What actually triggers an add-on in a transfer fee

The fixed fee is the only committed number

When a transfer is reported as "€40m plus add-ons" or "up to €50m", the only part that is committed on the day the clubs shake hands is the fixed fee. Add-ons are extra payments that fall due only if conditions written into the contract are met. If they are not met, nothing is paid. The "up to" figure is a ceiling, not a price.

The awkward part is working out which number you are being shown. The reporting of Cristian Romero's move from Tottenham to Atlético de Madrid on 12 August 2026 is the illustration to hand. Sky Sports described "a package in the region of £34.2m (€40m) with add-ons", and Fabrizio Romano called it a €40m package, add-ons included. Cadena SER's El Larguero, meanwhile, reports the clubs settling on €35m fixed with €5m in variables. None of that conflicts, but one deal is being published two ways: some outlets give the ceiling, others give the split. The body text tells you which you are looking at. Follow the headline alone and €40m reads as the guaranteed fee.

None of those figures came from either club; all of them came from journalists. Whether or not an official announcement follows, one deal was circulating at two different levels of detail on the same day. Everything below about Romero is reporting.

The money goes to the selling club, not the player

Add-ons are an arrangement between two clubs. The €40m in "€40m plus add-ons" is income for the seller and says nothing about what the player earns. UEFA's regulations draw the same line in the accounting: payments to the player, such as signing-on fees, are staff costs rather than part of the cost of acquiring his registration. Since the word "bonus" gets used for both, work out which one a report means before anything else.

The structure usually starts with the buyer's problem. Tying part of the price to results limits the damage if the signing does not work, and it keeps cash free now. For the seller it is mostly something to accept rather than something to ask for, though it does retain a claim on the upside after letting a player go cheaply. In practice the most common use is simpler than either: the gap that cannot be closed in the fixed fee gets pushed into conditions instead.

Appearances, team success and new contracts

Individual contracts stay private, but there is one unusually good window into what conditions clubs actually write. Accounting standards require unsettled obligations of this kind to be disclosed, and because Manchester United are listed in the United States, theirs are published where anyone can read them. At 30 June 2025 the outstanding total was £135.8m: roughly £117.2m tied to appearances, team success and new contracts; £3.7m to international appearances; £13.6m to awards and future transfers; and £1.2m to everything else. A year earlier the total was £115.6m; nine months later, at 31 March 2026, it stood at £144.4m. It does not simply climb: at 31 March 2025 it was £138.8m, above the June figure, because signings, conditions falling due and revised estimates all move it in both directions. What the number does show is that one club can be carrying well over £100m of unrealised conditional payments.

A caveat on reading it. The disclosure bundles conditions together: "awards and future transfers" is one line, and the largest category combines three things at once. So it cannot tell you what a Ballon d'Or clause is worth on its own, nor what the new-contract portion is. What it does establish is that the overwhelming share of the money sits in the appearances, team success and new contracts category. Even at that resolution, the centre of gravity is clearly the everyday conditions rather than the eye-catching ones.

Inside the game these conditions get split into "easy" and "difficult" ones, with appearance counts in the first group and a Champions League win or an individual award in the second. That is trade shorthand, not a defined category. No governing body recognises it and there is no threshold that separates the two. When a report calls an add-on "easily achievable", that is the journalist's read, not a contractual label.

Add-ons are also capped, and the cap is what United put in this note: "the maximum additional amount that could be payable", which is not the same thing as the amount they expect to pay.

What UEFA counts is what has become payable

There are two accounting treatments here, not one, and they do not agree. How an add-on is handled inside UEFA's financial rules is not how it is handled in a club's own accounts. Running the two together produces nonsense.

Take UEFA first. Article 4 of the Club Licensing and Financial Sustainability Regulations (2026 edition) defines the cost of a player's registration as fixed transfer compensation, "realised conditional transfer compensation for amounts which have become payable during the period", and other directly attributable costs such as agents' fees. The load-bearing word is "realised". Probability does not count. Until a condition is met and the money is owed, it is not part of the registration cost for UEFA's purposes. The squad cost rule measures wages, registration amortisation, net loan costs and agents' fees against income, so an unrealised add-on does not push that ratio up. Once a realised add-on joins the registration cost, it feeds through into amortisation from there. While it stays unrealised, a club can add to the maximum headline value of a deal without adding anything to its registration cost in UEFA's eyes.

A club's own books work differently. Manchester United estimate the fair value of contingent consideration at the date of acquisition, based on the probability of the conditions being met, and carry it in the cost of the registration; amounts not yet recognised are added once payment becomes probable. Money can therefore sit in the asset before any condition has formally fired. So "nothing hits the books until the condition is met" is not true as a general statement about football accounting. What stays out until the condition fires is the registration cost as UEFA measures it.

The regulations also say what happens when the two diverge. Article 67(6) requires a club whose annual financial statements do not meet Annex G's accounting requirements to submit restated financial information that does, along with a management declaration and an auditor's report on it. A club can capitalise an add-on early in its own accounts; for licensing purposes it still gets restated down to what has actually become payable.

Unrealised amounts do not disappear from UEFA's view either. Article 71 is the licensing test for having no overdue payables to other clubs, and the transfer information submitted under it must include conditional amounts not yet recognised on the balance sheet. That submission goes to the licensor — the national association, or a league to which the licensing system has been delegated — rather than to UEFA directly.

Note also that the threshold differs depending on which way the money flows. United capitalise what they may have to pay once it is probable, but do not recognise contingent amounts receivable on players they have sold until receipt is virtually certain. At 30 June 2025 the amount of such receipt considered probable was £nil, and the same note confirms that contracts entitling them to further payments do exist. Only the probable amount gets a figure, so "we are still owed a fortune in add-ons" never becomes a number in a set of accounts until it is nearly banked.

A player's own performance bonuses travel by a different route again. They enter the squad cost ratio as staff costs, and never touch the cost of the registration at all. For the wider financial picture, see our piece on financial sustainability and FFP and the one on the squad cost rule.

Leeds, Leipzig and the end of a season

Disputes over conditional payments are rarely about vague wording. They are about deadlines and definitions.

On 25 January 2020, Leeds United signed an agreement to take Jean-Kévin Augustin on loan from RB Leipzig until 30 June. Their own announcement two days later described an option to sign him permanently; the agreement in fact carried an obligation to buy him for €21m, payable in three instalments, if Leeds went up at the end of the 2019-20 season. Covid stretched the season and Leeds' promotion was sealed in July. Leeds argued the window for the condition had closed; Leipzig argued the obligation had been triggered. FIFA found for Leipzig on 1 June 2021, and on 4 November 2022 the Court of Arbitration for Sport dismissed Leeds' appeal, holding that the obligation had been triggered at the end of the 2019-20 season even though that season had finished later than expected because of the pandemic. The clubs announced an amicable resolution on 5 December 2022; the terms are confidential.

Whether a club won promotion is about as clear-cut as a condition gets. It still took nearly three years, because what moved was not the condition but the calendar underneath the phrase "end of the season". A purchase obligation is not an add-on, but it is written the same way, and it goes wrong for the same reason.

Inside Atlético's own add-on clauses

Atlético almost never put a figure in a transfer announcement. The numbers usually come from the other side.

Nahuel Molina's move to Roma is the most recent example. The club announced an agreement on 12 August 2026 without a fee. Il Messaggero put the fixed amount at €14m with €4m in bonuses, split into €2m for winning the Scudetto or reaching a Champions League semi-final and €2m tied to Molina's appearances and Roma qualifying for next season's Champions League. Sky Sport Italia has the fixed fee at €13m. Seeing the triggers themselves reported is unusual.

Álex Baena joined from Villarreal in July 2025. Villarreal published no figures, but EFE and Europa Press both reported €42m fixed plus up to €13m in variables. Of that €13m, €8m sits behind conditions described as straightforward — Atlético staying up, matches Baena plays and goals he scores were the examples given, not a complete list — and €5m behind trophies the club wins over the next five years.

Robin Le Normand's 2024 transfer is the rare one where a club official said it out loud. Real Sociedad president Jokin Aperribay, speaking at Javi López's presentation in late July, put it at "34.5 million plus five in variables". Atlético had chosen to negotiate rather than pay Le Normand's release clause. Reading the variables as the thing that bridged the gap the fixed fee could not is my inference rather than Aperribay's point, but as a demonstration of what add-ons are for I have not found a cleaner one.

Julián Álvarez, meanwhile, was reported at €75m plus €20m two years ago and the €20m has never been explained. I cannot find a single outlet that has reported what triggers it. The most expensive signing in the club's history, and the back end of the fee is a blank. That, more than any definitional problem, is why add-ons stay fuzzy for supporters. There is nothing complicated about the mechanism. Nobody publishes what is in it.

What the Griezmann clause showed

Atlético have been on the wrong end of a conditional-payment argument themselves, and it is instructive.

The two announcements of his 2021 return did not say the same amount. Barcelona's, on 31 August, described a season-long loan with the option of a further season and stated that "there is a compulsory permanent transfer clause". Atlético's, the following day, described a one-year loan with an option to extend renewable by either party, and mentioned no obligation at all. Neither was untrue; they disclosed different quantities of the same deal. Neither published the €40m figure or the playing-time condition attached to it.

The terms surfaced a year later. In September 2022 Cadena SER's El Larguero reported the contents of the agreement, saying it had obtained them: Atlético would owe €40m if Griezmann exceeded 50% of his available matches by the end of the loan, an available match being one where no medical report explained his absence, and a match counting once he played 45 minutes or more whether he started or came off the bench. He had already cleared 50% in the first season. Atlético's answer, per the same report, was to hold his minutes down in the second. That summer they had offered €20m to buy him outright; Barcelona refused and pointed at the contract.

How he was actually used is on the record. Griezmann entered in the 62nd minute against Getafe on 15 August 2022, the 62nd against Villarreal on 21 August and the 64th against Valencia on 29 August. Asked about it after those three matches, on 2 September, Diego Simeone said: "I am a club man." He did not deny it. The pattern held the next day against Real Sociedad, on 63 minutes, and in the Champions League against Porto on 7 September, on 61. None of them reaches 45 minutes.

It ended in October. On the 9th, Barcelona president Joan Laporta told the club's assembly the deal was €20m fixed plus up to €4m in variables, with Barcelona entitled to the difference up to the original €40m if Atlético later sold him. The following day Atlético announced the permanent signing, with no figure.

How far this generalises is limited. International caps, or another club qualifying for the Champions League, are conditions nobody in the deal can switch off. Playing time is the exception: whether that kind of condition fires depends on selection, and selection belongs to the buyer. When you read "up to €50m", the useful question is whose hands the conditions are in.

How add-ons differ from instalments and sell-ons

Three neighbouring mechanisms get mistaken for add-ons.

Instalments spread a fixed, certain fee across several years. The total does not move. "€30m over three years" and "€30m plus add-ons" describe completely different levels of certainty.

A sell-on clause gives the original selling club a share of what the buying club receives if it sells the player on later. Sky Sports report that Romero's move includes a 15% sell-on, though Sky do not say who holds it or whether it applies to the gross fee or the profit. I take that clause on its own in a separate piece.

A player's bonus is employment. The goal that triggers a seven-figure payment between two clubs and the goal that triggers a five-figure payment to the player are two entries in two different sets of books.

One more note, on Transfermarkt. You will see it repeated everywhere that the site records only guaranteed fees. Its published guidance for contributors does tell them to enter the fee as a figure in the fee field and to put contract information such as release clauses and add-ons in a free-text "Further Information" field, so the two are certainly kept apart at the point of entry. What I could not find is any published rule stating that the fee figures shown across the site uniformly exclude add-ons. Treat it as probably true rather than as something to cite.

And the reported fixed fee can simply be wrong. Viktor Gyökeres' move to Arsenal was reported in advance at €63.5m fixed plus €10m in add-ons, up to €73.5m. Sporting's filing to the Portuguese market regulator gave €65,764,103 fixed plus €10,256,410 in variables, a total of €76,020,513. Part of the difference was a financing cost Arsenal absorbed in return for paying earlier.

Atlético give no numbers and Sporting give them to the euro, and that is not one club being secretive and the other being generous. A listed issuer like Sporting can be required to disclose a material transaction under market rules; an unlisted club is not. Atlético are not withholding the figure so much as never being asked for it. The same kind of deal reaches the public at a different resolution depending on who was selling.

The rest of the transfer fee

This piece covers one layer: a fixed fee with conditional money on top. A real agreement also has instalment terms, an amortisation profile on the buyer's books, a route in through either a release clause or a negotiation, and deductions for solidarity payments to the clubs that trained the player.

The natural next step is the sell-on clause that appeared in the same Romero reporting. It pays out at a different moment and is calculated on a different base. If you would rather come at this from the regulatory side, start with financial sustainability and FFP.