The setup
This Hopper case was about pricing a trip protection add-on, not in theory but in unit economics. I compared 80 percent and 90 percent refund coverage across premiums, attach rates, and loss ratios to see which variant actually makes money per offer. The goal was to pick a static premium, then sketch smarter experiments on top of it.
The actual question
On the surface, 90 percent coverage is the "nicer" product and usually drives higher attach rate. The real question is whether that extra demand ever compensates for how often Hopper has to pay out. Which variant wins on profit per offer once you factor in both price and risk, and where does 90 percent become worth the volatility?
The analysis
At current prices, 80 percent coverage is the clear winner. It earns about $0.80 profit per offer versus $0.67 for 90 percent, with a healthier 72.6 percent loss ratio versus 81 percent. Attach rate is slightly lower for 80 percent, but the extra demand for 90 percent doesn't offset how expensive those payouts are at low premiums.
80% stays inside the healthy 65–75% band. 90% only crosses into safe pricing past 17% premium.
Across the premium range, 80 percent looks like a stable workhorse. It mostly sits in a 65 to 75 percent loss ratio band, while 90 percent swings hard and stays loss-making around the 10 to 11 percent premium mark. 90 percent only starts to beat 80 percent on profit per offer once you push premium to roughly 17 percent of fare.
90% only beats 80% above 17% premium. Below it, 90% bleeds money.
The attach-rate charts explain the risk story. Travelers love 90 percent coverage when it's cheap, so attach rate spikes right where the economics are worst. That combination of high attach and thin pricing makes 90 percent a high-beta SKU: great upside when priced high enough, very ugly downside when it isn't.
Because of that, my static picks are: for 80 percent coverage, a 15 percent premium where profit per offer is around $1.07 with a loss ratio of roughly 66 percent and a 5.9 percent attach rate. For 90 percent coverage, a 17 percent premium where profit per offer jumps to about $1.39 with a 62.3 percent loss ratio and a 6.1 percent attach rate. At current prices overall, the data says 80 percent beats 90 percent about 86 percent of the time on profit per offer.
What I would do about it
First, I'd treat 80 percent at 15 percent premium as the default safe SKU, and use 90 percent at 17 percent premium as the "go big" option in markets where Hopper is comfortable with more variance. From there, I'd run two experiments. One is adding a fixed deductible, like "up to 90 percent refund minus $25," to cut loss ratios without killing perceived value. The other is moving toward dynamic pricing by capping high-risk users at 80 percent and reserving 90 or 100 percent coverage for lower-risk segments based on historic cancellation behavior.
What I would want to validate
I'd want to validate that adding a $25 deductible keeps attach rates roughly flat while meaningfully lifting profit per offer, and that the dynamic-pricing rules actually shrink loss ratios for repeat cancellers instead of just shifting risk around. I'd also re-run the profit-per-offer simulations as new cohorts come in to confirm that 80 percent stays the reliable baseline while 90 percent pays off only at higher premiums.
Hopper should treat 80 percent coverage as the dependable cash engine and only unleash 90 percent when pricing and risk controls are tight enough to handle its mood swings.