£24m a year of incremental upgrade revenue
A small number of changes to the upgrade programme already running across the network would add £20.5m incremental revenue a year for DfTO operators, on top of what upgrades earn today. The changes are simple: adjust three settings within a week. None of them needs new technology, a fare change or a subsidy.
This paper sets out what the opportunity is worth by operator, what needs to be adjusted, and how it is controlled. We want to run it with you; every lever measured against a control from day one, reported monthly, reviewed together at the end of March.
The £24m comes from three levers
Three small changes to how upgrades are run add £24m a year across the nine operators already live. None of them needs new technology, a fare change or a subsidy. We can push these changes live within 1 week.
The number comes from measuring every First Class service on Seatfrog routes from 1 June to 13 September 2026 (4,226 services, 166,707 departures) and sorting each by how full Standard is and how much of that demand First Class converts. The table shows where each change lands by operator.
18.5% of services (nearly one in five) run with no upgrade available. These are decided manually. Run at least one seat on every eligible service and sell to the significant demand already on each service.
9.1 million First Class seats a year travel empty on services where Standard is filling. Release one in twenty empty seats (about half a seat per service per day) and the revenue generated will be significant.
Bring SWR and GTR live on the platform to generate significant revenue for the DfTO group. Based on a highly conservative model already presented to each TOC.
| Operator | DfTO ownership | 1 · Upgrades on every service | 2 · One seat in twenty | 3 · New operators | Incremental revenue annually |
|---|---|---|---|---|---|
| LNER | Yes | £2.07m | £2.29m | — | £4.36m |
| TPE | Yes | £0.20m | £0.56m | — | £0.76m |
| Greater Anglia | Yes | £0.14m | £0.38m | — | £0.52m |
| SWR | Yes | — | — | £0.72m | £0.72m |
| GTR | Yes | — | — | £2.80m | £2.80m |
| Four operators transferring to DfTO shortly | No | £3.00m | £11.62m | — | £14.62m |
| Two operators outside DfTO | No | £0.08m | £0.17m | — | £0.25m |
| All operators | £5.49m | £15.02m | £3.52m | £24m |
Gross a year incremental: reservations, June to September 2026. SWR and GTR data taken from each operator's financial model, year one.
Three decisions worth £24m a year
Run every operator on the agreed framework
Nearly one in five services runs with no upgrade on sale at all: 18.5% of service-dates. A third of those are trains where Standard is already more than 40% full. The decision is made by hand, service by service, and the seats go empty.
The framework agreed with DfTO in April replaces that manual call with a set of controls. Seats are released from live sales data, only where they are forecast to be empty at departure, on every eligible service. Guesswork comes out, revenue goes up, and the operator keeps the controls.
Release 5% of empty First Class seats where Standard is filling
Seats are capped by hand, from last month’s load, which already includes the upgrades sold. On these services 9.1 million First Class seats a year travel empty: 5.7 million where Standard is busy, 3.4 million where it’s quieter. Demand is on the train and the cabin stays shut.
One seat in twenty is released, only where it is forecast to go empty at departure using real time data, not the prior month. About half a seat more per service per day. Priced to the service, so a quiet train sells cheaper than a busy one, and the operator sets the floor.
Bring SWR and GTR live
Two DfTO operators run First Class with no digital upgrade channel. Every empty seat goes unsold, every day, to a passenger who would have paid for it.
Live within about four weeks of a decision, as Greater Anglia and TfW were. Same controls from day one: seats released to forecast, reserves on the framework, monthly reporting.
Three ways into First Class
81% of First Class tickets are sold before the day; every Seatfrog upgrade is bought in the final 24 hours before departure.
Timeline: illustrative upgrades on one service.
A different customer.
2.5m+ passengers who buy in a different way: 185,000 pieces of feedback.
Seatfrog brings a national audience to each operator’s First Class: budget-conscious commuters and leisure travellers who cannot afford to travel First Class at the fare. 96.5% come from non-operator channels, and the competition in auctions means 73.1% of them do not win.
Across 185,000 pieces of customer feedback, three things decide whether a passenger upgrades:
- Price. 83% of customers say price decides whether they upgrade. The first question is always “is this a good price?” and “is it worth it?”
- Value. What kind of day have they had, how do they feel? What do I get in First Class? 52% say they aren't sure First Class is worth it or what they get, and that the day they've had, the journey length or time of day is a big driver.
- Certainty. Will they win, how many others are bidding, how much time do I have? Uncertainty is the mechanic, and it is also the biggest reason people don’t bid.
2026 Seatfrog upgrade research covering surveys of 17,500+ unique customers and 185,000+ pieces of customer feedback through Seatfrog Pulse always-on research, alongside moderated user research and testing.
Manual restrictions on upgrades show declines in total revenue and First Class load factor.
Test: remove upgrades from 395 services to see whether more full-fare First Class tickets would sell.
Outcome: upgrade revenue fell £3.6m annualised. First Class ticket sales decreased faster on services where upgrades were removed.
Has removing upgrades increased First Class ticket sales? No.
First Class bookings, 4–31 July against 1–28 August, same services and days of the week, excluding upgrades. Bookings, not revenue.
Test: raise First Class reserve prices 30% to drive higher total revenue.
Outcome: each seat earned 30% more in yield, but a third fewer sold and total revenue fell 11%.
Every decision has a floor, a cap, a control group and a stop.
Each lever is measured at service level against a matched control, reported in real time, and paused if First Class ticket revenue moves. DfTO and the operators keep every control.
- At least one seat on sale on every service with empty First Class seats.
- Reserve set by the framework.
- Reserves adjusted on thin-demand services to lift participation.
- Excluded services and operational rules stay in place.
- Seats released only where live RARS data forecasts the seat travelling empty.
- Capped at 5% of empty First Class seats. 95% untouched. The 5% can rise if the operator or DfTO wants to further increase revenue.
- Released inside the final 24 hours, after 81% of First Class tickets are sold.
- SWR and GTR start on the same framework, controls and reporting from day one.
- Control group in place before the first auction runs.
- Live within four weeks.
Control group
A matched set of services, so every number is a difference against control.
Real-time readout
Upgrade revenue, First Class load and First Class ticket sales, by operator, inside Seatfrog Intelligence.
Pause rule
Any lever pauses if First Class ticket revenue falls against control beyond a threshold agreed with the operator.
Measured today
Win rate, participation, yield and revenue per auction on every service.
A decision by 31 October allows go-live in November and a joint review at the end of March
to start
by 31 March
incremental a year
How the numbers are built
- Decision 1 (£5.49m): £5.08m from running auctions on the 18.5% of service-dates that carry none, each priced at the 75th percentile of revenue per auction for its operator and demand band; £0.41m from lowering reserves by about £5 on quieter services, where every £1 off the reserve raised revenue by 4.4% in targeted pricing trials on quiet services.
- Decision 2 (£15.0m): 5% of empty First Class seats released, priced at the 75th percentile of revenue per seat for each operator and demand band: £7.44m where Standard is over 40% full and First Class is 10 points or more behind similar services; £4.67m on a set of services 20–40% full; £2.90m on services already converting but turning bidders away.
- By 31 March (£3.6m): LNER, TPE and Greater Anglia from November and GWR from 13 December, at the full annual rate from go-live.