The ridiculousness of sentiment analysis

Profile picture for user gonzodaddy By Den Howlett March 14, 2018
Summary:
Sentiment analysis is an oft flaunted but utterly useless undertaking in the B2C world. Here's why.

Yesterday I tweeted this message:

This followed an unexpected phone call from a marketer who had emailed me asking for a bunch of data I would never pass to a no-name research outfit at any time and about which I had made the mistake of informing her in email. Vanity stats (page views, volumes etc) are irrelevant to what we measure and while we look at those metrics, they don't especially figure into the way we analyze performance or put our efforts.

It turns out that the marketer had made the mistake of assuming that because their client was mentioned in an article that this amounted to some sort of marketing. I guess they work on the 'any coverage is good coverage' principle whereas we were talking about the company in a specific context that could only be tangentially construed as marketing.

Needless to say, I almost blew a fuse, especially given that the email turned out to be deceptive in nature and hiding the specifics of why they wanted the data  in the first place.

Mindless sentiment analysis

I say nearly because an incoming call I needed to take interrupted my tirade conversation. None of this was helped by the fact that Yet Another Mindless PR Company I'd received a pitch that included a set of pretty illustrations talking about United Airlines sentiment analysis trends.

There's no surprise in the fact UA are having a tough time given that one of their Flight Attendants stuffed a pet dog into an overhead bin, leaving the owner/passenger to discover on landing that the poor animal was dead, presumably through suffocation. As Jon Tweeted:

FWIW I've not flown UA since just after 9/11 and have no intention of going back. Everything I hear tells me that they continue to shale frack nickle and dime customers at every turn, even those that are most loyal. They're not alone of course, but UA does have a storied history of customer service foul-ups. Who can forget United Breaks Guitars and the near 11 million negative views on YouTube?

At least this time around, UA has quickly made a grovelling apology. But that's not my point and, I'm sure, this tragic incident will play out in the Theater of Ambulance Chasers courthouse in due course.

Back to the Yet Another Mindless PR Person. Here's the sentiment analysis that was sent over. Not once, not twice but three freaking times.

 

from a stupid PR person
UA sentiment analysis 14th March, 2018

Breathtaking

It was accompanied by the dross, breathtakingly insightful, no sh&t Sherlock commentary that:

This latest incident could be another massive blow to United's reputation unless upper management takes control of the storm on social in an effective way. In this instance, United will need to do more than just apologize -- they will need to provide solutions and reassure their wide customer base, and to do so intelligently, they must be mindful of the overwhelming responses they're already receiving about the incident. Leveraging social listening during a crisis can help any brand gauge the right response, and hopefully, make a comeback.

And your point is.....????

I don't know if these people have noticed, but following the United Breaks Guitars fiasco, the company barely missed a beat in reporting earnings.

The most recent incident, while wildly more egregious than the earlier one, will have almost zero effect on United. Unless...a few large corporate specifiers put ethics to the front of their choice parameters and say enough is enough.

The likelihood of that happening is almost zero because, like the other major U.S. airlines, United operates what are near monopoly hubs that act as choke points for others. You wanna go to the Bay Area from Chicago, Houston, Frankfurt or Denver? UA is pretty much your only realistic choice. In short, the U.S. airline majors operate as a set of cartels, ostensibly in competition, but in reality, having 'safe' harbors into and out of which they are the mob bosses owners with very little to lose.

The second problem with the illustration is that it is missing a factor. I've corrected that:

stupid PR sentiment analysis

But the overarching problem is much more worrying. You might think that x-million negative comments on social media might damage a brand but it doesn't because each of those are individual. That's why I say that it would take large contracted corporates that spend millions with specified airlines for any impact to be felt - if that was possible or practical.

How about dieselgate?

It's worse. Remember the 2015 VW diesel emissions cheating issue? It occupies an interesting and exhaustive section in the company's 2017 annual report. You might have thought that the levying of multi-billion dollar fines and court awards against the company which, I understand now exceed $20 billion would have had a significant impact on performance in successive years. Especially given the evidence heavily suggests that high-level VW executives lied were economical with the truth to regulators to the point where criminal charges were in play. Yes, there was a certain amount of sword falling but imagine my surprise when I saw this table of executive pay:

VW exec pay

OK - so the pay figures are actually less than would have been allowed under the pre-2016 exec pay scheme but that doesn't prevent the CEO from bumping up against the overall comp limit. And, VW made more money.

Which suggests to me that apart from taking calculated risks that can, and in fact did backfire, (sic) this company (and it is one of many) is immune from the impact of government imposed penalties, let alone the outrage we see in media and social media, which continues to this day.

My take

I could go on with examples of marketers and PRs pitching or measuring the wrong stuff. It happens on a daily basis. What these folks have clearly missed is that sentiment analysis, government fines and moral outrage have little impact on business results.

The acts of the individual are rarely, if ever effective in this context. We have certainly seen examples where a negative comment via social media - often Twitter - from a celebrity or high profile politician has a massive impact on the share price but those are exceptions and often a short-lived problem.

What about the flip side? Where there are genuine choices then sentiment analysis can matter and especially among collective groups of interested parties. And it is here where we see the greatest impact.

As for the rest of us unwashed, railing against The Man? Not so much. So please Mr/Ms Marketer/PR wonk. Just stop it and apply attention to more useful matters. If you're an agency and making this a big deal for your client then I suggest you're doing them a dis-service.