Infobird is no longer a software compounder. The Nasdaq ticker is a Cayman listing wrapper that sold the original customer-engagement stack for a token Hong Kong price, then spent the remaining cash pile to buy a mainland digital-advertising variable-interest vehicle and, within a year, wrote off most of the premium paid for that vehicle. The April annual filing is the first clean look at a full year of the new agency, and the look is harsh: revenue recovered into the high single millions while a goodwill charge near $55 million flattened book value and advertised that the Pure Tech purchase price did not survive contact with a year of actual cash conversion. The equity is a call on whether two advertising buyers keep feeding a thin agency layer inside an untested contractual structure, not a claim on a durable software franchise.
The load-bearing tension is the gap between a recovered top line and an asset base that just confessed the purchase was too expensive. Continuing operations printed about $9 million of revenue against roughly $3 million of gross profit, which is a real agency, not a zero. The same year took a $55 million goodwill impairment and a net loss near $57 million, so the accounting year is a write-down story rather than an operating-turn story. Cash at year-end still sat near $5 million against modest borrowings, which is why the tape can look cheap on book and sales even as the franchise quality collapsed. Cheap on a cleaned-up stub is not the same thing as cheap on a franchise.
What actually moved under the hood is customer concentration plus a control failure, not mix or price. Two buyers accounted for about four fifths and about one tenth of fiscal 2025 revenue, and those same names dominated receivables, so the recovered print is a two-ticket book rather than a diversified agency. Management also concluded that disclosure controls were ineffective and identified a cluster of material weaknesses spanning information-technology policy, vendor oversight, data security, and segregation of duties. The June amendment then came back to correct identified accounting errors in the same annual statements. A listing that restates after writing down most of a fresh acquisition is telling the market that the books, not just the model, are still being rebuilt.
The August succession is the latest governance event, and it does not close the operating question. Xiangyang Wen left the chief-executive and chair roles in late August, and Yaqing Wei stepped in from clinic management and consumer-marketing posts rather than from a disclosed digital-advertising operating seat. The next dated test is the still-absent first-half interim, which foreign-private-issuer rules put on the calendar before month-end September. Does a two-buyer agency inside a contractual China structure, run by a new chair whose resume is not ad-tech, convert the recovered revenue into cash without another control incident, or does the stub keep decaying toward residual cash?