INTRODUCTION: GOING VIRAL
The growth of the internet and a hyper-connected, hyper-mediated world poses enormous challenges for consumer products companies. The numbers alone are staggering: 2.2 billion monthly active Facebook users, 1.8 billionmonthly logged-in YouTube users, 1 billion monthly active Instagram users, 336 million monthly active Twitter users, 289 million monthly active Snapchat users, 250 million monthly Pinterest users, 173 million monthly active Yelp users.
And that doesn’t count blogs, reviews on retail and brand websites, celebrity appearances and endorsements, and countless other forms of public expression.
Although people have shared their opinions for as long as they’ve had language, they’ve never had so many opportunities to do so as they do today. Telling the world about a product you just bought now takes only one click.
Equally simple is finding out what the world thinks of the product you're considering buying. And people do pay attention to others’ opinions; PwC’s 2018 Global Consumer Insights Survey found that social networks were the No. 1 choice of online media that survey respondents used to find inspiration for their purchases.
The second big change is in the speed of opinion sharing. Information and ideas that once disseminated over months, years, or decades now seem to travel with the speed of light. Fifteen to 20 years ago, in the dawn of the social media age, the phrase “going viral” began to be used. Like viruses, expressions pass rapidly from one person to the next and suddenly appear everywhere.
Apparel and other consumer products companies were early to seize on social media’s potential for marketing and customer care. For example, brands routinely use their Facebook pages to generate excitement about new products, promotions or store openings, or to resolve customer problems that weren't addressed through standard channels.
But the impact of social media on the supply chain is more complex, and it has taken longer to recognize and deal with. Messages that “go viral” can make or break a style or even a brand, and a company can quickly find itself with too much or too little product on hand.
Learning to respond to — and stay on top of — the social media cycle is now a necessity for supply chain managers, but solutions are available. Experts in the field say that companies need to consider changes in technology, processes and even corporate culture.
Negative social media
There are endless opportunities for customers or others to criticize a company’s products on the internet. Unfavorable reviews — genuine or orchestrated — can appear on multiple websites and dampen demand for a particular product. A customer may complain to her friends on Facebook about an ill-fitting garment or post photos on Instagram of a product that arrived damaged.
Some social media commentary involves a brand or company as a whole. Commenters may object to a company’s business practices (or perceived practices) or even the personal activities or political opinions of a CEO or a celebrity endorser. In a time of polarized views, it may be impossible to avoid negative reactions altogether.
The challenge for companies is to capture this information and respond to it quickly. If social media commentary is based on incorrect information, this requires correcting customers’ misperceptions. If the negative commentary is valid, it offers an opportunity to fix the problem — for example, by redesigning a style for better fit or by publicly committing to change business practices that customers object to.
When there is no way to counter negative perceptions, a company may have to engage in costly demand shaping — extra marketing activities, price reductions, or special promotions — to clear the excess supply. In all cases, the information about reduced demand should be quickly applied to planning and forecasting to avoid building up excess inventory.
Positive social media
More often, customers enjoy sharing happy experiences. They post positive reviews, tell their contacts how much they like the products they buy, and even take the trouble to create videos on YouTube explaining how they use the products.
Influencers — paid and unpaid — are trend-setters whose recommendations carry extra weight. Mountain climbers are trusted recommenders of technical clothing and gear. Athlete Serena Williams, embroiled in a controversy about “appropriate” tennis wear, recently appeared on court in a tutu and garnered attention for the designer. The outfits of princesses, first ladies, and Academy Awards attendees are scrutinized, discussed, and emulated by millions.
The challenge for apparel companies is to anticipate as much of this buzz as possible, then capture, analyze and amplify it, ensuring that even more people see it and talk about it. In addition, they need to feed this information into demand forecasting models to keep up with the increased demand and not leave their customers unsatisfied.
Perhaps the toughest challenge is the unexpected fashion event — good or bad. In one of the more talked-about fashion statements of 2018, first lady Melania Trump was photographed wearing a jacket with a controversial message when she went to visit detained immigrant children. The White House offered at least three different interpretations of what the message meant, and outside observers proposed still other explanations. Another clothing company took advantage of the uproar to create jackets and t-shirts printed with the opposite message, made them available for order online within a day or two (though they would not ship for several weeks), and sold out almost instantly. Proceeds were donated to charity.
As this example demonstrates, an extremely flexible supply chain is required to capitalize on the short-lived social media buzz about an unexpected event.
The social media tsunami
The world of social media may seem chaotic and ungovernable, but the tools exist to make sense of the chaos, react quickly to the trends it reveals, and even (in many instances) stay a step ahead of it.
The challenge for apparel companies is to be proactive in dealing with social media, rather than being swept away by the wave. This means knowing your customers well enough to be attuned to what inspires them; helping to shape and guide the online conversation about your company and products; and having the agility to take advantage of new opportunities that arise on a moment’s notice.
In another well-publicized 2018 incident, Nike demonstrated a proactive approach by diving headfirst into the controversy over NFL players’ protests of police brutality and launching an ad campaign that featured quarterback Colin Kaepernick, a leader of the protesters. Customers who disapproved of Kaepernick responded by burning Nike products and threatening boycotts. However, more customers approved than disapproved. Nike’s risky move paid off; it immediately added an enormous number of followers on Instagram, Facebook, and Twitter. Christopher Svezia, an analyst at Wedbush Securities, wrote in a research note that Nike management “knows its American consumer well and the campaign featuring Mr. Kaepernick is a positive for the brand and likely its sales.” The company’s stock price hit a record high shortly afterward.
AI for demand forecasting
Traditionally, companies forecast demand on the basis of historical sales records. With social media driving rapid changes in tastes and opinions, historical sales data are no longer sufficient. Companies need to use more current data, and they need to look at data from outside the company. Social media represent a rich source of external information, and big-data analysis techniques, including artificial intelligence (AI), offer new ways to incorporate unstructured as well as structured data into forecasting.
Forecasters and planners can use social media data in several ways. One is to track online conversations about general fashion subjects, such as styles, colors, fabrics, and features, to gain a deeper understanding of current trends. For example, if the social media world is buzzing about the color yellow, planners might adjust their assortments accordingly.
Annibal Sodero, a professor of supply chain management at the University of Arkansas, notes, “Particularly for fashion products, the selling season is too short, and the replenishment cycle is too long, so we need to get it right before the season.” Sodero says social media conversations — not just what is said, but the intensity of user engagement, as measured by metrics such as likes and retweets or shares, as well as the number of connections each commenter has — can be used to calibrate forecasting models with information that is more up-to-date than last year’s sales.
A second approach is to track online conversations about specific products. According to Arijit Choudhuri, senior manager of the North America Digital Supply Chain Practice for Capgemini, ratings on retail websites should be monitored for each product. To eliminate the “noise” in the data, analysts ignore the outliers, or most extreme ratings, and track the averages over time. The ratings history for any product or category can be correlated with demand and added to the forecasting model. Sudden, sharp upward or downward trends may call for demand-shaping measures, including price adjustments.
Joe Vernon, senior manager of Capgemini’s Supply Chain Transformation and Analytics practice, says that, in addition to tracking ratings, companies “troll” social media conversations looking for mentions of the company name and images of products. Search dictionaries may include lists of positive and negative terms (“fabulous,” “doesn’t fit,” and so forth). Findings don’t have to be given equal weight; they can be evaluated based on the intensity of engagement and the connectedness of the commenters.
Findings can also be corroborated through focus groups or surveys. Vernon says, “If social media indicates that teens are not taking to a product as we thought — well, let’s go talk to some teens.” The results of these analytical findings and follow-up surveys should also be incorporated into demand forecasts.
Rather than replacing existing systems, analyses of social media conversations will form a “smart layer” on top of them, Vernon explains. Artificial intelligence can be embedded into supply-chain systems, adding functionality without changing systems that already work well. “That’s one of the biggest things the cloud has conquered,” Vernon says. “Integration used to be so complicated, but new cloud architectures have opened things up. It’s been a huge accelerator.”
Flexible supply chain
Once a company captures and analyzes social media data, it must be agile enough to respond quickly to the trends it discovers. Sometimes, this requires making design changes on the fly — for example, if color trends prove different from what was expected, it can change the dyes and try again. Or, if many customers complain online that product sizing is not what they expected, the company may want to re-size the product and put the new iteration out on the shelves as fast as possible. “That can be frustrating, but it’s helpful to be able to take a better run at it,” Vernon says. More often, the appropriate response to real-time social media data is to manufacture more or less of a product than planned, redistribute it to different stores or channels than planned, or try to shorten cycle time altogether. Any of these responses requires a flexible supply chain.
To position themselves to take advantage of real-time social media information, companies should consider making changes to their supply chain technology, processes and culture. Though agility is usually aligned with efficiency, in some instances (such as redundancy), the two goals may conflict. Each company must consider this trade-off in light of its particular situation; brands that emphasize current trends will find agility relatively more valuable than brands that specialize in timeless classics.
Here's what to aim for in increasing supply chain flexibility:
· Visibility. To shorten cycle times, retailers, manufacturers, and their suppliers must all work from the same information in a cloud-based or other open system. Similarly, supply-chain managers can't wait for marketing managers to tell them how products are being received; they need accurate information in real time. In the same way that social media information “goes viral,” information about product development, order status, inventory status, and sales must also go viral, and should be disseminated instantly to everyone who needs to see it. Batch systems that delay dissemination and siloed applications that can hold inconsistent data are no longer viable in today’s world. “If everyone sees everything, that’s a real enabler for agility,” Sodero says.
· Collaboration. Simply seeing the same information isn't enough. Working collaboratively, both inside the organization and with external partners, is critical. Every partner in the supply chain should understand its role and be able to trust the other partners. For many companies, developing such relationships would represent a cultural change; Sodero says U.S. companies are further behind than European companies in achieving collaborative, long-term relationships.
· Reliability. Reducing variability in lead times can be even more important than reducing lead times themselves, according to Sodero. Consistency can be improved by using a system with visibility into the entire supply chain, rather than depending on offline messages and spreadsheets to fill in gaps.
In addition, companies should find and reward suppliers that meet their production commitments reliably, and they should seek out third-party logistics partners that can be counted on to deliver shipments on time. Contracts should include strict standards and penalties for not meeting the standards. When agility is extremely important, companies may consider bringing production closer to their consumers to avoid weather-related shipping delays or unpredictable waits at border crossings.
· Redundancy. Companies should avoid being hemmed in by inflexible contracts with suppliers, especially when it concerns rapidly Ihe ncreasing production volumes. Some companies find that keeping backup suppliers on standby is a worthwhile investment. Others prefer having fewer contracts, but ensuring that the contracts allow for ramping production up or down as needed or altered as needed.
· Omnichannel fulfillment. The ability to redirect inventory easily to different retail customers or stores, or between online and brick-and-mortar channels, is important when social media buzz affects different channels in different ways or in different time frames.
Apparel companies need to understand how social media drives fashion trends and disseminates opinions about their products and brands. To take advantage of the opportunities that social media presents — and avoid being caught short by media-driven changes — they should analyze online conversations in real time and respond as close to real time as possible. This requires a flexible supply chain, with technology, processes and culture that enable them to rapidly adapt to changing demands.