In the final issue of Game Developer magazine (June/July 2013), editor-in-chief Brandon Sheffield did something the publication had spent two decades asking of other developers: he wrote an honest postmortem of his own operation, a few months shy of its twentieth birthday.

What went right:

  • Print prestige — contributors would rather be published in print than on the sister website, and Sheffield is clear-eyed about why: it is the same instinct that “we grew up with it, so it must be better,” the one that sends indie games to consoles that sell worse than the alternatives.
  • Annual state-of-the-industry reports — the Salary Survey and Front Line Awards ran for over a dozen years and remain the only contiguous data on developer pay and tool rankings.
  • A hardcore crew — from around 2005 the magazine was produced by two full-time people plus part-timers, shipping 52 to 96 pages a month with multiple edit and art passes.
  • A frank, honest tone — the postmortem format itself. Sheffield treats that as the magazine’s main public good: the industry has few places where developers are encouraged to be honest with each other.

What went wrong:

  • One revenue stream, consolidating — B2B ads meant industry contraction hit hard: tool vendors merged until “a hydra of advertisers” became a few sales points, and recruitment ads moved to the web. For years nobody was assigned to magazine sales, and at one point pages were sold for less than they cost to print.
  • Digital too little, too late — the app shipped in the final year, and subscribers reported a better experience reading downloaded PDFs than using it.
  • No internal support — the art director was budgeted four or five days of work per issue; new ideas had no backing until the last year, when support arrived and was filed as overhead.
  • Slow bleed into a skeleton crew — editors absorbed production work that once belonged to a five-person department, which made seeing the long view harder with every cut.
  • Never really knowing the audience — audio professionals were 2 percent of surveyed readers and the audio column survived anyway; programmers complained the articles were not innovative enough, then went silent when asked what they wanted to read.

The blunt part is the accounting. The magazine was profitable every year it existed — not every month, but every year. That was not enough, because the contribution kept shrinking and the parent company would not keep skin in the game. A small, declining profit on a lean operation is a slow exit, not a reprieve.

Two lessons transfer cleanly to indie studios. Revenue concentrated in a few consolidating buyers is fragile no matter how profitable this year looks, and asking your audience what they want reliably produces criticism, not specifications.