VI
Mark học AIBlockbuster Inc. · Dallas, Texas · 2004–2010
Source: Blockbuster 10-K for fiscal 2004 (revenue); 8-K of 23/9/2010; bankruptcy petition, Exhibit A, via Bloomberg 23/9/2010 and a University of Tennessee College of Law paper (secondary source).
Source: Blockbuster 10-K 2005, 2006, 2007; Q1 2007 release (2/5/2007) and investor presentation of 28/6/2007 (8-K); 8-K of 2/7/2007; Netflix 10-K 2007, competition section; Blockbuster 8-K of 23/9/2010.
A company with $6.1 billion in revenue, whose online service was slowing its rival down, why did it file for bankruptcy less than four years later? Not because it failed to see Netflix coming. Because a long-term project was judged with a short-term yardstick, while the company carried debt and changed who was at the wheel.
Source: Blockbuster 10-K 2004 ($6.1 billion revenue); Netflix 10-K 2007 (growth slowed because of Blockbuster Online); Blockbuster 8-K of 23/9/2010.
Dates and figures come from Blockbuster's SEC filings; anything found only in the press or in memoirs is marked as a secondary source.
Source: Randolph (2019) and Hastings (2020) (secondary); Blockbuster 10-K 2005 (online 8/2004, 14%), Q2 2005 release in 8-K ($140 million); press releases of the California and New York Attorneys General (3/2005); Blockbuster proxy filings 2005 and 2007; 10-K 2006 and release of 1/11/2006; Q1 2007 release; Netflix release of 16/1/2007; 8-K of 2/7/2007; Netflix 10-K 2007; Blockbuster 10-K 2008; 8-K of 23/9/2010; DISH 8-K of 6/4/2011; DISH release of 6/11/2013 (reprinted by Engadget, secondary).
13 things often said about Blockbuster, checked against SEC filings, the companies' own releases and those of government agencies. Tap each line to see why.
10-K 2005: extended viewing fees were about 14% of 2004 rental revenue. Q2 2005 release: rental revenue lost almost $140 million in the quarter, largely offset by more base rentals. Full-year 2005 rental revenue fell only 5.0%.
Per the California and New York Attorneys General (3/2005): keep a disc more than 7 days and on day 8 it became a purchase; returning it within 30 days still cost a $1.25 fee. 47 states and DC settled; Blockbuster refunded customers and paid the states $630,000.
Netflix itself wrote in its 2007 10-K that subscriber growth in 2007 slowed mainly because of the rapid expansion of Blockbuster Online.
Netflix 10-K 2007: in the second half of 2007 Blockbuster moved to "profitable growth", cut marketing and raised Total Access prices. Blockbuster's 2008 annual report says it positioned Total Access as a "profitable and stable business".
8-K of 23/9/2010: a plan to cut debt from nearly $1 billion to about $100 million or less. The petition listed about $1.46 billion of debts against about $1.02 billion of assets (via Bloomberg and University of Tennessee College of Law). The auditor had warned in the 2008 report.
This is the account of Netflix's two founders in their books. Antioco says there were no serious talks. Some sources say the offer was for 49%, without enough evidence. There is no original document.
The image comes from Marc Randolph's memoir (2019), repeated by the press. There is no third witness or record.
Netflix competition was real (20.01 million subscribers at the end of 2010, per its 10-K). But there were also cheap rental kiosks, streaming, debt, and internal decisions. Keyes blames the studios' payment terms; that is his view.
The "radar screen" line is real (Motley Fool, 10/12/2008). But he named Wal-Mart and Apple as the worries, and in 2007 the company had bought Movielink for downloads. One quote is not a whole strategy.
Easy to say in hindsight. Total Access really lost money: Q1 2007 net loss of $46.4 million. The lesson is not "spend without limit", but which yardstick to judge a long-term project by and which milestone to stop at.
The filings show the opposite: online rental 8/2004, no late fees 1/1/2005, Total Access 1/11/2006, Movielink bought in 2007 for $7.7 million (10-K 2005, 2006, 2007; 8-K Exhibit 99.2).
It had online rental with over 3 million subscribers (Q1 2007) and Movielink for downloads, being integrated into Blockbuster.com (10-K 2007). The problem was scale and persistence, not absence.
DISH bought the assets in 4/2011 (about $320 million), and the company stores all closed in early 2014. One franchise store in Bend, Oregon still carries the Blockbuster name (AP, 2019, secondary).
Source: the documents named in each line; full list in Part 8.
Scores are Mark's assessment based on the sources in Part 8, not data from a study.
Blockbuster did not lose because it could not see the future. It saw it, ran for it, and was ahead for one leg. It lost because it changed the yardstick halfway, before the project could turn a profit.
The in-store rental market shrinking was an industry-wide shift, not Blockbuster's alone. No decision would have kept $6.1 billion of store revenue.
What was avoidable is how it fell: a company with a large online subscriber base, which its rival admitted was a threat, could have transformed instead of entering bankruptcy with debts larger than its assets.
Source for "More": Blockbuster 10-K 2004 ($6.1 billion); Netflix 10-K 2007; bankruptcy petition via Bloomberg (secondary); the rest is Mark's commentary.
Applies to every "pay first, earn later" project: a new product, a new sales channel, training a team, a digital shift, even your own learning plan.
| Team self-check (5 minutes) | Yes | Not yet |
|---|---|---|
| Does your long-term project have its own scorecard, separate from the cash cow's? | ☐ | ☐ |
| Are the stop rules (money, time, metric) written down and signed? | ☐ | ☐ |
| Is at least one leading indicator reviewed monthly, not only profit? | ☐ | ☐ |
| If the team lead changed tomorrow, is there one page explaining why the project exists? | ☐ | ☐ |
| Does the team have enough runway (money, time) to reach the first milestone? | ☐ | ☐ |
Three or more "Not yet": your long-term project is being judged by a short-term yardstick.
Primary documents first: SEC filings, the bankruptcy filing, releases from the companies and from government agencies. During research sec.gov could not be opened directly; excerpts were checked through search, as noted in the research file.