Frontdoor's $9 Million Summer

The hottest summer on record cost Frontdoor nine million dollars of gross margin — and the stock still printed a 90-day gain of 22%. The story is in what management quietly stopped saying.

FTDR · Earnings Call · 2026-11-05

Frontdoor's $9 Million Summer

Frontdoor reported its third quarter on November 5: revenue up 6% to $655 million, adjusted EBITDA up 5% to $205 million, net income up 10% to $116 million, adjusted EPS up 15% to $1.81. Shares have run 22% in ninety days and sit near the high of a year-long climb. The headline is a boring beat-and-raise. The substance is three things management told you, and one thing it stopped telling you.

The heat bill, and the proof it bought

This was the quarter the margin story got stress-tested. In the second quarter the company banked a roughly $5 million weather benefit; it warned that would reverse. It reversed with interest.

This resulted in gross margin of 55%, a 200 basis point decline from the prior year period, primarily driven by $9 million of unfavorable weather, more than reversing the favorability we saw in the second quarter.
Jason Bailey, Senior Vice President and CFO · 2026-11-05

And still adjusted EBITDA landed above the midpoint of guidance. That is the whole thesis in one quarter: a 200bp gross-margin hit absorbed without cutting service, without cutting claims, and without an earnings miss. The company's own language for the mechanism — operational improvement, its single highest-momentum keyword this quarter — is vague, but the compounding process improvements behind it are not: job routing inside the preferred contractor network, state-by-state HVAC pricing, an virtual technician that resolves some issues without a truck roll, and five-star/star rating scores at record highs and lows respectively. The structural version of this shows up in the filed numbers. Gross margin finished the June quarter at 58.6%, up 13 points over three years, which is why the company felt able to lift its long-term adjusted EBITDA margin target into the mid-20% range and defend it through a record-hot summer.

What actually broke: real estate

The change this quarter is the housing channel. Three quarters of language tell the story. In April the CEO was citing a 6% attach rate in March and eight consecutive quarters of improvement. In August, real estate member count grew 7% and he called it “a good showing for Q2” — William Cobb, Chairman and CEO · 2026-08-06. On this call, that channel got guided down.

Make no mistake, the housing market is effectively frozen, and conditions deteriorated as we moved through the quarter.
Bill Cobb, Chairman and CEO · 2026-11-05

The specifics matter: prices up 39 straight months to roughly $420,000, mortgage rates back above 7%, the Pending Home Sales Index at its lowest on record, about 15% of signed contracts falling through in September, and existing home sales just under a 4 million seasonally adjusted annual rate. Frontdoor now expects real estate ending member count to be flat for 2026 rather than growing. That channel is only ~7% of revenue — but it was the fastest-growing one. The offset is the counterintuitive part, and it is genuinely interesting: a frozen market is a tailwind for the core. Fewer people move, so fewer people get a home warranty handed to them at closing and more people buy budget protection directly. Renewals edged back into growth at ~1.6 million members, direct-to-consumer member count rose 4% for an eighth consecutive quarter, and retention hit 79.7%, up 30 basis points and higher for three straight quarters. Total member growth of about 1% survives despite the housing channel going to zero. Note the trade-off though: DTC revenue fell 1% even as members rose 4%, because promotional pricing is buying volume with price — a deliberate swap the CEO says keeps paying because “the promotional cohort actually renewing at a higher rate than people who don't come into us with a discount” — Bill Cobb, Chairman and CEO · 2026-11-05. Better customer acquisition cost is the receipt.

The agent in the room

New this quarter, and clearly forced by the calendar rather than the P&L: agentic AI entered the narrative as the company's second-ranked theme, alongside AI agents. For context, the only AI language in prior quarters was about arming inside sales reps with tools. Now the market itself has a new keyword — Muse agent cracked the global top-50 this quarter — and Frontdoor is standing in the same wave. Analysts asked the right question, which is not "what's the upside" but "what happens when an agent makes cancelling frictionless?" The answer was more candid than the press release version: “We have people who try to cancel all the time... We track cancels very closely. It only turns out to be about 1% of the time.” — Bill Cobb, Chairman and CEO · 2026-11-05 And on whether this is happening now rather than someday: “I think it's mainly in the future. We have seen very little, and obviously we're quite attuned to this at this point, but I think adoption of personal agents has probably been a little slow.” — Bill Cobb, Chairman and CEO · 2026-11-05 That is refreshingly honest — a CEO naming a theme the tape cares about while conceding it does not yet touch the numbers.

What stopped being said

The quiet omission is dynamic pricing model. It was Frontdoor's number-one keyword as recently as the prior quarter and number two the quarter before that; in August the CFO attributed a raised realized-price outlook “mostly to the optimization around dynamic pricing” — Jason Bailey, Senior Vice President and CFO · 2026-08-06. This quarter it is gone from the top of the conversation entirely, replaced by state-by-state HVAC pricing, renewal process mechanics, and promotional cohorts. Realized price guidance actually went up to roughly 4%. So the engine may simply be running in the background — but the narrative has shifted from mechanism to outcome, and outcome is what they are selling. Two smaller things that slipped rather than broke: the appliance rollout, promised for the fourth quarter in February, is now a 2027 story ("a little slower by design" — contractor training), and the new HVAC system program keeps scaling but at a structurally lower margin. Non-warranty revenue grew 32%, on track for ~$175 million this year versus $13 million four years ago, on only 3% member penetration — but the CFO was explicit that “while that gross margin is lower than the home warranty business, it still drops through at a pretty high rate” — Jason Bailey, Senior Vice President and CFO · 2026-11-05. Mix is a slow, permanent headwind to the 58% gross margin, offset by SG&A leverage.

The cash-return kicker

$259 million of buybacks year-to-date, 3.7 million shares, about 5% of the float, 87% of free cash flow. The $650 million authorization completes a year early; the board replaced it with $1 billion over three years. Quarterly repurchases reached $91 million, up 42% year over year — and the stock is not cheap on this basis: 15.4x price to free cash flow, up from 10.6x two quarters ago. Interest coverage of 8.8x comfortably carries the $547 million net debt position. What to watch is simple. The fourth-quarter guide implies revenue of $453–463 million and adjusted EBITDA of $62–72 million — a seasonal margin trough — so the full-year confidence rests on renewals continuing to grow, DTC holding its promotional cadence, and real estate merely going flat rather than shrinking. Everything else — the HVAC runway, the appliance trade, the agentic optionality — is a 2027 argument. This was a quarter where the operation proved itself and the market proved itself harder to read.