Guidance and Control

Guidance and Control

Sportradar has been a public company for five completed fiscal years, and in that time it has issued a full-year revenue range every March, revised it most Novembers, and reported against it every March after that. That sequence is a public ledger. It can be walked line by line, and the outcome of each entry is a matter of record rather than of interpretation.

The ledger is mostly good. It also contains one entry where the standard itself moved, and one where the disclosure that proved a promise was retired the year after the promise was met. Both changes were properly disclosed and neither is unusual on its own. What follows sets out the whole record — the floors raised and beaten, the single guide broken, the bar lowered, the segment line retired, the cash commitment kept — and then establishes who issues the word in the first place, on what mandate, and for how much longer that mandate runs.

Sources: guidance record compiled from the nineteen earnings-call transcripts, Q3 2021 to Q1 2026 [1]; voting power from the FY2025 Form 20-F beneficial-ownership table as of March 12, 2026 [2]; economic ownership derived from the same table on a converted basis.

Five years of guidance

The pattern is consistent enough to state plainly. In four of the five years the company set a number, raised it during the year, and finished above it. In one year it set a number, reaffirmed it in August, and cut it in November.

Sources: initial and revised guidance from the earnings-call transcripts — FY2021 [3], FY2022 [4] and its November revision [5], FY2023 [6] and its November cut [7], FY2024 [8] and its raise [9], FY2025 [10] and its raise [11]; delivered revenue and Adjusted EBITDA from the FY2025 Form 20-F [12] and the FY2023 Form 20-F [13].

Two features of the table are worth separating. The first is the construction of the guide. FY2022 and FY2023 were issued as two-sided ranges; those are the two years that produced the only narrowed Adjusted EBITDA band and the only revenue cut. From FY2024 the company switched to "at least" floors — at least 20% growth in revenue and Adjusted EBITDA, equating to €1.050 billion and €200 million [14] — and every floor since has been raised in-year and cleared. A floor is a weaker commitment than a range, and it has a correspondingly better hit rate.

The second is the shape of the one miss. On August 9, 2023, asked whether a €10 million currency headwind pushed the company toward the low end of its range, the CFO answered that "we're reaffirming that we believe we will land in the guidance ranges that we set at the beginning of the year" [15]. Eleven weeks later, on November 1, the range came down to €870–880 million, attributed to a stronger euro against the dollar and to third-quarter softness in Managed Trading Services caused by a run of bettor-favourable soccer results [16]. The cut carried little warning, and the second of its two causes — client gaming revenue moving against the company through revenue-share contracts — is the mechanism Rights Bought on Credit sets out.

Attached to that same call was a cost commitment: a global workforce reduction that "should result in an approximate 10% reduction of the company's 2023 labor costs run rate" [17]. Adjusted personnel expense was €257.5 million in FY2023, €282.8 million in FY2024 and €310.8 million in FY2025 [18]. Against revenue those are 29.3%, 25.6% and 24.1%. The promise was delivered as a ratio — five points of operating leverage over two years — and not as a reduction in the absolute bill, which is the reading the words themselves supported.

The long-term bar

Alongside the annual guide the company has carried a multi-year standard, and that standard was reset once.

On March 20, 2024, reporting FY2023 and guiding FY2024, the CFO put it this way: "we are well on track to deliver on the long-term financial targets we outlined at the time of our IPO, namely, revenue growth of at least 20% and adjusted EBITDA margins in the 25% to 30% range" [19]. Twelve months and twelve days later, the Investor Day deck of April 1, 2025 published a three-year outlook: at least 15% revenue CAGR from €1.1 billion in 2024 to about €1.7 billion in 2027, Adjusted EBITDA margin of 27% in 2027 against 20% in 2024, and free cash flow conversion of 60% [20].

Source: revenue growth as computed in this report's fact table from revenue as filed, FY2021–FY2025 [21]; the 20% bar as reasserted on the FY2023 results call [22] and the 15% bar from the Investor Day 2025 deck [23].

FY2025 delivered 16.6% revenue growth. That clears the newer bar by 1.6 points and misses the older one by 3.4. The reset was published twelve months before the first year that would have failed the standard it replaced.

The margin half of the change runs the other way, and it belongs in the same paragraph. Adjusted EBITDA margin has moved 18.2%, 17.2%, 19.0%, 20.1% and 23.0% across the five years [24][25]. The IPO-era framing of a 25–30% margin was never within reach on the trajectory that existed when it was reasserted; the 27% target for 2027 sits inside the old band and asks for four points of expansion in two years off a base that has just produced 2.9 points in one.

Sources: realized margins from the FY2023 Form 20-F for FY2021–FY2023 [26] and the FY2025 Form 20-F for FY2024–FY2025 [27]; the 2027 target from the Investor Day 2025 deck [28].

Two readings of the reset are available and the record does not settle between them. One is that the company recalibrated to a post-IPO reality it could actually hit, and paired the lower growth number with a higher committed margin — a trade of top line for profitability that the realized margin series supports. The other is that the standard moved to meet the performance rather than the reverse. What would decide it is the 2027 outcome itself: 27% delivered on roughly €1.7 billion validates the first reading, and a second reset validates the second.

Disclosure retired after the proof

The clearest kept promise in the archive is also the one that can no longer be checked the same way.

On August 17, 2022 — with the United States segment losing money and the shares well below the offer price — the CEO told investors: "we now expect to achieve profitability in the U.S. at least 12 months ahead of the original 2025 target date" [29]. The company delivered. United States segment Adjusted EBITDA went from minus €22.6 million in 2021 to minus €4.1 million in 2022 to positive €18.9 million in 2023, on segment revenue that more than doubled from €71.7 million to €165.5 million [30]. A 2025 target was met in 2023.

Source: FY2023 Form 20-F segment tables; no United States segment figures exist for 2024 or 2025 because segment reporting was collapsed to a single reportable segment effective January 1, 2024 with prior periods restated [31][32].

The blank years are the point. In October 2023 the company reorganized; the restructuring completed in January 2024; and in reassessing its segment identification under IFRS 8 it "concluded that discrete financial information was available to allocate resources solely on a consolidated basis," making it one operating and reportable segment effective January 1, 2024, with historic periods restated [33][34].

The accounting rationale is standard and the trigger — a chief operating decision maker who allocates on a consolidated basis — is the correct test. The consequence is nonetheless specific: the line item on which the US commitment was verified stopped being published in the first year after the commitment was met, and there is now no continuous segment series against which to test any geographic mix claim. Management still describes US performance verbally — the US grew 23% in FY2025 and is 25% of total revenue, per the FY2025 results call [35] — but a spoken growth rate and an audited segment profit are different classes of evidence. This is the second of two disclosure changes in three years: four geographic segments through FY2022, two product segments in FY2023, one segment from FY2024.

The cash promise

Against those two, one multi-year commitment has been made and kept without amendment.

On August 17, 2022 — the same call as the US-profitability statement, and at a point when free cash flow for the year would come in at €5.5 million — the CFO said: "we believe that our business model can achieve the 55% to 60% free cash flow conversion targets over the long term" [36]. In March 2023, walking analysts through a year in which the cash balance had fallen from €743 million to €244 million, the interim CFO added: "we will remain cash flow positive" [37].

Source: free cash flow as computed in this report's fact table on an operating-cash-flow-less-capex basis; the underlying reconciliation is in the FY2025 Form 20-F, which also reports the company's own free cash flow of €167.2 million after lease principal [38].

Free cash flow has risen in every year since the promise, from €5.5 million to €174.7 million on the operating-cash-flow-less-capex basis charted above, and margin from 0.8% to 13.5%; on the company's own definition, which deducts lease principal as well, FY2025 free cash flow is €167.2 million. The two are the same year on two bases, not two readings of one. On the company's own definition and denominator, FY2025 conversion was 56% against 53% in 2024 and 30% in 2023 [39], and the first quarter of 2026 converted at 67% on €44 million of free cash flow [40]. The 55–60% band stated at the low point has been reached.

The counter-fact belongs in the same breath, and it is established rather than asserted: Rights Bought on Credit shows what carries that conversion — rights amortization that is non-cash in the year it is charged, rights principal that settles inside investing rather than operating, and a growing share of rights acquired without cash at all. The commitment was made about a ratio, and the ratio was delivered. Whether the ratio measures the same thing it did in 2022 is a separate question, answered there.

What the IPO cash did

The prospectus of September 14, 2021 priced 19,000,000 Class A shares at $27.00 and disclosed net proceeds to the company from the offering and concurrent private placements of approximately $634.6 million, intended "for working capital, to fund incremental growth and future acquisition of, or investment in, companies, technologies, products or assets that complement our business and other general corporate purposes," with the explicit caveat that the board would have broad discretion and no definitive plans existed [41].

What the cash actually did was retire debt. On July 14 and December 14, 2022 the company prepaid €200.0 million and €220.0 million of Facility B, "thereby reducing the outstanding Facility B commitments to zero," writing off €6.8 million of unamortized issuance costs in the process [42]. That is a defensible use of proceeds in a rising-rate year and it is the reason the borrowings line has been empty ever since — a fact whose significance Rights Bought on Credit develops.

The one discretionary venture investment made from that cash went the other way. On August 4, 2022 Sportradar bought 100% of Bettech Gaming (PYTY) Ltd from Carsten Koerl and minority shareholders for €7.0 million, and immediately contributed those shares plus €13.7 million of cash and a €14.3 million equalization payment to Ringier for a 49% interest in a new Swiss holding company, SportTech AG. The same disclosure notes that "Sportradar's director Marc Walder also serves as a director for Ringier" [43]. On May 31, 2023 — ten months later — the 49% was sold back to Ringier, producing a €13.6 million loss on disposal [44], on top of €4.0 million and €3.7 million of share-of-loss in 2022 and 2023.

The sums are small against a company now turning over €1.3 billion. The structure is what makes the episode worth recording: an asset sourced from the chief executive, contributed to a joint venture with a company whose chief executive sat on Sportradar's board, and unwound at a loss inside a year. It is the only transaction of its kind in the record, and the related-party disclosure around it was complete and contemporaneous.

Votes, economics and a dated sunset

Every commitment above was made by an executive team that serves at the pleasure of one shareholder.

Carsten Koerl holds all 783,607,701 outstanding Class B ordinary shares plus 1,840,883 Class A, giving him 78.7% of combined voting power as of March 12, 2026 [45]. The mechanism is nominal value rather than a separate voting right: Class B shares carry one vote each at one-tenth the nominal value of Class A, so the same capital buys ten times the votes, and each ten Class B convert into one Class A [46].

The gap that produces is visible in the same table. Canada Pension Plan Investment Board owns 31.8% of the Class A and casts 6.8% of the votes; Technology Crossover Management owns 13.6% and casts 2.9% [47]. Reinforcing it, the Articles strip voting rights from any newly acquired Class A stake above 10% of registered share capital, while grandfathering holders who were already above that line before the Articles were registered [48]. A rival bloc cannot assemble votes by buying stock.

Source: FY2025 Form 20-F, dual-class risk factor [49].

The date trigger requires no event and no decision. On the current share count, when it passes, the founder's voting power falls from 78.7% to roughly his economic stake — a little under 27%. Two years and two months separate this report from that conversion, and the fourth trigger runs in parallel: enough further selling of Class B, and the sunset arrives early.

The direction of the founder's stake

Across the two years leading into that clock, the direction of the founder's holdings has been one way.

Sources: beneficial-ownership tables as of March 1, 2024 [50], March 1, 2025 [51] and March 12, 2026 [52]; economic ownership derived from those tables on a fully converted Class A-equivalent basis.

Between the first and second rows the Class A holding fell by 3.50 million shares. Between the second and third it rose by 1.81 million — but only after a conversion in between: on April 16, 2025 Koerl converted 120,000,000 Class B into 12,000,000 Class A, with the Class B remaining in treasury pending cancellation [53]. Net of that conversion, roughly 13.7 million Class A shares were disposed of across the two years, and economic ownership on a converted basis fell from about 31.5% to about 26.8% while voting power fell 2.9 points.

Eight days after the conversion, on April 24, 2025, a secondary offering of 23 million Class A shares priced at $22.50. The selling shareholders included a CPPIB affiliate, TCV and Koerl; Sportradar concurrently repurchased 3 million shares, up to $75 million, under its own buyback authorization [54].

None of this was visible in real time. As a foreign private issuer, Sportradar's directors and officers were not required to file insider reports under Section 16(a) of the Exchange Act until March 18, 2026; principal shareholders remain exempt from Section 16(a) altogether, and officers, directors and principal shareholders all remain exempt from the Section 16(b) short-swing profit rules [55]. The sales are reconstructible from annual beneficial-ownership tables filed a year in arrears, and from a secondary offering prospectus. They were not reportable as they happened.

Then, on the April 28, 2026 first-quarter call — with the shares far below both the $22.50 secondary print and the $27.00 offer price — the founder said: "I believe the company's current valuation does not reflect the strength of our business and our long-term prospects, and I'm confident in the path we are on. Accordingly, I intend to personally purchase $10 million worth of shares in Sportradar when our trading window opens" [56].

Stated intent and realized cash flow point in opposite directions across the same two-year window, and $10 million is about 3% of the roughly $308 million those 13.7 million shares would have been worth at the $22.50 secondary price. Whether the purchase was executed is not answerable from this corpus: no post-May-2026 insider filing, 6-K or news item covering the period after the trading window opened appears in the record. It is, however, now a reportable event — the Section 16(a) obligation began on March 18, 2026, six weeks before the pledge — so it is dated and checkable in a way that the preceding two years of selling was not.

Who sits beneath, and what they are paid for

The layer under the founder is new. As of December 31, 2023, Executive Management comprised Ger Griffin as chief financial officer, Eduard Blonk as chief commercial officer, Ulrich Harmuth as chief strategy officer and Lynn McCreary as chief legal officer [57]. As of March 1, 2026 the executive officers are Koerl, Craig Felenstein as CFO since June 2024, and Michael Miller as chief administrative and legal officer [58]. A chief operating officer, Sameer Deen, was announced alongside the first-quarter results and started on May 18, 2026 [59]. The finance seat has turned three times since 2020: Alexander Gersh from July 2020 [60], whose departure Koerl announced on the same August 2022 call as the US-profitability commitment [61], then Griffin, then Felenstein. Institutional memory below the founder is under two years old.

What that layer is paid for is set out precisely. Long-term incentives run entirely on total shareholder return relative to the constituents of the Standard and Poor's 500 Information Technology index over two-, three- and four-year periods, on a curve that pays nothing below the 40th percentile, 50% at the 40th, 100% at the 60th, 150% at the 80th and 200% at the 95th or higher. Performance stock units were 70% of long-term grant-date value for executives other than the CEO in March 2025 and 100% for the CEO, who did not participate in the plan's first year [62].

That index contains no sports-data company, no betting supplier and no sports-rights holder. Payout therefore turns on how Sportradar's share price behaves against US mega-cap software and semiconductors, not on how it performs against Genius Sports or Kambi — the rivals The Barrier and the Bill benchmarks the operating record against. The 2025 tranches were struck at a weighted-average grant-date fair value of $26.35, against $12.59 for the 2024 grants and $11.40 for 2023 [63]. The shares closed at $14.68 on July 27, 2026, the last session in this report's price record, so the highest-struck tranche is also the one furthest from its threshold.

The annual bonus runs on company-wide Adjusted EBITDA, revenue and cash flow [64]. Adjusted EBITDA as the company defines it excludes share-based compensation of €56.1 million, non-routine litigation costs of €35.2 million, transaction-related costs of €11.6 million and the €2.2 million cost of the secondary offering in which insiders sold — all in FY2025 alone [65]. The composition of that gap and its recurrence are analysed in Rights Bought on Credit; the point here is narrower. Several of the excluded categories are costs management itself elects to incur, and the metric that determines the cash bonus is measured after they are removed.

Total FY2025 compensation for current directors and executive officers was CHF 13.6 million [66], of which the CEO's package was CHF 7.80 million — CHF 0.62 million of salary, CHF 0.61 million of bonus and CHF 6.31 million of stock awards, with no separate board fee [67]. Against a stake of roughly 80 million Class A-equivalent shares, the package is two orders of magnitude smaller than the holding; the incentive that matters is the direction of the stake, not the grant.

Five years of ineffective control

Every number in this report passes through a financial reporting system in which the company has reported a material weakness in every one of the five years it has been public.

Sources: FY2021 Form 20-F, material weakness identified as of December 31, 2021 [68] with no management report due to the SEC transition period [69]; FY2022 not remediated [70]; FY2023 conclusion [71] and adverse opinion [72]; FY2024 conclusion [73] and adverse opinion [74]; FY2025 conclusion [75] and adverse opinion [76].

The weakness was first identified in the run-up to the IPO and related to "insufficient design and implementation of controls, IT systems and segregation of duties" [77]. Five years later the FY2025 filing reports that despite "significant progress in our remediation efforts during the year," control over financial reporting "was not effective as of December 31, 2025," because the weakness "as reported in the prior year" has "not been fully remediated" [78]. It now reflects "insufficient design and implementation of control activities in certain financial reporting processes" plus "an insufficient complement of personnel with appropriate levels of knowledge, experience, and training" [79]. The processes affected are not named, so the exposure cannot be mapped to particular balances.

Three things about the table deserve to be read together rather than separately. KPMG AG has been the auditor since 2014 and has issued an unqualified opinion on the financial statements in every year, including the three in which it issued an adverse opinion on the control environment producing them [80][81]. Those two opinions are consistent — the statements can be fairly stated while the system that produced them is judged unreliable, because the company performed "additional analysis and procedures" to compensate [82]. Compensating procedures are, by construction, manual and after the fact.

And the same auditor flags two critical audit matters, both attaching to the largest items on the balance sheet. The first, present in both FY2024 and FY2025, is the capitalization assessment for newly acquired or modified sport rights licences and the identification of the fixed minimum and variable payments used to measure them, requiring "significant and complex auditor judgment." The second, new in FY2025, is the valuation of the €302.4 million of licence intangibles acquired with IMG ARENA, where "a high degree of subjective auditor judgement was required to evaluate the projected margins and discount rate" [83]. The asset the auditor names as hardest to audit is the same asset the company's entire capital deployment runs through, and it is the one whose measurement — described in Rights Bought on Credit — sets both the largest asset and the largest liability.

There is a plain reading in the company's favour: the remediation disclosure says a "significant number of deficiencies across several affected financial reporting process areas" were successfully addressed during 2025, and the direction of travel over five years has been toward fewer deficiencies, not more. There is no restatement in the record: the FY2025 cover page reports no correction of an error to previously issued financial statements and no error correction requiring a recovery analysis of incentive-based compensation [84]. The finding is duration, not error: nothing has yet gone visibly wrong, and the control environment has been formally inadequate for the entire life of the listing.

The record that is clean

Founder-controlled structures usually come with a related-party surface. This one does not.

The only recurring related-party item is €0.1 million of revenue in each of 2023, 2024 and 2025 with UAB TV Zaidimai, a Lithuanian company in which Koerl holds 33% [85]. Every related-party transaction must be reviewed and approved or ratified by the audit committee under a written policy, which requires an assessment of whether terms are comparable to arm's length [86]. No dividend has ever been declared since incorporation [87], which is consistent with a reinvestment-first policy rather than with extraction.

Nine of the ten directors are determined independent under Nasdaq rules, each elected individually and annually by the general meeting, with a non-executive chairman and Koerl as the only executive on the board [88]. That is a stronger formal structure than a controlled company is obliged to maintain.

Two qualifications sit alongside it, both from the filings themselves. First, all of those annual elections are decided by a holder with 78.7% of the votes, and the CEO's own compensation is reserved for the full board rather than for the compensation committee [89]. Second, two of the nine independent directors are employed by shareholders: John Doran is at TCV, which holds 13.6% of the Class A, and Pascal Keutgens at CPP Investments, which holds 31.8%. Both sit on the compensation committee that sets pay for executives other than the CEO [90], both take no compensation for board service [91], and the Shareholders' Agreement among Koerl, CPP and TCV provides director-nomination rights until a party falls below 7.5% of share capital [92]. Board composition is partly contractual rather than purely elective.

Authorized and executed

The last commitment on the ledger is capital return, where the authorization and the execution are two different numbers.

Sources: authorization steps and cumulative repurchases of $111.2 million through December 31, 2025, of which $90.9 million was executed during 2025, from FY2025 Form 20-F Note 19.3 — the December 2024 figure of roughly $20 million is the difference [93]; the October 2025 cumulative figure from the Q3 2025 call [94]; the February 2026 increase to $1 billion, the $171 million repurchased as of February 27, 2026 and cumulative repurchases of $228 million as of April 24, 2026 from the Q4 2025 and Q1 2026 results releases [95][96].

The authorization went from $200 million in March 2024 to $300 million in October 2025 to $1.0 billion in February 2026 — a fivefold increase in under two years [97][98]. Execution over the same period was 12.5 million shares for $228 million, an average of roughly $18.24, against the $27.00 offer price [99]. Management has been explicit throughout about why the two diverge: "our capital allocation priority is investing in expanding the long-term growth potential of the company, and we will weigh returning capital to shareholders versus additional organic and M&A investment opportunities," a formulation repeated near-verbatim on the Q3 2024 and Q3 2025 calls [100][101]. A large authorization with modest execution is what that stated hierarchy predicts.

The effect on the share base is the part worth stating precisely, because it runs against the usual intuition about buybacks.

Source: weighted-average diluted Class A and Class B share counts from FY2025 Form 20-F Note 12; the Class A-equivalent column adds Class B at its one-tenth dividend entitlement and is derived from the same table [102].

Three consecutive years of repurchases — €9.0 million, €28.7 million and €105.2 million of treasury purchases [103] — have coincided with a diluted Class A count rising from 226.6 million to 237.5 million, and a combined Class A-equivalent base rising from about 317.0 million to about 319.4 million [104]. Issuance has outrun retirement. In FY2025 alone, 3.1 million treasury shares were surrendered back into vesting equity awards [105], and €34.2 million of shares were granted to a sport rights licensor, following €52.0 million granted to a licensor in 2023 [106].

That licensor equity connects to a statement made on the first call the company ever held as a public company. On November 17, 2021, disclosing the NBA's stake alongside the eight-year data agreement, the CEO said: "As a very general statement, I think you will not see too many equity deals from Sportradar, like we did it now with the NBA," and later on the same call, "from the scope of Sportradar, you will see not many of those deals following now the scope of the NBA" [107][108]. Two further grants totalling €86.2 million have followed, in 2023 and 2025 [109]. Two is not many, and the statement was hedged; the direction is nonetheless the opposite of the one it pointed in, and it is the mechanism by which the rightsholders described in The Barrier and the Bill became shareholders as well as creditors.

April 2026 is the first genuine change of pace. On April 28 the company announced a $250 million enhanced open-market repurchase to be completed in roughly three months, having repurchased about $90 million in the first quarter [110][111]. One quarter's programme now exceeds the $228 million executed across the preceding twenty-five months. The share-count effect of it, and the market context in which it is running, belong to Priced for an Allegation.

Where the record stands

The five-year ledger reads as follows. Four annual revenue guides raised and beaten; one cut, eleven weeks after being reaffirmed. A US-profitability commitment met a year early and then made unverifiable in the same form when segment reporting collapsed to a single line. A cash-conversion band promised at the low point and reached. A long-term growth bar lowered by five points twelve months before the first year that would have failed the old one, paired with a committed 2027 margin of 27% that sits inside the old 25–30% band and seven points above the 2024 level. IPO proceeds that retired debt rather than funding the growth and acquisitions the prospectus named, and one venture bet sourced from the CEO and unwound at a €13.6 million loss inside ten months. A related-party record that is close to empty, a board that is formally independent and elected annually by one holder, and a control environment that has been judged ineffective in every year of the listing.

Four things on that ledger are open, dated and checkable.

Sources: the FY2026 outlook and its reaffirmation from the Q1 2026 deck [112] [113]; the Q2 2026 reporting date from the news archive [114]; the repurchase programme and the personal-purchase statement from the Q1 2026 call [115]; the Section 16(a) effective date from the FY2025 Form 20-F [116]; the conversion date from the dual-class risk factor [117].

The read this chapter lands on is that the record is mixed and legible rather than uniformly good or bad: the annual word has been kept far more often than not, while the two commitments that were most testable — the long-term growth bar and the US segment — are the two that can no longer be tested the way they were made, one having been reset and the other retired from disclosure. The strongest fact against that read is the cash-conversion promise, which was made at the worst possible moment, involved no restatement of the standard, and was delivered. What would move the read either way is the 2027 margin outcome against the 27% bar, and whether the FY2026 guide survives contact with the second half without amendment.

None of that is what has actually moved the share price. The FY2026 guide was reaffirmed six days after the shares fell from $16.84 to $13.04 in a single session [118], and that fall attached to neither a reported result nor a missed guide.