New competitors and new channels mean a new mandate for the consumer packaged goods industry: Engage more directly with consumers. These views from industry and academia lay out the opportunities and specific tactics enabled by digital technology.
An executive perspective by Sudhakar Gudala, Vice President and Global Head, CPG and Distribution, TCS
The consumer packaged goods (CPG) industry is undergoing a dramatic shift to compete in a marketplace where technology-savvy consumers have ever more choices in how they buy products and interact with brands. CPG companies can no longer rely exclusively on retailers for sales or to mediate feedback from the end consumers. Instead, brands must put consumers at the heart of their business and take a direct-to-consumer (D2C) approach across the life span of the consumer relationship. That approach may include personalized marketing and digital engagement, real-world brand experience stores, online sales channels, and consumer involvement in the cocreation of new products. Adapting to the emerging D2C channel can help CPG companies reaffirm brand loyalty and gain competitive advantage.
E-commerce is expected to account for 10% of overall CPG sales by 2022. To reach consumers directly, CPG companies are forecasted to spend $11 billion on digital advertising in the U.S. this year.
The D2C strategy requires a strong foundation in technology to gather and leverage consumer data, as well as a willingness to rethink organizational priorities, IT investments, and skills. Processes such as supply chains must be redesigned to deliver on consumer demand across any channel. Companies in regulated segments such as alcoholic beverages must plan how to provide customized experiences within a compliance framework, while those establishing their own D2C sales must consider how to avoid channel conflict with retailers.
The shift requires significant effort and commitment, but CPG companies that delay adopting a D2C strategy risk losing ground to not only new market entrants but also traditional rivals that are willing to make the change. The better consumer insights afforded by D2C will allow these competitors to outflank the laggards with more innovative new products, hyper-personalized offers, and better demand forecasting.
Meanwhile, those who do not adopt D2C will be less able to provide immersive consumer experiences, have limited control of their revenue streams, and lack the ability to counter disruptive innovation with new business models and sales channels.
D2C Starts With Consumer Data
Direct relationships with consumers provide direct access to data, which fuels the advanced analytics that drive D2C benefits. This data is the top prerequisite to tapping artificial intelligence/machine learning capabilities that can personalize every consumer interaction, at scale and in real time.
By gaining insights into demand patterns and unique needs from customer order data on its e-commerce platform, a global CPG company developed a strategy for prepackaging goods. That resulted in a significant increase in packaging productivity and speedier order delivery.
Brands taking a D2C approach have used advanced analytics to optimize pricing, promotion, and demand forecasting. For example, one large CPG company sought to deliver a hassle-free customer experience during its peak season. By using a digital command center to proactively monitor and plan for demand through its D2C channel, it was able to fulfill orders faster, improve satisfaction, and realize more revenue.
CPG companies are also using consumer data with AI to accelerate the purchase cycle. One such brand that operated mainly in traditional retail channels wanted to also provide an immersive experience to consumers. With more and better data from D2C and predictive machine learning, the brand could customize content and offer a more personalized experience. This approach increased online sales by 5% and drove a 15% increase in website traffic.
Since consumer data underlies the success of a broad D2C initiative, all consumer engagement that generates data — and hence, insights — must be undertaken with security and privacy as a paramount concern. CPG companies that have not historically collected personal data must understand and implement regulations and best practices for data collection, storage, and use.
Make New Channels But Keep the Old
D2C does not necessarily equate to selling products through a brand’s own e-commerce portal. Rather, it is about adding more touchpoints to the consumer relationship and developing a broader footprint, such as brand-owned brick-and-mortar stores, social media engagement, and mobile apps.
While direct sales channels provide more options to consumers as well as more immediate access to data, CPG companies selling direct need to devise an appropriate strategy to avoid channel conflict. One such approach is to adopt exclusivity as a strategy. For example, a CPG company that sells via multiple partner e-commerce channels offers personalized products and services only through its brand website. Bundling products and product giveaways can also help brands avoid conflicts. Another scenario under which channel conflict can be managed is for the brand to share some of the data generated from its direct sales with its retail partners to help them better target consumers, a win-win for both parties.
Building the Organization to Deliver D2C
A successful D2C initiative must start with full commitment from top leaders; since it is fundamentally a digital strategy, the chief digital officer (CDO) is often best positioned to orchestrate its implementation. The CDO is pivotal in creating a unified digital collaboration platform to connect stakeholders across the company. He or she must work in partnership with business heads to align digital capabilities to strategic priorities and act as a transformation agent for digital innovation.
The top priorities to operationalize D2C are to build up a robust data analytics team and ensure that the organizational processes are in place to enable open communication with marketing, product development, supply chain, and logistics. Collaboration between analytics experts and business domain experts helps keep the company’s strategic priorities aligned with D2C goals and builds a culture of data-driven decision-making. It is imperative that the closer engagement with consumers that D2C enables also feeds new product development. The organization should develop ways to gather and analyze real-time consumer feedback as it conceptualizes new products and services.
By placing the consumer at the center of all activities, a D2C strategy creates a positive feedback loop of sustainable value for both CPG companies and their customers. Brands gain increasingly better insight into the experiences that consumers want and can activate more channels through which to control and deliver those experiences, while consumers benefit from more-targeted, useful interactions and products that better meet their needs. The closer relationship can foster increased loyalty in a marketplace where consumers face a dizzying array of choices and help CPG companies maintain their edge against a growing array of competitors.
Mining Online Content for Customer Needs, With Help From Machine Learning
A scholar perspective by John R. Hauser, Kirin Professor of Marketing at the MIT Sloan School of Management, and Artem Timoshenko, Ph.D. 2019, MIT Sloan School of Management
Insights into customer needs are critical to help companies spot new product opportunities and improve new product designs, existing products, and services. Today, consumers are creating a wealth of new content that speaks to their product needs as they search for, buy, review, and chat about purchases online. But is this trove of data a practical source of information for CPG product innovators, and is it as valuable as traditional focus groups and experiential interviews?
In our recent research1 at the MIT Sloan School of Management, we found that, yes, by using machine learning techniques along with human analysts, important customer needs can be found efficiently and cost-effectively in online user-generated content (UGC).
There are really two questions we asked: What is the value of information surfaced from UGC content, and can a machine augment a human to make this process fast and efficient?
To investigate the first question, we worked with professional consultants who are experts in identifying customer needs from experiential interviews and had them review both randomly selected interview transcripts and information from user-generated online product reviews. While there may be an expectation that online reviews are biased due to self-selection — that people only bother to write a review if they are very pleased or very unhappy — we found that the UGC contained almost all the needs related to the particular product that we studied. We also found some needs in UGC that didn’t surface in interviews.
The second question is whether machine learning can successfully augment a human being in the task of sifting through high volumes of content to extract meaningful information for review. The answer there is yes as well. Strong algorithms can identify UGC content that’s rich in information about customer needs, as well as UGC content that’s redundant, so that humans don’t have to look at all of it. This not only saves many expensive hours of professional readers’ time but also is far more practical, and faster, than manually reviewing thousands of reviews, and potentially tweets and other social media posts. Machine learning augmentation is particularly valuable as the frequency with which a need is talked about is not very highly correlated with how important it is, so it is easy to miss significant but rarely mentioned customer needs.
It’s important to note that machine learning doesn’t fully replace humans in the process. There are pieces that machines are particularly good at, such as content selection, but at least today, humans are better than machines at understanding what customers want and relating to the customer experience. And so in our approach, we rely on the human analysts to formulate the precise customer needs that people have expressed.
Although our research used online product reviews, a similar approach can be taken to extract customer needs from social media posts and tweets. We may find that different kinds of customer needs are expressed in different types of user-generated content: You may be chattier with friends on Twitter than on an e-commerce site. In fact, in one of our studies, a company combined social media data and online reviews and found that the two sources were complementary in terms of what kind of customer needs they could identify. And given the efficiency gains we found using a machine-learning-augmented approach with online reviews, we would expect even greater gains with social media data, as it contains far more redundant and irrelevant content that can be automatically filtered.
What we were able to achieve with current machine learning and natural-language-processing technology is already quite effective, but we’re confident that what we’ll have five years from now will be absolutely amazing. At this point, every company in the CPG area should be considering this approach to surfacing customer-needs information from the content that consumers are creating online.