The three questions that decide it
Every repair-or-replace call we make comes down to three questions. Answer them honestly about your roof and you'll usually land where we would.
- Is the problem local or general? One flashing failure is local. Granule loss across every slope is general. Repairs fix local problems; they can't fix general ones.
- How much paid-for life is left? A $1,500 repair on a roof with ten good years left costs $150 a year. The same repair on a roof with two years left costs $750 a year, and you'll still buy the replacement.
- Is this the first repair or the fourth? Repair frequency is the roof telling you its own diagnosis. Once fixes become a subscription, the math has already flipped.
When repair is clearly right
Storm damage on a healthy roof. A failed pipe boot or flashing detail at any age short of decrepit. A leak with a single traceable source on a roof that's otherwise holding its granules and lying flat. In all of these, the system is sound and one component failed; replace the component.
This describes most of the leaks we see, by the way. The average leak call ends in a repair, not a pitch. (What repairs run, honestly.)
When replacement is clearly right
Widespread granule loss, multiple slopes with curling or cracked shingles, leaks that move around, decking that flexes underfoot, or a repair bill quoting a meaningful fraction of a replacement. Past a point, repairs on a failing roof are interest payments on a loan you haven't accepted yet.
There's also the bundling effect: replacement resets everything at once (shingles, underlayment, flashing, boots), and those components fail on similar clocks. Replacing a roof that's failing generally costs less than chasing its failures one at a time.
The middle: how to decide under uncertainty
For the genuinely torn cases, our advice is to buy information before buying anything else. A thorough assessment (with photos, free from us) turns 'the roof is oldish and leaks sometimes' into 'two slopes are sound, one is failing, the flashing is from another era.' Decisions get easy when the facts get specific.
Then run the per-year math from question two above, and factor your own horizon: staying fifteen years favors replacing a borderline roof on your schedule rather than its schedule. Selling in two favors the documented repair. Whatever you decide, the Replacement Guide covers the road ahead if the answer is replace.