When to defer (and the cost of deferring)
Every capex item can be deferred. The question is what deferral costs. Owners and asset managers respect deferral analysis. They lose respect for "we should do this" without a number attached to "we should not do this." The honest deferral conversation is one of the highest-leverage skills in operations — and one of the rarest.
The three deferral costs
Cost one: physical degradation. The asset gets worse during the deferral period. Carpet that should have been replaced in year 1 looks tired in year 2 and embarrassing in year 3. The replacement happens eventually, but the property absorbed two years of guest impression damage. The cost is real but hard to quantify — it shows up in NPS, in OTA reviews, in conversion rate on the brand site.
Cost two: price escalation. The €380k corridor refurbishment quoted in year 1 is €420k in year 2 and €470k in year 3 because materials inflate at 4-6% and labour at 5-8%. Deferring by two years costs the property €90k of extra spend on the same scope. This cost is quantifiable and should appear in every deferral memo.
Cost three: cascading failure. The HVAC chiller deferred from year 1 to year 2 fails in month 14, six months before the planned replacement, during peak summer. The emergency replacement costs €185k vs. the planned €120k, plus three weeks of partial AC failure, plus the guest comp bill, plus the negative reviews. This cost is high-variance — sometimes deferral is fine, sometimes it triggers a five-figure recovery — and the asset manager wants to see the engineering judgment on which risk level applies.
When deferral is the right call
Three conditions need to be met. The deferral cost is genuinely modest (price escalation is manageable, no cascading failure risk, the degradation is below guest-visible threshold). The capital being freed has a higher-yielding use (a PIP that prevents brand termination, a ROI project with a stronger IRR, a debt paydown that the lender is pushing for). And the engineering team signs off in writing that the risk is acceptable.
I deferred a guestroom soft goods refresh at the Bodrum property from year 1 to year 2 because the FF&E reserve was needed to fund an emergency kitchen ventilation replacement. The deferral cost: €34k of price escalation and a half-percent NPS drag for one summer. The alternative cost: closing the buffet restaurant for 11 weeks at the start of high season. The deferral was obviously correct. The memo to the owner ran one page. The owner approved by email within an hour.
When deferral is the wrong call
The deferral memo
Every deferral generates a one-page memo: the item, the original year, the deferred year, the three deferral costs (each quantified or labelled "not material" with reasoning), the alternative use of capital, the engineering sign-off, and a single recommendation line. The asset manager reads it, the owner approves or pushes back, the decision is logged in the capex tracker. Six months later when someone asks "why did we defer the corridor work," the memo is in the file. That answer matters more than the decision itself.