigger to someone other than the owner. The person tracking receivables should be the person who calls crews off. Owners negotiate; operators enforce.
- Notify before you demobilize. The client should know the rule exists from day one and get written notice when the clock starts, so pulling crews is a consequence they chose, not a surprise.
Consult your own counsel on notice requirements and lien deadlines in your state before you write the clause — the mechanics vary and the timing matters.
He just gave us enough to kind of keep us going, until I realized — dude, my company’s going to go down if I don’t stop this.
— Tye Shumway, Moab, Utah builder and investor
Why One Client Owning Five of Your Jobs Is the Real Risk
The stop-work rule limits the size of the loss. Client concentration determines whether the loss is survivable at all.
Five simultaneous projects with one counterparty meant Tye’s entire production schedule had a single point of failure. When the payments stopped, he wasn’t pulling off one job and backfilling with others — he was pulling off everything, with two more ground-up builds queued that evaporated at the same time. Crews, material commitments, and forward calendar all went at once. Recovery required finding replacement projects from a standing start while carrying the receivable.
There is a second, quieter cost to concentration: it destroys your leverage to enforce the rule you just wrote. A client with five of your jobs knows you cannot afford to walk. You know it too, which is why the negotiation in the moment always ends with crews staying on site.
Deadline pressure compounds it. Tye’s client was putting the finished remodels straight onto the nightly rental market and had already booked guests. Tye describes finishing that first remodel at two in the morning, wiping down counters and checking light switches, because the unit was rented that day.
When you are working against someone else’s revenue calendar, their deadline becomes your deadline, and every conversation about money happens under a clock that belongs to them. That is a terrible position from which to demand payment. The fix is upstream: no single client should command enough of your capacity that stopping work would end your company.
Structuring the Business So One Bad Client Can’t Kill It
Tye describes his business as a three-legged stool, and the structure is the reason a $1M loss was a setback rather than a closure.
- The construction company — client work, the leg that generated the loss.
- The development company — raw land development and fix-and-flips, where he controls the timeline and the counterparty is the market, not a single payer.
- The real estate holding company — rentals producing income that does not depend on anyone approving an invoice.
The order those legs were built in matters more than the list. Tye had run the contracting business as a side hustle since 2011 while working commercial construction jobs. What let him go full-time in 2020 was not the contracting revenue — it was that the holding company had accumulated enough rental income to cover the household. The construction company was launched from a base of passive income, not the other way around.
That sequencing is the practical takeaway for anyone whose income is entirely client-dependent. Contracting revenue is the most fragile of the three legs: it is collected after the work is performed, it is concentrated in a small number of payers, and its collection depends on someone else’s solvency. Development income is lumpy but self-directed. Rental income is the only one that shows up whether or not a client answers the phone.
Tye’s read on corporate employment applies equally to a single big client — he learned in a previous job that an entire income can disappear in a day when someone else changes direction. A contractor with one dominant client has rebuilt exactly that exposure with extra steps.
Recovering: Systems, an Integrator, and Letting Go
The loss did something useful: it exposed where the company was actually weak. Tye’s phrasing is that it “showed the weak links in our company” — the receivables tracking, the missing thresholds, the absence of anything that would have flagged the pattern before month four.
He is also candid that building those systems is not his skill set. “I can put a bag on and build something all day long,” he says, but sitting behind a desk documenting a process is the hardest thing he does. This is where a lot of contractor recovery plans stall — the owner knows what needs to exist and never produces it.
His solution was a person, not a resolution. Shae Batt, his right-hand man, functions as the integrator to Tye’s visionary role: he watches how the work actually gets done, converts it into process, and — this is the part that matters — tells Tye to his face when he isn’t holding people accountable to the system he asked for.
The follow-through lesson is worth as much as the system itself. Tye describes calling Shae on a Thursday because nothing was on fire: crews were on job sites, materials were ordered, clients were happy. His instinct was to find a problem. Shae’s reminder the next morning was to stop creating fires that don’t exist. Tye left a meeting that day because his team had it handled.
If you build the process and then keep overriding it, you have not bought yourself anything. The point of the stop-work rule is that it fires without you.
Using Construction Skill as the Value-Add Edge
Tye calls construction knowledge a superpower, and he means something specific: most investors evaluate a property against what it currently is, while a builder evaluates it against what the lot and code allow.
Two examples of what that produces. His first investment was a 900-square-foot, two-bed one-bath home built in 1906 — unremarkable on paper. The lot was large. He added a triplex alongside the existing house and converted the garage into a unit, taking a single-family property to five doors.
The second was a two-acre parcel a partner picked up at a tax sale in 2019 and closed on in 2020. The partner had no plan for it. Tye’s read was that Moab needed housing for the seasonal workforce that arrives for the river and mountain bike seasons — people who need a room, not a family home. He initially estimated 30 to 40 units. The final project was a 47-unit condo complex of studios, one-bedrooms and two-bedroom micro units, and it fit legally without any upzoning.
The math scales down to ordinary rentals too. Mike Hambright described adding a bedroom and bathroom to a house he had owned for 13 years, on the logic that Dallas market pricing was running around $250 per square foot while the addition cost roughly half that. Deals he would have passed on years ago now pencil, because the spread between construction cost and market value per square foot has widened.
If you cannot do the work yourself, the substitute is a builder relationship close enough that you can ask “what fits here legally” before you make an offer.
Frequently asked questions
At what point should a contractor stop work on a project for nonpayment?
At the threshold you agreed to in writing before the project started — not at the point where the balance finally scares you. Tye Shumway’s rule after his $1M loss is that an unpaid invoice means crews come off the job until it is paid, with no in-the-moment negotiation. The specific number matters less than the fact that it is pre-agreed and enforced by someone other than the owner.
The trap is judging by total exposure rather than payment behavior, because total exposure always looks recoverable next to the contract value still ahead of you. Check your state’s notice and lien requirements with counsel before demobilizing, since the timing rules vary.
What are the early warning signs a client is going to stop paying invoices?
The clearest signal is partial payment paired with verbal reassurance. A client who pays enough to keep your crews on site while the balance grows is financing their project with your working capital, and that pattern almost always precedes a total stop. Tye’s client never refused to pay — invoices simply arrived slower over roughly six months while he was told everything was fine.
Two secondary tells: the client is working against a hard revenue deadline of their own, like a booked rental or a scheduled closing, and the volume of promises starts exceeding the volume of payments. Track days-outstanding per client rather than dollars owed, because days-outstanding shows direction of travel.
How many projects should one client represent in a construction company’s pipeline?
Few enough that you could stop work on all of them and still make payroll. Tye had five simultaneous projects with one counterparty plus two more queued, which meant a single default took down his entire production schedule at once.
Concentration also destroys the leverage you need to enforce a stop-work rule. A client who controls most of your capacity knows you cannot afford to walk, and so do you — which is why the conversation always ends with crews staying on site.
How can a contractor build income streams that survive a client default?
Separate the revenue that depends on a client approving an invoice from the revenue that doesn’t. Tye runs three entities: a construction company, a development company doing raw land and fix-and-flips, and a real estate holding company producing rental income.
The rental leg came first in practice. It was accumulated rental income, not contracting profit, that let him leave corporate employment in 2020 and go full-time on a construction business he had run as a side hustle since 2011. Contracting is the most fragile leg because it is collected after the work is done and depends entirely on someone else’s solvency.
Do you need to rezone a lot to add units to an existing single-family property?
Not necessarily — existing zoning often permits more density than the current structure uses. Tye added a triplex alongside a 900-square-foot 1906 house on a large lot and converted the garage, taking the property from one unit to five. On a separate two-acre parcel in Moab he developed 47 condo units, all fitted legally without any upzoning.
Zoning and density rules are entirely local, so the only reliable answer comes from your own municipality’s code and planning department before you write an offer. The general point holds: what a lot currently contains is rarely what it is allowed to contain.
The bottom line
Write your stop-work threshold into your next contract this week, and hand enforcement to whoever tracks receivables rather than keeping it yourself — that single change is what stands between a slow-paying client and a six-figure receivable you never collect.
