B90 Holdings plc's significant shareholdings are disclosed through a dated issuer table and separate notifications of holdings and directors' dealings. At 31 July 2026, the largest named position was Winforton Investments Ltd: 85,520,000 ordinary shares, reported as 19.5%. That identifies the largest line in that disclosure, without establishing a single ultimate owner.
Our analysis separates named holdings, attributed interests and voting rights. The B90 Holdings company overview explains the performance marketing business; this page examines the ownership evidence and its limits.
Significant shareholders: the dated issuer disclosure
The issuer's AIM Rule 26 page lists the following significant shareholders holding above 3% at 31 July 2026. It states that 440,814,739 ordinary shares were in issue on that date. The page's last-updated date is 1 August 2026; our research date is separate from both.
| Holder as disclosed | Ordinary shares | Percentage as disclosed |
|---|---|---|
| Winforton Investments Ltd | 85,520,000 | 19.5% |
| P. Westerterp | 37,842,240 | 8.6% |
| Performance Media SIA | 28,784,449 | 6.6% |
| Diman BV | 25,178,432 | 5.7% |
| H.M. Hansen | 23,907,004 | 5.4% |
| M. Rosman | 23,419,019 | 5.3% |
| Dowgate Group Ltd | 22,025,000 | 5.0% |
| Ronny Breivik | 20,197,047 | 4.6% |
Source: B90 Holdings plc, AIM Rule 26, significant shareholders at 31 July 2026; accessed 17 November 2026.
The issuer adds that Ronny Breivik owns 34.65% of Performance Media SIA and reports a 7.0% interest for him when its attributed shares are included. That broader interest is explained below; it is not an additional independent row of shares.
We retain the published numbers, although the percentages do not all reconcile with the stated share total. For example, dividing 85,520,000 by 440,814,739 gives approximately 19.40%, against the reported 19.5%. This calculation identifies a reconciliation issue; it does not establish a corrected register or explain the difference.
The table is a significant-holder disclosure, rather than a complete register of every shareholder. Absence from it does not establish that a person has no interest. Nor does reading the page in November make all eight positions a November snapshot. Later individual notices are presented separately.
The denominator concerns ordinary shares already in issue. Options disclosed alongside directors' shareholdings are a separate category; adding unexercised awards would change the basis of the comparison and would not reproduce this issuer table.
The instrument and its market data belong in share price and listing. They provide a different answer from the identity and scope of a disclosed holding.
How direct and indirect interests overlap
Mark Blandford's TR-1, published on 1 March 2024, records a threshold event on 28 February and notification to the issuer on 29 February. It attributes 85,520,000 voting rights, reported as 19.5%, to him. Section 9 lists Winforton Investments Ltd at 18.7% and Mark Blandford at 0.8%. However, section 8A places the full 85,520,000 in the direct voting-rights column, with a dash for indirect rights. We retain that inconsistency in the reproduced form rather than presenting it as a fully reconciled direct-and-indirect split.
The later issuer table places the same 85,520,000 shares and 19.5% against Winforton alone. These are different dated presentations of attribution. Adding them as two separate 19.5% holdings would count the same disclosed total twice. The documents do not explain the change of presentation sufficiently to infer a subsequent transfer between Blandford and Winforton.
The connection is evidenced, but matching totals do not establish a reconciled personal economic percentage or explain the form's inconsistent presentation.
Breivik presents a second overlap. The July table gives him a separate 20,197,047-share line at 4.6%, followed by the broader 7.0% including Performance Media shares. The 34.65% figure describes his interest in Performance Media SIA, not a 34.65% holding in B90. Combining the 4.6% line with the 7.0% aggregate would count part of the interest twice.
Performance Media also requires care within the annual report itself. Its directors' interests footnote, covering 31 December 2025, names Performance Media Ltd with 31,084,450 B90 shares. Elsewhere in the same report, the significant-shareholder table, stated as at the report date, names Performance Media SIA with 28,784,449. The July 2026 webpage repeats the latter name and count. The footnote also identifies Entercreation Ltd as wholly owned by Breivik, holding 8,600,000 B90 shares.
These passages have different dates and scopes; we cannot establish whether they describe one legal entity or explain the share-count difference. We preserve the labels and reported aggregates without adding every component to a later total again.
Voting rights, custodians and ownership
Dowgate Group Limited's TR-1 reports 22,025,000 voting rights, or 5.01%, following a threshold event on 23 August 2024. The issuer was notified on 27 August; its RNS archive dates publication to that day. The July 2026 issuer table reports the same share count at 5.0%; both the date and reported precision differ.
| Label in the Dowgate notification | What the document identifies | Limit of the inference |
|---|---|---|
| Person subject to notification | Dowgate Group Limited | The notifying entity is not automatically the economic owner of every client holding. |
| Controlled undertaking in the chain | Dowgate Wealth Limited | The chain describes how the notified voting interest is attributed. |
| Named shareholder entries | Bank of New York Mellon on behalf of Dowgate Wealth Limited (Discretionary); Bank of New York Mellon on behalf of Onward Opportunities Plc; Royal Bank of Canada on behalf of SVS Dowgate Wealth UK Small Cap Fund | Custody and account labels do not create extra independent holdings to add to the notified total. |
The filing distinguishes these roles within one notification. It does not support treating the full notified position as Dowgate's own investment portfolio, or assigning each named bank a separate additional stake of the same size.
A further distinction concerns the issuer's 32.6% not held in public hands at 31 July 2026. Its definition includes directors, officers and major shareholders above 10%. That classification answers a different question from beneficial ownership. It is not interchangeable with an aggregator's insider or free-float category whose date and methodology may differ.
What later disclosures change
Two September 2026 announcements provide updates after the July table. Their transaction dates and publication dates are separate, and the resulting interests have different scopes.
| Person | Transaction date | Announcement date | Reported interest after the purchase |
|---|---|---|---|
| Ronny Breivik | 24 September 2026 | 25 September 2026 | 31,308,423 shares, approximately 7.1%, directly and indirectly in aggregate |
| Andrew McIver | 25 September 2026 | 25 September 2026 | 1,530,709 ordinary shares, 0.35% |
Breivik's announcement records a purchase of 258,448 shares. The 31,308,423 figure is the resulting direct and indirect interest, so adding the purchase to it again would overstate the disclosed total. It also cannot be added to July's separate Breivik row as a new, non-overlapping holding. The announcement does not provide a fresh breakdown reconciling every underlying company interest.
McIver's announcement records a purchase of 244,200 shares and the resulting position shown above. His reported 0.35% is below the July table's stated above-3% criterion. A director's dealing disclosure and a significant-holder table can therefore include different people because they serve different purposes.
These notices update two reported positions; they do not verify that all other shareholders remained unchanged or produce a complete September shareholder register. Our analysis retains the July baseline and labels the later evidence separately. The directors and governance page covers the responsibilities attached to the offices, rather than treating a job title as a shareholding.
What the disclosures show about control
In Note 25 of the 2025 annual accounts, the directors state that, at 31 December 2025, they did not believe there was a single controlling party. That is a dated directors' assessment, rather than an independent confirmation of the position on our research date.
Companies House identifies Mark Robert Blandford as a person with significant control of Winforton Investments Limited, company number 13689166, on the basis of significant influence or control. That entry does not state an exact percentage of Winforton's shares. The register separately identifies Sophie May Blandford as its director. The roles concern Winforton and do not by themselves establish sole control of B90.
We can identify the largest disclosed line, an attributed Blandford interest, Breivik's reported aggregate and a notified Dowgate voting chain. These documents do not resolve every ultimate economic interest or establish that the largest holder alone controls the plc.
Ownership of shares in B90 is also a different level from B90's ownership of companies. The subsidiaries and markets page covers that group structure; the shareholder table is not a list of subsidiaries.
Sources and method: The baseline is the issuer's Rule 26 disclosure. Separate TR-1 notifications, directors' dealing announcements, the 2025 annual accounts and Winforton's Companies House entries supply attribution and later context. The Blandford TR-1 was read as reproduced RNS text, separately from the host's automated summary. Published percentages remain labelled as reported; the illustrative ratio above is our calculation. None of the separate dates is presented as a complete ownership refresh. Source websites: b90holdings.com, ticker.app, company-information.service.gov.uk.
