Among content producers the criteria Google News uses to choose its top stories is infamous for its inscrutable, often perversely unpredictable choices.
Break a huge story in a U.S.-based publication about an important change in politics, technology or social crisis and, by the time it shows up on news.google.com's home page your story will be in the No. 39 spot, well below the See All xxx Sources link, while the top headline and photo will be a version rewritten from a wire-service summary of your headline, heavily localized and republished by the Hindustan Times, which may mention the wire service as a source, but not you.
The news judgment – or at least the sourcing – of the algorithmic Google Gods is more predictable now than a year or two ago, but the priorities remain somewhat opaque and often self contradictory.
The top item in the Technology section headlines at 9 a.m., for example was packed with photos of Virginia (Ginni) Rometty, who was elected yesterday as the first female CEO in the long, iconic history and Neolithic culture of IBM.
True, the transfer of power was so smooth that ITWorld columnist Chris Nerney suggested it should be the model by which other corporate transfers of power should be made. So maybe the lack of drama reduced its chance for screaming headlines.
Still. Despite the photos of Rometty, almost all the headlines referred to the other decision IBM's board made yesterday – to buy back $7 billion worth of its own stock.
By 9:30 the featured top spot reoriented to highlight stories focused on Ms. Rometty's promotion, background and likely priorities for the largest but least-influential-for-its-size company in the computer business.
The early focus on dividends and stock buybacks might have based on the audience available at the time: movement of any pile of money too large to fit in an attaché case is more important to early-rising, early-trading investment managers than a revolutionary choice of chief executive at even a large, wealthy company.
Just as likely, though for the same reason, was that the bulk of stories available on IBM during the early morning hours were from financial-news sources that would focus on dividends and stock buybacks as being more important than the movement of people.
Or the placement might have been because the announcement of the stock buyback and dividend to stockholders hit the wires a few minutes earlier than the announcement that Big Blue promoted Virginia Rometty from senior VP in charge of sales, marketing and strategy to president and CEO, reporting to current CEO Sam Palmisano, who remains chairman of the board.
It may be lingering sexism, though that's unlikely in the genderless algorithm that chooses the "most important" stories for Google.
Or it may have been an understandable bias of the programmers creeping in to the news-judging criteria that would reduce the perceived importance of any announcement involving personages identified in their own press releases using excessively preppy or cute nicknames (Ginni in this case, though "Skip," "Buffy," "Trip" or "Turd Blossom" would also qualify).
By 10 a.m. stories on Rometty's promotion dominated the top spot and the number under the "See All XXXX sources>>" link was growing as the number of available stories increased.
I'm not sure how important the process or criteria by which stories are chosen for the most prominent spots in newspapers, web sites and other news outlets actually are to most readers, though they're among the most frequent sources of complaint from those who measure bias against their own favorite causes according to those indirect criteria.
SEO specialists make analysis of content and placement an obsession, in order to improve their own results by aping positive treatment of others – which must be frustrating when you're parsing results from interconnected networks dominated by one player whose choices are (literally) formulaic and whose formulae change specifically to prevent people like you from gaming their system.
Some news- and content-presentation techniques driven by SEO – endless repetition of keywords that may or may not seem appropriate in a particular story, for example – should be important to readers, if only so they know when they're being fed content chosen for its "importance" by criteria a processor would consider valid and critical, no matter how self-servingly trivial they would seem to a human.
I've worked in the news business for hundreds of years, so these kinds of questions are obviously more important to me than to most (if only in judging how important my own stories seem to other editors).
It's still not much more interesting than it was in journalism school, when I studied statistical analysis of word choice, story placement and other criteria of news judgment in classes filled with people who like me, who had to phone expert sources to help with any math more complex than advanced arithmetic.
It's entirely possible the whole question of story placement and how it changes is irrelevant to anyone but journalists or activists.
If so, you won't have read even this far and your inattention will show even those of us deprived of the ability to do higher mathematics that we should continue to ignore our obsessions to focus on those of our audience.
I just thought the choices and placement, at least in this one instance, were a good reminder that no matter where you actually get your news or whose reporting you believe, online, even your ability to find the stories that would interest you are heavily influenced by the assumptions and judgment of people with biases different from yours, hard-coded into search engines whose only concern is to get you coming back for more.
Read more of Kevin Fogarty's CoreIT blog and follow the latest IT news at ITworld. Follow Kevin on Twitter at @KevinFogarty. For the latest IT news, analysis and how-tos, follow ITworld on Twitter and Facebook.