Editing data

(Picture by David of Earth)

The mighty, multi-format Ben Hammersley has been doing some thinking about how magazines should work to make the most of every format that they are in, and the need for metadata to be part of a digital magazine’s workflow.

You can catch up with his thinking starting here, then here and here. Here’s my thoughts on a bit of where he’s going with that.

He’s very right in saying that “every medium… is uniquely gifted in a particular way.” He could also add that most publishers in every single given medium aren’t very good at taking advantage of that. However, for the sake of argument, let’s go with his assumption that publishers are looking to make the most of digital platforms, and prepared to alter their workflows accordingly.

He goes on to talk about the need for metadata to feature in articles, and gives this example in pseudocode:

<subject identifier=”Design”/>

<subject identifier=”Industrial Design”/>

<subject identifier=”Bill Moggridge”/>

<subject identifier=”IDEO”/>

It’s the hot design company hired by <company identifier=”APPL”>Apple</company> to create its first mouse, (and by <company identifier=”MSFT”>Microsoft</company> to create its second), by the <company identifier=”UKPO”>Post Office</company> to rework the postbox, by <company identifier=”MUJI”>Muji</company> to create its wall-mounted CD player and by <company identifier=”PANDG”>Procter & Gamble</company> to reinvent toothpaste tubes. It made the <company identifier=”NOKIA”>Nokia</company> N-gage, the <product identifier=”PALMV”>Palm V</product> and the <product identifier=”HEADAIRFLOW”>Head Airflow</product> tennis racquet. Now <company identifier=”IDEO”>IDEO</company> is being retained by <person identifier=”Barack Obama POTUS”>Barack Obama</person>’s White House to help to reinvigorate the American civil service; by the government of <country identifier=”ICELAND”>Iceland</country> to help the country to innovate its way out of <newsevent identifier=”2008-9FinancialCrisis”>financial crisis</newevent>; and by the <company identifier=”KLLGF”>Kellogg Foundation</company> to reinvent education.

Now think of all of the stuff in that we can index against. No use at all for the print magazine, but once you come to put it into digital form an archive full of typescripts like this would be so full of options it starts to get giddy. Give me a thoughtfully built mark-up standard and a year’s worth of, say, Vogue, and I’ll break your heart with beauty.

I understand how such a page would be beautifully indexable in various ways by any decent content management system, and hugely useful for publishers wanting to create future compilations of articles on different subjects (Ben uses the example of a century of Christian Dior dresses). I agree with his analysis, but would like to add a qualification.

If they are designed to be standard throughout, I implore publishers not to make these identifiers visible in the text. I can’t stand it when an online article is filled with distracting hyperlinks that don’t relate to me, most of which seem to be there mostly as googlebait. I know who Barack Obama is, without the need for a hyperlink either to his own website, or an automatically indexed page filled with thousands of other articles that mention him – many of which, as in Ben’s example, probably hardly relate to him at all. And if for some reason I don’t know who he is, I can always Google him. I don’t need a magazine helpfully explaining every other word with a hyperlink, just in case.

I guess what I’m trying to say is that publishers should avoid providing reader-visible metadata that’s not edited or curated. It’s just irritating. Every time I mouseover a link supplied in an article on, say, The Guardian website, and find that it’s just linking to an automatically generated list of other articles on that theme, a little part of me screams inside. And sometimes, outside.

A great print magazine invention is the sidebar – a place where supplementary information, glossary terms and other, related stories go. Carefully chosen, and there to read if you’re interested in the topic and want to learn more, allowing the writer to keep the narrative tight.

In my opinion, digital magazine metadata and extra info/photos/graphics should work like enhanced sidebars, that is to say context-relevant, and part of an editor’s job (and/or journalist’s) will be the careful curation of that information, as well as of the article itself. Track down the relevant extra information, put in audio snippets from the journalist’s interview, personally choose other articles where you should read on, point us to the sources that the writer used in writing the piece.

Anything on show to the reader must be treated with the same care as the content itself – and that includes links and metadata. Give the geeks more to scrape if you must, but hide it away where most of us don’t have to see it. We’re trying to read.

In summary: what needs to be avoided are automated structures or thoughtless implementations that always point to the same company profile or share price graph, regardless of whether the story is in a sports feature or a business one. Involve us in the story, then use metadata to pull us in deeper. Take the time, and do it properly, through human intervention. Don’t push us away into automated, irrelevant cul-de-sacs just because the technology means that you can.