Retail, footwear and apparel (RFA) leaders have long invested in advanced analytics initiatives. Historically, most investments have focused on end-market or customer-facing applications, like inventory optimization or real-time tracking of product reviews and social media sentiment. But applications for advanced analytics and artificial intelligence abound across the digital value chain, from the earliest product discovery conversations, to design and product creation activities, to vendor management and manufacturing.
RFA product leaders are showing increasing interest in advanced analytics and understand the benefits to financial metrics like topline revenue and margin, as well as impacts on retail-specific indicators like product returns and design adoption. A 2020 Kalypso Digital Adoption survey revealed that 81% of product leaders see advanced analytics as a critical investment for 2021.
So Where is Everyone?
Despite overwhelming excitement for the application of advanced analytics in product discovery, creation and manufacturing, few product leaders are witnessing that vision come to life. According to an August 2020 analytics report by McKinsey, retailers expect to invest over $5 billion in advanced analytics in 2021. We have found that they are dedicating only 15% of investments to product analytics, with most funding still earmarked for end-market activities like e-commerce and channel analytics.
Product leaders have valid concerns about going “all in” on advanced analytics. They have witnessed other enterprise analytics initiatives fail to achieve impact, despite the significant financial cost and upfront promises. They worry initiatives with multi-year payback horizons will lose steam, or that leadership turnover will result in divestiture before projects deliver their intended outcomes. Often, they simply struggle to gain the trust of business partners, who associate mentions of artificial intelligence and automation with significant role changes or inevitable layoffs.
Pitfalls & Misconceptions
Advanced analytics initiatives do not have to be a high-risk, multi-year struggle. In our work with dozens of retail, footwear and apparel organizations, we have identified the eight most common missteps of advanced analytics and the shift in mindset, strategy and approach required to overcome them.
1. Bringing a Solution in Search of a Problem
Many initiatives begin with a vendor’s persuasive sales pitch or an internal analytics team eager to deliver a tool or platform, rather than a frank discussion about the challenges facing the business. This leads to low-impact initiatives that fail to excite the business and compromise the likelihood of approving future projects. Instead, source real challenges from business partners and determine a solution only after identifying a high-priority problem.
2. Treating All Opportunities Equally
Business partners face a plethora of challenges that could be addressed by a backlog of analytics initiatives, and teams disagree on which challenges are most urgent to solve. Product analytics leaders asked to mediate these disagreements often face pressure to prioritize the most influential business partners, rather than the most impactful issues, resulting in disengagement from overlooked business partners and investments with minimal impact. Instead, rank opportunities by impact and complexity; start with the least complex projects with the greatest impact on the business.
3. Getting Lost in Translation
Business partners and data scientists speak different languages. Data scientists use terminology unfamiliar to business leaders, while business leaders make data requests without understanding data science principles and limitations, leading to poorly scoped projects that fail to meet business objectives. Instead, identify “analytics translators” – business-savvy team members with technical know-how who can bridge the gap.
4. Making Investments with a Multi-Year Payback
Teams energized about the possibilities of advanced analytics often their sights on the most challenging or complex problems to solve. Often these initiatives take years to pay off, resulting in projects that are stalled or defunded after leadership turnover or budget challenges. Instead, initially focus on a set of small, practical use cases that generate near-term business value and help create a sustainable, self-funding analytics program.
5. Replacing Human Intuition with Machines
Activities like product design or assortment creation will always require the experience and intuition and intuition of talented teams. Some business partners fear that advanced analytics initiatives will automate necessary functions or replace their jobs entirely, which leads to a lack of funding from key partners like design or merchandising, or poor adoption of solutions that are approved and implemented. Instead, focus on solutions that automate their most frustrated or disliked tasks or that provide them with transformative new insights that augment their capabilities.
6. Stopping at Insights Without Integration
From PLM systems to POS systems, retailers have no shortage of product data. Organizations are increasingly growing their analytics functions to make sense of this overwhelming amount of information, but the role of an analytics group typically ends when insights are handed off to the business, who often lack the time, resources, or tools to act. As a result, few business partners can trace their team’s success to analytics teams, leading senior leadership to question continued investment. Instead, engage business partners from the beginning and include integration activities (i.e., process design, decision governance, change management) in project plans.
7. Relying on “Black Box” Models
Analytics teams can mistakenly believe it’s easier to spare business partners the details of advanced analytics initiatives due to technical complexity, choosing to communicate at the beginning of the project when they require data, and at the end of a project when they have insights to share. As a result, analytics programs take on an air of mystery, generating skepticism and distrust from business partners and resulting in solutions with poor adoption. Instead, adopt a mantra of radical transparency, communicating openly with stakeholders about the research approach and potential outcomes. Leverage analytics translators to help business partners decipher technical language.
8. Mistaking Business Intelligence for AI
Most organizations have invested in business intelligence functions, which usually take the form of dashboarding or reporting tools. These tools may describe or diagnose business challenges, but they fail to predict outcomes or prescribe actions, the defining characteristic of artificial intelligence. Senior leaders without a deep understanding of data science may decline to invest in additional advanced analytics capabilities, ultimately threatening their organization’s ability to remain competitive among other retailers. Help leaders recognize AI as a step-change in analytics capabilities and invest in up-leveling your people and technology accordingly.
More Important than Ever
Navigating the analytics minefield is more critical than ever for retail, footwear, and apparel organizations. As margins tighten and digitally native competitors double down in their data investments, product leaders cannot afford to deprioritize advanced analytics initiatives or delay investment for another year.
While these pitfalls and misconceptions are numerous, they share common themes of distrust among teams and questionable returns on investment. By starting small, focusing on real problems, bridging communication gaps, and collaborating with business partners on implementation, product leaders can finally translate their intention into action, pursuing the analytics use cases with significant, long-term impacts on their organizations.
Drew is a principal in the Digital Strategy & Operations practice, bringing subject matter expertise & practical experience in leading global transformation programs across the retail, consumer and life science industries.