662
post-template-default,single,single-post,postid-662,single-format-video,theme-stockholm,qode-social-login-1.1.3,qode-restaurant-1.1.1,stockholm-core-1.2.1,woocommerce-no-js,select-theme-ver-5.2.1,ajax_fade,page_not_loaded,menu-animation-underline,side_area_slide_with_content,wpb-js-composer js-comp-ver-6.1,vc_responsive

Ron Howard is right: disinformation killed “Solo: a Star Wars Story”

How many of you saw Ron Howard’s “Solo: a Star Wars Story” when it came out in May 2018? I, for one am not ashamed to admit I listened to the whispers of why the movie wasn’t going to do well, read about the premiere buzz and promptly waited a couple of weeks before seeing it in the theater. I almost waited for it to come out on iTunes.

Man, was I wrong. The movie was better than Episodes 1-3 of the original Star Wars franchise; maybe better than the thrilling but Ewok-ridden “Return of the Jedi.” I remember hearing that there was something going on with the negative reviews on Rotten Tomatoes, but I figured it wasn’t worth more time. The movie was good and that’s all that mattered.

Then came this Guardian.co.uk story today about Howard and his Star Wars movie:

Speaking on the Happy Sad Confused podcast, the film’s director said that, while he noticed “pushback” from the previous Star Wars film, The Last Jedi, he felt “aggressive trolling” was a crucial factor.

 

“It was pretty interesting,” Howard said. “It was especially noticeable prior to the release of the movie. Several of the algorithms, whether it was Metacritic or Rotten Tomatoes, there was an inordinate push down on the ‘Want to see’ [score on Rotten Tomatoes] and on the fan voting.”

 

Now that I’m aware of computational propaganda–defined as “the use of algorithms, automation, and human curation to purposefully distribute misleading information over social media networks”–it becomes easier to see that the trolls used “Solo” as an opportunity to up their game.

 

Why are bad reviews a threat? It’s just a movie…

Ask Disney’s accountants and lawyers if it’s just a movie. The two movies prior to “Solo” printed money at the box office. “Rogue One” grossed a little over $1bn and “The Last Jedi earned $1.3 bn in theaters worldwide. “Jedi” had its own trouble with trolls and manufactured outrage. To date, “Solo” has underperformed, with just $448 million in revenue from ticket and home video sales. In the midst of the run, Disney Chairman Bob Iger even admitted an error in releasing the movie within six months of another franchise tentpole. Had the movie performed in line with its predecessors, what would have been the impact on licensing agreements? Stock and financial performance?

But what if all of this was engineered outside of their control? Would someone do that? It’s a lot of work, why would they even try? That’s the thing with computational propaganda, it is a threat that hides in our inner dissonance. On the one hand, we are all coming around to understanding that there are people out there who can and have hacked and deceived us online. On the other hand, we discount that they’d come after us. This is the blind spot necessary for propagandists to operate in.

 

Pull back the camera to see the usual suspects

In a March 2019 story on troll-proofing movie review sites, Adi Robinson at The Verge says “over the past couple of years, the highest-profile review-bombing campaigns have targeted blockbuster films for the sin of casting too many women and people of color.” In the case of “Solo”, the woman up for crucifixion was Kathleen Kennedy, head of Disney Studios. Kennedy and the two original directors of “Solo” clashed over vision, and she put Howard in place. Prior to this, Kennedy had been excoriated by the alt-right for engineering the feminist takeover of “The Last Jedi”. Too many female heroes. One group even took credit for manipulating Rotten Tomatoes against the movie.

So the campaign against “Solo” was an exercise in using computational propaganda to mop up anything else Kennedy touched. And it worked. What’s my proof? Also in March, Rotten Tomatoes took steps to protect the release of Marvel’s “Captain America” from trolls. Instead of allowing trolls to manipulate a metric around interest in seeing the movie, they shut the analytic down until the movie opened. Although a year late, the biggest of the movie review sites clearly understands it’s being played and is starting to fight back.

 

What’s the takeaway for communicators?

Disinformation is like home security. Be vigilant based on your risk. Communicators need to understand a few things about the conversations taking place around them.

  1. What are the top 500 trending keywords, phrases and images in #crazyland–what are they saying about you on Reddit, 4Chan, 8Chan and other alt-right channels?
  2. What’s the risk exposure; meaning are there troubling overlaps in what they say versus language in your channels?
  3. Are they saying your name or using your logo on the rest of the public internet? If so, where and when?
  4. Are these keywords and media being used in your channels; and in what context?
  5. If so, who is using them? What do their networks look like, and is there a way to examine their conversations?

 

To continue the home security comparison, you should always lock your doors–meaning be aware of the conversations about you and your industry on the public internet. If you have more things of value, take further precautions. So if your agency represents women, journalists, women journalists, people of color, themes, industries or products of interest to the alt-right, you should execute a monitoring strategy as part of a larger effort to build a threat profile.

A threat profile, according to Stephen Irwin “includes information about critical assets, threat actors, and threat scenarios. A threat scenario is an illustration in which one or more threat actors can mount one or more threat actions in an attempt to compromise an identified critical asset by exploiting both vulnerabilities and inadequate safeguards (Dziadyk, 2011). A threat scenario campaign is a series of related threat scenarios that are used together as part of an APT for a common objective. An organization’s threat profile includes all of this threat information and presents a clear and detailed illustration of how each of these components are used together.”

In English, this type of threat profile is a very prominent part of your crisis plan that requires constant attention. Good PR people shouldn’t be new to the idea of threat profiles.

If you work for a movie studio or similar major content creator, you should have already called an AI-powered social media intelligence firm like Brandwatch/Crimson Hexagon or Austin, TX-based New Knowledge. Five years ago.

 


Important read:

Bitnami