By David Pring-Mill

The following text has been excerpted from Sections 3.8—3.8.3 of the Policy2050 report “D2C and Other Digital Adaptations During COVID-19,” in order to serve as a product sample and fulfill Policy2050’s mission “to keep the most socially-relevant insights outside of any paywall.”

It’s hard to even scan a business article without seeing a mention of the word “data.” To see through hype and actually improve conversions or other business metrics, businesses may need to rethink their perspectives and approaches.

This involves distinguishing between what is good for business, and what makes a business good. It involves teasing out an accurate market view from the potential amplification of false signals.

Data is advantageous but a vicious cycle is not, and so businesses must view data with granularity.

Good for business, or a good business?

Even when changing technologies and/or trends are “good for business,” that doesn’t mean that they result in a good business.

Digital commerce and information has brought with it digital analytics, which means that nearly every type of economic activity has a real-time, sophisticated, and increasingly integrated algorithmic feedback loop.

However, not everything should have a feedback loop. Many metrics aren’t indicative of long-term success.

Inserting a feedback loop into the news media, for example, has turned a lot of journalism into junk food. People still consume it and validate it but they hate themselves for doing so and hate the product. It also becomes a less valuable product, which means that the idea of paywalling content and moving away from a purely ad-supported revenue model may become less viable as the particular media brand and culture has its identity cemented. The algorithms of aggregators play an additional role in determining what type of news spreads the farthest and the fastest.

The popularity of content, or the provocativeness of that content, or its shareability, are not equivalent to truth.

There’s obviously a lot of junk food overtaking the digital shelves on which our minds browse. Brands must carefully tailor their content to the audiences that matter, and strive for virality without compromising their defining values, or obfuscating the value that they provide.

Accuracy, or amplification?

Executives are understandably motivated to implement, monitor, and respond to feedback loops, as they can be used to optimize their operations, but at times, this is also distorting operations because they’re not actually responding to a fair representation of data.

They’re responding to the amplification of data through algorithms, to editorial selection dictated by commercial imperatives, and to especially vocal subgroups that might even be using botnets.

Even if a business can see through these funhouse mirrors, the public may not. And that is still a problem for businesses because they’re commercially incentivized to stay in-sync with consumers, even when they’re missing a bigger picture.

A lot of this polarization, distortion, so-called “cancel culture,” and tech-lash has happened within a sociopolitical context. To provide a little respite from politics, let’s look at an example that is strictly business:

The “Digital Adaptations for Retailers” section of this Policy2050 report observes that L.L.Bean had to modify its overly generous return policy, given the costly and accelerating nature of the abuse of that policy.

The Maine-based retailer’s internal surveys indicated that 85% of their customers didn’t have a problem with this policy modification.

However, on social media, some of the critics seemed to be very loud and dramatic, proclaiming that they would never buy from the brand again. It’s suspect if they ever did buy from the brand in the first place; even if they did, some of that “buying” may have resulted in a long-term loss when they brought the well-worn items back for a refund or exchange. Social media chatter may have even spread awareness of the original, vulnerable policy.

If the brand had listened to the feedback on social media, instead of their surveys, they might have rewarded people who were never really their customers, at the expense of their most loyal customers.

D2C brands must ignore feedback loops when they spit out distortions. This requires both attention to detail and discipline.

Business interpretations and technical approaches

Within this report’s “D2C & Amazon Factors” section, the “Optimizing Discoverability within Amazon” subsection outlines the importance of clearly and efficiently signalling the nature of a product to buyers and the platform’s algorithm.

Product listing titles, descriptions, and imagery should also be based on an understanding of the customer base — what appeals to them, what language they use in their search terms, how factors rank in their purchase decisions.

A failure to optimize accordingly could falsely lead a D2C brand into the conclusion that they haven’t found product/market fit, when they simply haven’t represented their product in a relevant or ideal way. Rewording a listing is always cheaper than remaking a product!

A well-optimized listing for a quality product, on the other hand, will likely create a virtuous cycle, with positive reviews leading to more sales, more reviews, and more refinement.

Very few products will be perfect out of the gate. But if you begin with a product of reasonable or good quality, and a very intentional, accurate, consumer needs-focused presentation of that product, then any data generated is more likely to be advantageous.

Many technological evolutions are obvious, especially in hindsight, but even with the benefit of hindsight, some of the underlying or resulting changes in consumer perceptions and behaviors have been subtler and harder to perceive. History often smooths over consumers’ initial perceptions and resistance.

D2C expert Alex Murray has commented that the feedback loop currently available to retailers is immediate and public. It’s easy to forget that there was once a question of whether someone does or doesn’t shop online; now everyone does. It’s hard also to imagine that frustrated consumers once penned letters of complaint to brands and retailers.

Murray said, “Retailers are really having to up their game in terms of how they connect with and how they talk to their customers. It’s a very different world and not all businesses have adapted so well to that.”

Consumers have come to expect key features from ecommerce platforms, such as reviews. Online retailers and D2C brands must meet these baseline expectations and can sometimes go wildly beyond that standard by becoming early adopters of the latest tech. However, as Alex Murray remarked, “I’ve still seen so many over-engineered technical solutions that you wonder [about] the cost versus the benefit.”

Looking ahead at the technical challenges for D2C over the next 2 to 5 years, Alex Murray expressed the following ideas in his D2C Leaders interview:

• Build the capability to iterate quickly. To stay relevant: iterate, test, update tools as needed.
• Question the value of technological investment plans that stretch over longer timeframes, because the rate of change is too fast.
• Buying SaaS might serve you better than buying a platform, locking in that investment, and having it depreciate, while lacking the flexibility to swap and modify aspects of the system.
• Choose wisely but don’t dither. Through his experiences in this space, he has noticed that “just the sheer volume of choice, for most retailers, is utterly bewildering.” There are thousands of indistinguishable digital marketing tools but “dithering is deadly.”
• Make data actionable. Data is involved in all of the aforementioned points but it’s only valuable when it’s actionable. Murray said, “I much prefer to see clearer, smaller datasets used well than huge datasets gathered and then a lot of speculation and hypotheses thrown over the top of that data.” Each hypothesis should have an execution at the end of it so that the business isn’t wasting time in order to essentially say “hmm, that was interesting information” and go no further.

Even when data has been well-integrated and turned into a success story, that story could be distorted by a focus on the wrong metrics or a lack of appreciation for the fuller context.

For example, Murray explained that Tesco did an amazing job of pioneering the online grocery market in the late 90s and early 2000s. He said, “They’re getting a little flat now because they’re losing market share, but the reason they’re losing market share is actually because they took so much market share in those years, and actually, the other retailers are catching up.”

Incidentally, Tesco is now involved in the pioneering of “unmanned stores” or AI-powered retail, which could conceivably improve their market share, their margins, and their ability to leverage customer data.

The tech industry’s view on data, broadly speaking, is that it can be useful when optimizing multiple areas of the business and creating real-time, personalized experiences based on hyper-targeting. But there’s actually an optimal way to use data to optimize.

The other, unavoidable variable here exists in the form of data regulations, which could conceivably affect certain tech stacks.

Many mar-tech vendors show confidence to their investors and customers by asserting that their specific offerings will be resilient in the face of increased regulations. However, they might secretly fear the loss of their market position and the disruption of proprietary methodologies that require access to particular datasets in a particular way. This might limit the usefulness of some of their public responses when assessing the situation.

Resistance to new regulations and behind-the-scenes negotiations signal the utility and high profitability of data usage, as well as the fear of increased costs associated with compliance.

The full report “D2C and Other Digital Adaptations During COVID-19” is now available for purchase on Policy2050.com.