A thought that showed up this morning on a site called Flarrio. I’ve been thinking and writing a lot about artificial intelligence lately, especially in connection with digital commerce and, more specifically, the use of intent data to predict customer behavior.
Tag Archives: artificial intelligence
I suppose it’s time to comment on Content Commoditization.
Several prospective clients have commented recently that they urgently “need lots of content.” This isn’t surprising as Content Marketing matures as a discipline. It’s now broadly accepted that marketers can attract a lot of promising leads by dangling Thought Leadership content for prospects to download from gated sites.
A lot of CMOs evidently are feeling that they’re getting a late start in the inbound marketing game. Our conversation may begin with a very specific need for an eBook or case study, but lately it’s often ended with a rueful admission: “We’ve got a lot of holes to fill in our content – we’ve got nothing.”
That’s both gratifying and troubling to hear. When multiple marketing execs express a need to populate a content library in a hurry, it’s natural to worry about the quality of the material and what’s likely to pass muster these days.
What started me thinking about this was Ann Rockley’s post from a year ago on “machined content,” which resurfaced this morning in an email from the Content Marketing Institute. I’ve been seeing a variety of references lately to robotically-constructed content – articles concatenated from raw data by algorithms. Wise guy friends who know I make a living as a writer share these things with me – they can’t help themselves.
In fairness, that’s not what Rockley is talking about. She’s suggesting that original content can be repurposed for use in multiple forms and media, with the aid of software. Which is fair enough, assuming the original source content is well thought out and well crafted – generally meaning, at least for now, that the content was written by a human.
Rockley’s original post was half of a point-counterpoint with Jay Acunzo, another content marketing pundit, who shoved back hard at the suggestion that robots will ever be literate enough to replace human authors. Both posts are worth reading for anyone with a professional interest in content.
I’ve seen some of the results of automated writing. My sense is that, if an algorithm can convert a box score into a short, pedestrian recap of a ball game, that’s probably a reasonable expedient for some media. On the other hand, I’ve seen attempts to teach artificially intelligent agents to write creatively, and come away with the sense that I’m likely to be in a comfortable retirement before a robot can replace a literate meat puppet.
I think automation is a side issue, but something worrisome is afoot. That something is commoditization of marketing content – a more general homogenizing of content into prosaic, plain vanilla, placeholder prose that fills a gap in the content arsenal, but only just.
I see evidence of commoditization every day. Content practitioners wind up on a lot of email lists, and we’re constantly getting each other’s stuff. And of course I notice – I have a professional interest in what passes for marketing content these days.
Apologies in advance if this sounds arrogant, but I invest a lot of myself in what I do, and a lot of the marketing content I read is crap.
I take it on faith that it’s written by carbon life forms. It’s search engine optimized to the nines – finding it certainly isn’t a problem. But it’s dull, weighted down with past-peak business jargon, and each piece is practically indistinguishable from what the sponsoring company’s competitors are tweeting this week.
Not all of it, of course. Content marketing is maturing but it hasn’t become decadent yet. Plenty of creatives are producing outstanding stuff that speaks to real customer needs, makes an elegant case for a product, answers serious questions and casts a flattering light on a brand.
But demand for content is increasingly universal, and the creative talent pool, like anything else, is defined by a bell curve. A lot of what I see seems to be sourced from the big hump in the middle of the curve.
My point is not to gripe about bush-league writing, but to caution against the “more is better” approach to content development, which might lead to a dumbing down of quality standards.
Automation, to return to my original point, is unlikely to be an important driver for this. There are other ways to source cheap, commodity content. There are plenty of low-end content mills farming out authoring to amateurs looking to break in to the writing game. If creative quality is a secondary concern and the need is simply to generate lots of verbiage in a hurry, that can look like an attractive option.
The problem, though, is that content quality actually matters. This is not just something content pros tell themselves to continue feeling relevant. It’s been an article of faith ever since content marketing became a thing, and there are pragmatic reasons for this. Bad quality content will frustrate your SEO efforts – Google degrades the position of lousy content in its search results, on purpose.
Net: I doubt that humans who create original content for a living have much to worry about from robotic competition. Ann Rockley probably has a point about using algorithms to aid in repackaging content for reuse. In a market where CMOs are trying to get caught up quickly, we mere mortals can use the help.