Anonymised, illustrative composite. A BC framing contractor's holdback was locked up for 55 days by law after a certificate of completion — payroll was not. A 13-week rolling forecast caught the week where those two timelines collided.
At a glance
A British Columbia framing contractor finished a multi-unit project and received a certificate of completion. The 10% holdback on the $540,000 contract — $54,000 — was withheld as required, but the money was not coming for a while, and the crew was already mobilizing onto the next job with its own costs starting immediately.
British Columbia’s Builders Lien Act fixes exactly how long that holdback sits: section 8(1) provides that where a certificate of completion is issued, the holdback period “expires at the end of 55 days after the certificate of completion is issued”, on top of the 10% retention section 4(1) requires in the first place. Payroll for 18 field staff does not pause for those 55 days, and the new job’s mobilization costs were landing in the same window.
Starting cash was $50,000. The forecast modeled a net weekly cash burn of $10,000 across the mobilization period — payroll plus new-job startup costs, net of other receivables coming in. Over seven weeks that is $70,000 of outflow against $50,000 on hand: a projected trough of −$20,000 in week 7. Fifty-five days is just under eight weeks, so the $54,000 holdback was due to land in week 8 — one week after the projected shortfall would have hit. The gap the forecast caught: exactly one week, five weeks before it would have happened.
The decisive fact was not accounting cleverness — it was reading BC’s holdback deadline correctly and mapping it against a real payroll calendar instead of assuming the money would show up when needed. British Columbia’s 10%-for-55-days rule is not the same clock as Ontario’s or Alberta’s: Ontario runs 60 days to preserve a lien and 90 to perfect it, and Alberta’s major lien fund holds for 60 days from substantial performance (90 for concrete or oil-and-gas work) — province-specific timelines that do not transfer from one job to the next if a contractor works across provincial lines.
Fifty-five days isn’t an arbitrary number either: BC’s Builders Lien Act gives anyone who supplied labour or materials up to 45 days after the certificate of completion to file a lien (s. 20(1)), and the extra 10 days in the 55-day holdback period is the margin the Act builds in for that lien window to actually close before the money moves.
With five weeks of lead time, the contractor arranged a $25,000 short-term line-of-credit draw timed to cover weeks 7 and 8, and repaid it in full the same week the $54,000 holdback posted. No payroll was missed, and no last-minute scramble was needed. The 13-week rolling forecast, tied to the certificate of completion date on every active contract, is now standard practice. For the tooling behind it, see building a rolling cash flow forecast for a contracting business and running Canadian construction payroll without the manual reconciliation.
Without the forecast, the same $20,000 trough would have arrived in week 7 as a payroll shortfall discovered in real time rather than five weeks in advance — at that point the only options are an emergency draw negotiated under time pressure, missed payroll, or personal funds injected on short notice, none on the same terms as a $25,000 line arranged with five weeks' lead time. $54,000 ÷ $10,000/week burn is roughly 5.4 weeks of runway the holdback alone would fund once released — confirming why a short bridge, not a permanent facility, was the right size of fix.
Build the forecast around the specific dates a holdback release actually depends on — the certificate of completion date, plus the province's fixed clock (55 days in BC, a different count elsewhere) — rather than a generic "money in 60 days" assumption. The gap that matters is not the size of the holdback, it is the number of weeks between when cash runs out and when the holdback is contractually available.
A $25,000 draw arranged five weeks ahead of the projected trough is a routine short-term facility a lender can underwrite on the contractor's normal terms; the same $25,000 requested with days of runway left is an emergency request, typically priced higher and harder to guarantee approval on in time. The forecast's real value was not the $25,000 figure itself — it was converting a crisis-shaped decision into an ordinary one, simply by moving the point of discovery five weeks earlier than the point of actual need. The contractor now runs the same 13-week model against the certificate-of-completion date on every active contract the moment that certificate is issued, not only when cash looks tight.
A 30-minute call is enough to tell you whether AI pays for itself here.
