All across the world, in cubicles and board rooms, bars and conferences, marketing executives are asking themselves: What do women want? (It’s usually followed by: Should we make a Pinterest account for the company?).

I can imagine the scene that must have unfolded at IBM leading up to last week. Under fluorescents in a boardroom, a group of such executives — spreadsheets strewn in front of them, maybe a few Jezebel blog posts open — had an epiphany. A fantastic idea. One sure to lure women closer to the brand. A fun, IBM-hosted #hackathon that would boost the image of the company — nay, the tech industry as a whole — as a female-friendly space.

Hairdryers. Women love hairdryers, someone in that IBM boardroom said out loud. Yes, the others said, looking up from their spreadsheets, we’ll do something with hairdryers. This is a good idea that we should pursue. 

Spaceballs hairdryer animated gif

And so, #HackaHairdryer was born. Women, the fairer, hairdryer-loving sex, were sure to jump on board. The company made the hair-centric hackathon live on their website in October, and it was shared sparingly on social media by other Twitter users through November.

On December 6, 2015 — in a tweet that’s since been deleted — IBM shared a link to the page to its (at the time) 265,000 followers. It was 26th anniversary of the Montreal Massacre, when a misogynist killed 14 women studying engineering while they were working in a lab.

Blow Out

#HackaHairdryer went over like a hairdryer on high dropped into a bathtub.

crowdbabble_social-media-analytics-how-to-track-measure_crisis-management_hackthehairdryer_angryreaction

The campaign short circuited on launch because it was wired to go viral in a bad way. First, the campaign was aimed at a large and very active segment of Twitter users: women in tech. Second, #HackaHairdryer touched a hot issue: women in tech. Third, because #HackaHairdryer touched the zeitgeist, it was picked up by media. Just as #AlexFromTarget gave news outlets a face for teen social media habits, #HackaHairdryer became an easy joke and a sure-to-sell-ad-space way in to the conversation around misogyny in STEM.

Using Crowdbabble, we analyzed how #HackTheHairdryer sent IBM into a downward shame spiral on Twitter. How did the failed initiative affect the brand?

Before and After

In the month before the crisis, IBM averaged 147.7 favourites, replies, or retweets per tweet. According to Crowdbabble, engagement was low at just 5.8% (a fraction of 263,606 followers engaged with IBM’s tweets). Below, a visualization of total favourites, retweets, replies, and mentions from November 5 to December 5 2015.

Engagement_Over_Time (7)

For the first ten days of December for the year prior, the statistics are similar: 3.2 tweets per day, with 88.5 engagements per tweet. @IBM was growing, albeit slowly.

Then, #HackAHairdryer happened and things got hairy for the brand. On December 6, engagement went through the floor.

Engagement_Rate

Engagement rose the next day as people reacted to IBM’s apology. For the first half of December, #HackaHairdryer pulled the engagement rate down to 2.3% — even more dire than the previous average of 5.8%.

The #HackAHairdryer discussion on Twitter detached from IBM’s handle, perhaps mercifully, as women in tech used it to vent all grievances relating to misogyny in tech. This negative attention grew exponentially when a much larger group — everyone who did not want to be viewed as anti-women — jumped on board to condemn the campaign and distance themselves from the brand.

As shown in Crowdbabble’s customer service analysis, mentions of @IBM were low before the campaign but have stayed uncharacteristically low since the fatal tweet.

Response time

On December 6, in a Gizmodo story categorized under “Pinkification of Science”, author Mika McKinnon wrote slammed the company for “dumbing shit down because dem womenz can only understand technology when it’s the tools of beauty.” I can imagine a lot of yelling about hairdryers ricocheting around the halls at IBM, emanating from the same boardroom where the campaign was conceived.

Adding a Diffuser

Within 24 hours, IBM issued a formal apology on Twitter and deleted the original #HackAHairdryer tweet. But it was too late. After Gizmodo, The New York TimesThe GuardianFortune, and others followed with stories and editorials slamming the campaign.

While the pizza conglomerate DiGiorno apologized to each offended person individually after its misogynist Twitter gaffe in 2012, IBM tweeted one message aimed at everyone. The response was mixed.

Mixed replies to apologyy

Women in tech were not done venting about the campaign, and others replied supporting IBM and insulting its “feminazi” critics. The support from misogynists of IBM and the continued criticism from women compounded the crisis, solidifying the brand’s new image as a creaky old tech giant not built for the future.

Crowdbabble’s top follower tool (which allows sorting by mentions, replies, and retweets) shows that before the crisis, two of IBM’s top ten followers on Twitter were women. In the week since, women now make up five of ten of IBM’ top followers. If this sounds like good news, it’s not.

IBM most influential followers

The most influential new female followers are engaging with IBM only to criticize the company.

crowdbabble_social-media-analytics-how-to-track-measure_crisis-management_hackthehairdryer_mostinfluentialfollowers-angrytweet

Ouch.

A Disaster That Might Not Grow Out

IBM’s attempt to reach out to women, an adventurous foray into the 21st century, could not have gone worse. The #HackAHairdryer campaign stamped IBM as outdated and out of touch. But if tech giants are still dominated by men, will branding itself as an old-fashioned boys club help or hurt the company in the long term?

Mad Men animated gif: what do women want, who cares

Follower growth actually spiked after the #HackAHairdryer tweet. In the days since, the growth rate has plateaued.

Follower_Growth (2)

Based on the rolling engagement trend over the past 12 days, the forecast for IBM looks murky. Followers may be up, but engagement overall is way down. Below, an engagement graph from Crowdbabble showing favourites, retweets, replies, and mentions from December 1 to December 13.

Engagement_Over_Time (8)

The stigma from the social media disaster will make Twitter users, and the media, much more sensitive to any future missteps from IBM. #HackAHairdryer’s reception could also paralyze innovation in marketing at the company. Like Audi, IBM may cling more tightly than ever to tried-and-tested strategies after the social media disaster stained the brand.

The impact of the crisis and apology on @IBM’s overall brand message on Twitter is shown below:

@IBM_Tweets_Word_Cloud_12012015_12132015

The keyword analysis, generated on Crowdbabble, shows the words most used by IBM and most retweeted by others since December 1 2015.

The #HackAHairdryer initiative will likely be held up in boardrooms at IBM and other tech giants as an example of the dangers of risk-taking on social media. Worse, the hairdryer misstep could be used as an excuse to avoid any initiatives aimed at attracting women to STEM. Let’s just make a Pinterest account for the company, misguided marketing executives might say, we don’t want to be IBM. 

How Not To Be IBM

On social media, intent is irrelevant. IBM’s effort towards being woman-friendly did not come through in the execution (or conception, or any part) of #HackAHairdryer. What IBM was trying to do didn’t matter. The company’s “let’s just paint it pink” approach revealed hair-raising misperceptions about women that were more deeply rooted than its hot air about including them. Below, a cached version of IBM’s #HackAHairdryer page:

crowdbabble_social-media-analytics-how-to-track-measure_crisis-management_hackthehairdryer_ibmwebsite

Users will reimagine a brand’s intent based on the final product campaign that brand shares — reverse engineering the process from the outcome. A sexist tweet, Twitter users vented, meant IBM’s entire project (and company) was sexist.

Twitter users rush to condemn offensive tweets, so as to not appear to condone them. If criticizing a brand becomes an “I’m not terrible” bandwagon on social media, the brand should act as quickly as IBM did to destroy all the evidence (tweet, website) and apologize. Online, erasing the evidence is impossible, but it has become shorthand for a sincere apology and admittance of wrongdoing.

Finally, brands should check their campaigns for crisis potential before launch by asking: what could prevent our target group from participating? Is there a downside for them? Those simple questions could have helped IBM see that #HackAHairdryer was not a campaign, but a clever trap for both women and the brand. #HackAHairdryer not only sullied the company’s reputation as a smart innovator, but made it impossible for women to participate without casting themselves as trifling girly-girls.

crowdbabble_social-media-analytics-how-to-track-measure_crisis-management_hackthehairdryer_spaceballs

The downside for participants was huge, but apparently unseen by IBM.

The #HackAHairdryer crisis still hasn’t been straightened out. As of December 11, there is a new tweet with the hashtag every five minutes. IBM’s follower growth on Twitter is in decline. As it turns out, women want more than hot air about their inclusion in STEM.

The quick fix fairytale of helping women into tech with stereotypes that keep them out of tech is over. Welcome to real life, IBM.