Blog > Text Analytics: Finally, Metrics that Guarantee Better Outcomes

Text Analytics: Finally, Metrics that Guarantee Better Outcomes

Vanity metrics look backwards. Text analytics are the opposite


From a content marketing standpoint, the real problem with vanity metrics – social media counts of likes and views – isn’t that they’re manipulated by platforms, subject to the whims of an algorithm or heavily influenced by status games.

It’s that they’re backwards looking. Vanity metrics tell you what captured a bit of attention in the past, not how to change what you’re doing to get better results.

For example, consider this post from Larry Fink, published in 2020 and denouncing the January 6th storming of the US Capitol and attempted coup.

BlackRock’s most successful social media post in Q1 2020

It was a severe outlier. Over a period of two months, BlackRock’s post engagement exceeded 300 reactions on only two other occasions.

BlackRock’s LinkedIn engagement profile,
posts published between January and mid-March, 2020
Source: FinText

But Larry Fink can’t wake up the morning after and denounce the coup again. Also, his day job is selling funds, not commenting on explosive political events. In other words, the above post’s success told BlackRock nothing about what it should do on any following day.

Don’t underestimate momentum

Company-generated content, especially for large companies, is process driven. It’s an industrial process, but it’s flying blind. It’ll tend to generate similar outputs to what it has before, with broadly similar results.

Is a company’s process, as constrained by its resources, the best it can be? How should companies improve their processes? Text analytics offers some tangible answers.

The corporate content machine
Source: FinText

What’s text analytics?

Any publisher that consistently captures an audience’s attention isn’t doing it by virtue of manic creativity. There’s a process behind it. For any new piece, many of its properties are unwittingly pre-determined before it’s even created!

Text analytics looks at the underlying properties in volumes of texts. They’re a full reflection of a company’s content-creation process: how much, who writes, on which topics…text analytics not only reveals these properties, but helps teams improve.

Getting back to BlackRock’s LinkedIn activity, text analysis of their social media posts showed the impact of including a direct call to action (CTA) verb.

Nearly 50% of BlackRock’s posts included a CTA verb (most commonly – ‘Learn more’, followed by ‘Read more’). In our BlackRock sample, average engagement on posts with a CTA was nearly twice as high as those without one.

Even better, applying the same analysis to different investment managers allows you to compare.

  • What’s the average output and who’s being extra prolific?
  • Is a company producing a lot of ESG content, or maybe not enough?
  • How should the content mix be changes to achieve better results?

Equally, one can apply the same analysis to other sources of information your audience enjoys.
Content published by trusted sources is more likely to be read – so what are they doing right?

Yes, What are they doing right?

Multiple research studies show that an audience’s attention is governed by trust.  We pay more attention to sources we’ve already had positive experiences with. And positive experience boils down to just three factors:

  • One, new information…
  • Two, about topics the reader cares about…
  • Three, served in a language readers enjoy

When we start thinking about text in bulk, we can uncover these patterns – good or bad – being shaped by the process. Either way, marketers need measurements to improve. Vanity metrics simply can’t be those metrics, but Text analytics can.