Marketers, Put on Your Steampunk Goggles
Cultural trend spotting, a long promised perk of social media, may soon be a common tool for retail practitioners as data analytics firms become more adept at the technique. The obvious goal of measuring sentiment running through online chatter — blogs, online discussions, Twitter, Facebook, Pinterest, etc. — is to identify trends before they become commercially influential, thus getting a jump on competition. IBM is demonstrating its skills in this burgeoning area with the recent announcement that it has identified "steampunk" as a style that will have a sizable influence on fashion, design and media in 2013.
The origins of steampunk can be traced to the science fiction of H.G. Wells and Jules Verne and emerged prominently in more recent media such as The Wild, Wild West film starring Will Smith and The Prestige. It is a retro-futuristic interpretation of Victorian era imagery with campy, goth/punk overtones. (Got it? Think Captain Nemo meets Lady Gaga.) The sensibilities of the "sub-genre" are diverse, but often involve cogs, clockwork, frock coats, brass, leather and goggles.
IBM, as part of its Birth of a Trend project, found steampunk bubbling up as an emerging style movement by analyzing over a half million social media posts and messages.
"We first started looking at steampunk around about 2008 — the beginning of 2009. We were starting to see this phrase being used in contexts which were in a different cultural domain from fiction," explained Dr. Trevor Davis, IBM consumer products subject expert, in an interview with RetailWire.
"Birth of a Trend is an initiative that we started to stretch our science and technology to find what I call slow-moving trends that have got commercial viability," said Dr. Davis. "So these are trends with staying power attractive to retailers, brand owners and manufacturers."
Social media analysis must tackle huge quantities of data quickly enough to be practical for businesspeople. There is also the need to apply natural language processing so that, beyond merely pulling search results on keywords, one can recognize when chatter is sarcastic vs. sincere and snatch important comments out of the "background noise."
"The software says, ‘Here are the top ten things.’ You then bookmark those and watch them for a few weeks to see if they move. The ones that don’t disappear, you then build some analytical models around and you track those over time, over geography or as they move across cultural domains."
Dr. Davis refers to "boiling points" or spikes in the social sentiment index that typically occur when the style is reported on in the media. Other sentiment analytics projects by IBM have looked into trends affecting high-end perfumes and food and drink products. Dr. Davis sees a great many benefits for the retail industry.
"By staying ahead of a trend as it develops," he stated in a news release, "a retailer can more effectively control critical merchandising, inventory and planning decisions. Technology can provide tremendous foresight to help businesses differentiate what is a fleeting fad, versus what is an enduring trend."
Discussion Questions
How do you see social sentiment analytics becoming useful in retail and CPG businesses? Which departments will benefit most: marketing, merchandising, product development, etc.?
I think 2013 has tremendous potential in utilizing social media far beyond just “listening” or heading off public relations nightmares. But before sentiment can be used more widely, retailers and brands are going to have to validate social media in the same way they validated that eCommerce trends could be used to help predict product and customer trends in stores. That means making sure that the people who are making noise on social media are the same kind of people who shop the retailer or brand, and making sure that the depth of their impact is measurable—in other words, if there’s big swing in sentiment, does that translate into a big swing in demand or some other visible impact on the brand?
When people are more comfortable that the noisemakers/influencers online are truly representative of the population as a whole, then we’ll see the movement towards using this information really accelerate.
First of all — Steampunk?
Really?
How totally last year!
Several years ago I wrote a book on this topic (The Deviant’s Advantage for any of you looking for that early holiday shopping experience). Rather than restate the whole book—fascinating as I’m sure that would be—let me jump to the bottom line.
IBM should have identified and capitalized on Steampunk prior to 2008, not in 2013.
The answer to today’s specific question is social sentiment analytics (PLEASE stop with these tortured linguistic aberrations) or, what I like to call being aware of what’s happening in the world beyond your little bubble, could be tremendously useful to all aspects of your business provided that you are nimble, plugged in and open minded and don’t move at glacial speed like big companies do.
But … please … avoid the trap of assembling the most spectacular aquarium of jumped sharks the world of marketing has ever seen.
Trend spotting has always been important and risky. Knowing what is of interest right now is a real advantage when making product and assortment decisions. Spotting the trend, making decisions, and getting inventory produced and delivered in a timely manner is the key to success. If the trend is a fad, it may be over before the whole process can be executed. If the trend does not catch on with your target market you may have a lot of leftover inventory.
Developing the tools that provide trend information for members of your target market is important. Your organization also needs to be agile enough to respond quickly.
As many retailers struggle to just keep up or even stay a step or two behind most emerging trends, the concept of “getting ahead of a trend” seems a bit untenable to me. But beyond that bit of cynicism, understanding the emotional influences on store selection and purchase decisions have always been important to retailing.
This is particularly true for retailers who have decided they cannot or will not compete predominantly on price, selection, and other more quantitative delivery points. While understanding consumer sentiment and emotions have application to all retailers at some level, those retail channels the inherently invoke emotional choices and attachments, like fashion and apparel are likely to benefit from emotional marketing techniques than others.
This seems just way too precious to me. It’s akin to Pantone announcing that 2013 will be the year of emerald green (Google it.) Do marketers think that edicts like this will influence customer behavior? Or drive sales in the real world?
This predictive use of available insights and technology is invaluable to all stakeholders of the product lifecycle ecosystem including product designers, marketers and merchandisers, if brands and retailers choose to listen and act on the insights.
In the cacophony of Big Data, the challenge is to take historic insights—which all data points are—and use those insights and turn them into forward-looking recommendations that are quantifiably validated to make sound business decisions. How many forests have been wiped out to produce the tons of paper on which endless charts and tables of data line the walls of brand marketers offices and cubicles? This approach can unveil trends and allow brands to react in-advance further increasing the ‘speed-to-market’!
This does not answer the question, but here I go. Take a young 8th grade girl in a fashionable community in Birmingham, AL. She finds some colorful duct tape and plasters it all over her tennis shoes. She and her gal pals love to instagram so she takes a picture. All her friends see this and run to the store to buy duct tape. Now, if you are a trendy store in B’Ham, preferably the area known as the Tiny Kingdom, and if you are trend spotting, I would suggest that you run to the hardware store and plaster duct tape in all colors all over your mannequins. This is a useful use of social sentiment, IMHO
Getting response and feedback via social media is quick and efficient. The key with any feedback, data or analytics is to make sure it is relevant. In other words, it has to give you information you can act on. Too many times I’ve seen lots of data that is simply a waste of time and energy to go through. It’s not actionable.
Which department will benefit most? Define what information you want and what you will do when you get it. It’s not what department benefits most. It’s what information is most relevant to which department.
Steampunk will definitely remain relevant through 2013 with the much anticipated release of BioShock Infinite—which will be one of the biggest video game launches of the year—so I don’t think anyone is going out on a limb here. Amazon is already touting a tie-in ebook to be released in a few weeks.
RyMat, you had me laughing out loud on this one.
Confirming social trends using brute force data analysis must be a backward-focused undertaking. No wonder IBM arrived tardy to the Steam Punk smorgasbord.
Trend-spotting is an undertaking best pursued by embedded agents. These are people who inhabit the movements themselves and bring them to the attention of commercial interests.
Agency handlers want to tap this energy. Viral marketing will surely follow. Attempts to adapt mass concepts will tend to create very short life cycles.
My rule of thumb: Any social movement that involves home-made costumes is likely to be narrow and brief.
I think the Tank Girl aesthetic is the one to exploit—it’s an aspirational persona in the steam punk world—or as she said:
“Listen up, cause I’m only telling you this once. I’m not bedtime story lady, so pay attention. It’s 2033. The world is *screwed* now. You see, a while ago this humongous comet came crashing into the earth. Bam, total devastation. End of the world as we know it. No celebrities, no cable TV, no water. It hasn’t rained in 11 years. Now 20 people gotta squeeze inside the same bathtub—so it ain’t all bad.”
Complicated subject, but one that is worth exploring and with tremendous potential. I do question the macro trend of layering on forms of artificial intelligence in the social graph to understand consumer trends and preferences instead of just talking with people directly.
As Vahe points out, Tank Girl was an iconic figure and—as I’m forced to point out—it came out in 1995.
Thanks Ryan—or how about Blade Runner (1982) *but* there is an interesting twist: Machine created trends connects with the posthuman/cyborg concept. So I suppose once Steampunk goes big grid, folks in SF/LA/NY will be transforming themselves into cyborgs. But it’s all just schmatta in the end, mark it up and move it out.
Social media analytics currently is being used very successfully in CPG and retail organizations. In fact, it has been being utilized for several years. As sentiment analysis and NLP see improvements, social media analysis will only increase in value. Areas that I have seen successfully use social media analysis are:
As for IBM identifying new trends, I have to agree with previous posters that taking several years to identify a new trend is not all that useful. However, there is no reason that analysis of this type needs to take years. In fact several social media analyses that I have been involved with have taken weeks, not years.