Summary

  • Yahoo completed its broadcast.com acquisition on 20 July 1999 by exchanging approximately 28.647 million Yahoo shares for the outstanding shares, according to its contemporaneous SEC filing. A later annual table shows 57.294 million shares because its share counts were adjusted for a stock split, not because Yahoo bought the company twice.
  • Yahoo’s FY1999 and FY2000 filings describe approximately 3,600 and 4,150 Internet events in those years. Business-services revenue rose from $56.4 million to $105.7 million, but the category mixed broadcasts with hosting and enterprise services.
  • Yahoo said it did not allocate cost of revenues or operating costs to its business-services segment. It also warned that broadcast.com’s originally unicast streaming service had to make multicasting work at scale. Event volume was visible; the standalone margin of the acquired operation was not.

The deal was shares, and the service had two doors

The standard tale of broadcast.com starts with a huge acquisition valuation and ends with a joke about the dot-com boom. Yahoo’s own filing gives a more useful inventory. It closed the acquisition on 20 July 1999 and exchanged about 28.647 million shares of its common stock for all outstanding broadcast.com shares. It also converted options held on the acquired company into Yahoo options. This was a stock transaction accounted for as a pooling of interests. The roughly $22.1 million one-time charge anticipated in that interim filing concerned acquisition expenses, including advisory and merger costs. It was not the purchase price and certainly not a profit figure for streaming.

A later FY1999 annual acquisition table lists 57.294 million Yahoo shares for broadcast.com. The annual figures were adjusted for a subsequent two-for-one split. The two share counts represent the same exchange in different share units. Treating them as separate issuances would invent a second deal; treating the one-time merger expense as the price paid would turn an accounting line into a headline it never was.

The acquired business also had two routes to users. Yahoo’s 1999 filing said broadcast.com strengthened audio and video programming across Yahoo properties and brought live and on-demand online events for corporate customers. The consumer-facing broadcast site carried sports, concerts, radio, books and company conference calls. Yahoo could distribute licensed material through its larger network. The other door was paid production and delivery for businesses that wanted a meeting, an investor event or another broadcast online. The acquisition joined an audience business to a service business; the portal made them look like one screen.

A scoreboard with events and a blended revenue line

Yahoo measured activity. Its FY1999 Form 10-K reported about 3,600 events broadcast that year, up from about 1,800 in 1998 and 742 in 1997. Its FY2000 Form 10-K reported about 4,150 events in 2000. That is 550 more events in a year. Each count describes broadcasts across the combined Yahoo business, not a separate broadcast.com subsidiary still standing outside Yahoo’s accounts.

The revenue line needs the same care. Business-services revenue was $56.4 million in 1999 and $105.7 million in 2000. In 1999 Yahoo defined that category as live and on-demand audio and video event broadcasting plus subscription-based hosting. In 2000 it also listed enterprise services under the Corporate Yahoo! product. Yahoo said its Store product was the most significant portion of hosting revenue, while Corporate Yahoo! revenue was not significant in 2000. No filing gives a clean slice of those $105.7 million for broadcast.com alone. Assigning all of it to the acquisition would be a tidy chart with the wrong owner.

The event count and the revenue category therefore answer different questions. One counts pieces of work delivered. The other counts money from a basket of services. Neither states what it cost to produce an event, acquire content, run delivery infrastructure or sell a business contract. Yahoo’s annual report listed live-event production costs, Internet connection charges, content fees, equipment depreciation and staff-related expenses among its company-wide cost of revenues.

It explicitly said it did not allocate cost of revenues or operating costs to its business-services segment because management did not use that information to measure segment performance. The public scoreboard had a volume column and a revenue column; the corresponding segment cost column was not there.

The stream had a cost mechanism

The engineering record inside the FY2000 filing makes the missing column more consequential. Yahoo said the streaming services acquired with broadcast.com had originally used unicasting: one company-originated stream per user. It had begun to deploy multicasting, which could serve multiple users from a company-originated stream, and said it still needed to test, deploy and scale the multicast network. Yahoo warned that demand for audio and video could outrun the economics of delivering it.

That is an operating constraint, not a calculation of broadcast.com’s loss. One stream per listener consumes delivery capacity differently from a shared stream, and rights, production and networking costs sit around either method. Yahoo’s filing describes the transition and its risk, but does not attach a broadcast.com-only cost schedule. A service-line profit-and-loss account for the acquired operation would change the margin judgment; the published annual reports offer no such account.

The missing information did not prevent Yahoo from selling the product. The company could count events, recognize customer-service revenue and advertise a larger media offering. The reporting choice did prevent an outside reader from extracting a standalone streaming margin from the blended business-services figure. The distinction matters because a rising event count can coexist with rising production and delivery costs. A growing service can be useful to the portal, valuable to customers and hard to price correctly at the same time.

A merger made the old company harder to measure

Pooling-of-interests accounting restated combined results as if acquired businesses had been together since inception. Yahoo’s annual filing used that treatment for broadcast.com along with other acquisitions. The acquisition therefore entered the public record in two ways at once: as a dated stock exchange, and as part of a combined operating history. Neither form is wrong. They answer different questions. The dated exchange says who took control and what securities changed hands; the combined series says how Yahoo presented the group after integration.

The business-services category covered hosting products that Yahoo already had, as well as broadcast delivery and later enterprise services. That made the acquisition’s separate economics harder for a reader to isolate. The financial statements reported the combined company’s results, and management stated that it did not measure segment performance by allocating those costs. This was a deliberate measurement boundary, not an arithmetic puzzle an analyst can solve by dividing $105.7 million by 4,150 events. Such a division would blend Store hosting revenue, enterprise activity and broadcasts, then pretend every event sold for the same amount.

Yahoo also reported about $20 million in actual broadcast.com-related acquisition costs in the FY1999 annual filing after the earlier interim estimate of $22.1 million. These are merger expenses. An adviser’s fee and a live-event production cost belong to different operating stories. Putting both in a single “cost of Broadcast.com” total would erase the difference between buying a company and running its servers.

The most revealing sentence is the plainest one in the filing: management did not use an allocated business-services cost figure to measure that segment. Yahoo could tell investors how many events ran and how much a mixed service category brought in. It could not, from those disclosed lines, show whether the acquired streaming operation earned a standalone operating profit. The famous stock deal changed ownership in one day. The work of producing and delivering another 550 events occupied the next year, in a cost column the public never received.