Open in interactive viewer → charts, metric popovers & call review

Xero Puts a Price on Its AI

A guidance upgrade, an A$1 billion buyback and a CFO exit all land on the same morning Xero finally admits it knows how to charge for agents
XRO.AX · Earnings Call · 2026-11-11

The hinge hiding inside a good half

The numbers were strong enough to speak for themselves: NZ$1.745 billion of operating revenue, up 46% reported and 27% pro forma, adjusted EBITDA of NZ$362 million (+23% pro forma), pro forma free cash flow of NZ$238 million (+32%), and a Rule of 40 score of 36%. Xero then lifted FY27 revenue to NZ$3.77–3.83 billion, raised the bottom end of its EBITDA range to NZ$880–920 million, and pushed its FY28 revenue aspiration from "more than NZ$4.2 billion" to "more than NZ$4.6 billion" — with a further NZ$70 million of revenue synergies still excluded from that number. But the loudest line item was capital return. The board approved an on-market buyback of up to A$1 billion over 12 months — against a company valued at roughly A$8.9 billion, that is more than a tenth of the equity, funded from excess cash and cash generation while holding net debt at 1.1x pro forma EBITDA. Claire Bramley, on her final call before leaving for what the CEO described as a larger CFO opportunity, was blunt about the logic: “we don't believe that the current share price is a true reflection of the value and future growth opportunity at Xero” — Claire Bramley, Chief Financial Officer (CFO) · 2026-11-11. Note the sequencing — a A$550 million dilution-neutralisation programme was completed, the undrawn revolver was repaid, and only then did the company graduate to shrinking the share count outright. The awkward part is that the same CFO who architected the Melio payments integration and the capital framework is walking out the door in January. Bramley framed her own epitaph in the prepared remarks — “That includes our cash generation, our disciplined approach to capital allocation, the Melio acquisition and today's share buyback” — Claire Bramley, Chief Financial Officer (CFO) · 2026-11-11 — and the very first analyst question was a numbers question directed at her, politely acknowledging it was probably her last. A buyback is a statement about the future; the person who would normally defend it on the next three calls will not be there.

From "we're not charging" to three price routes

The genuinely new material is not financial. Six months ago, asked directly whether Xero would monetise AI, the CEO was unambiguous:

we are not monetizing AI this year explicitly... I don't think we have landed, Paul, yet on what model we will use this year.

Sukhinder Cassidy, Chief Executive Officer · 2025-11-13
Pressed on XeroForce specifically, she volunteered: “I freely admit I don't know yet how we're going to charge for Xero Force.” — Sukhinder Cassidy, Chief Executive Officer · 2026-05-14 This week that ambiguity was replaced with a framework.

We see 3 routes that we can monetize this AI transformation for customers... We are at the start of AI monetization.

Sukhinder Cassidy, Chief Executive Officer (CEO) · 2026-11-11
Route one is bundling: high-frequency features such as auto bank rec — around 90% of users returning monthly, more than 100 million transactions reconciled, 96%-plus accuracy — sit inside existing plans and support price. Route two is consumption-based credits and add-ons, already being tested on lower-priced UK tiers where price sensitivity is highest. Route three is solution selling: Xforce, an umbrella brand that wraps the original XeroForce agent builder with forward-deployed engineers and an "AI teammate" modelled explicitly on Cognition in engineering and Harvey in legal, with Azets, Nexia and Shepherd Hensman as design partners. The segmentation logic is the interesting bit — it is a size story, not a market story. “there may be, example, more desire for metering, understanding, and willingness to pay the larger small business... whereas the sole proprietor... maybe consumption metering is too much versus certainty.” — Sukhinder Cassidy, Chief Executive Officer (CEO) · 2026-11-11 That is a mature pricing argument from a company that has historically taken a granular, one-size-per-segment approach to price rises. Management was also careful to say the monetisation is not in the upgrade — analysts asked twice and were told not to read any of it into the guidance. The AI monetization story is optionality, not yet revenue. One global thread worth flagging: asked whether Xforce implies building on open-weight models, Sukhinder Singh Cassidy said “We are model agnostic. And we are built to take advantage of both proprietary and open weight models.” — Sukhinder Cassidy, Chief Executive Officer (CEO) · 2026-11-11 That puts Xero squarely inside the market's open weight model debate, and its "AI teammate" ambition sits alongside the broader personal AI agents theme that has been climbing the global keyword tables.

What Xero quietly stopped saying

The most revealing artefact is what fell out of the company's own vocabulary. Two quarters ago, subscriber growth was the single top-ranked theme for Xero. It is now absent from the top ranks entirely — replaced by average revenue per customer, gross profit dollars and cash. Likewise long idle subscriber cleanup and business edition mix, two subjects that dominated 2024 and 2025 calls, have vanished conceptually. The pivot is deliberate: ARPC reached NZ$62.21, up 25% (20% constant currency), of which Melio contributed NZ$5.81 and organic ARPC grew 14% — and management stressed that ARPC is per business, not per seat, so AI-enabled headcount reductions at customers cannot dent it. That reframing does real work. It lets Xero accept a lower gross margin percentage (76% pro forma, down 3.1 percentage points, purely from payments mix) while arguing the only number that matters is absolute gross profit dollars. It also explains why core Xero margins were described as "stable" three separate times under analyst pressure about the U.S. — the subscription gross margin is fine, the mix is the story. Payments are now 22% of revenue versus 15% two years ago, and CashFlow Central has 43 partners live including U.S. Bank, with PNC slated for calendar 2027. The one metric that moved the wrong way got new disclosure rather than a new answer: monthly recurring revenue churn of 1.19% is modestly above pre-pandemic levels, but cohort churn was broadly flat at 0.83%, and management attributes the gap to faster-growing international and direct channels. They have now broken out regional churn to prove it. Bullish or defensive, depending on your prior.

The questions the upgrade did not answer

Two tensions linger. First, revenue went up but EBITDA guidance only moved at the bottom end, because a weaker New Zealand dollar lifts both revenue and the offshore cost base, Melio adds gross profit at a lower margin, and management chose to reinvest some upside into AI capability and cloud infrastructure — including “tactical onetime investments in platform resilience and cloud infrastructure stability” — Claire Bramley, Chief Financial Officer (CFO) · 2026-11-11. That line reads very differently if you remember the May call, when the same executive was apologising for intermittent outages during Australian tax filing season and offering customer credits. Second, Melio's adjusted EBITDA breakeven is still pinned to the second half of FY28 despite accelerating payments growth, and analysts pushed on why it could not come sooner. The answer — that FY28 confidence is spread across the whole business, "not 100% Melio" — is reasonable but evasive in the same way it was a year ago. Meanwhile unaided U.S. brand awareness rose 4 percentage points, which is real progress on a multi-year spend with deliberately deferred payback. So: a company hitting an inflection on cash, buying back a tenth of itself, and putting its first real price list on AI — while losing the CFO who built the model and asking investors to trust a lower-margin, higher-dollar future. The buyback says the board believes the story. The next two quarters have to prove the story can be told without Bramley telling it.