For those who aren't familiar with the above acronym, millennials — the see-now, buy-now generation — also coined the phrase “I want what I want when I want it” as something of a call to action to retailers in the early part of the last decade.
Since then, retailers have been trying to rapidly adapt to consumer shopping preferences with a series of activities to please this customer, resulting in new business processes with new acronyms and innovations such as:
While millennials demand sales and service in the IWWIWWIWI mode, many retailers — although they’re running as fast as they can to keep up — are relying on inflexible legacy technology solutions. This has led to a series of compromises, with short-term gains barely standing in the way of long-term trouble. As retailers, why can't technology solutions share the need for speed and agility? Why can't they have the solutions they want when they want them?
Retailers have typically invested less in technology solutions than other industries. Further, many solutions were implemented and cost-justified to support high-scale capabilities with reduced costs. This achievement relies on hardened, consistent business practices that trade agility for these benefits.
For some business processes, this is a good path to follow, typically in relation to transactional and execution systems. Having few purchase order formats that are processed with industry standard methods creates efficiency and lowers cost without sacrificing any critical organizational flexibilities.
Managing the customer order acquisition process, whether in-store via point of sale systems or on the web, with a single method creates useful consistency and raises conversion rates while minimizing exceptions with, again, little or no trade-off of desired capabilities.
Posting to the general ledger, paying vendors and closing books are all processes that should happen with little effort, intervention or interruption — and the organization benefits from cost efficiencies.
In general these transactional solutions tend to use large-scale, expensive software products that take a lot of time and money to implement, but once stabilized, they are changed or evolved infrequently. Once the system and processes are well embedded, the opportunities for further learning and major ROI are limited.
For a different class of solution, specifically those that support business processes that need to keep up with the consumer (typically decision support solutions or customer experience solutions), hardened, scale-based, fixed solutions will reduce a retailer’s ability to be nimble.
Keeping up with the customer requires the ability to make well-informed decisions, move rapidly through trial and error, learn and change and, in general, practice a high degree of agility. These characteristics are the antithesis of the typical retail system lifecycle of requirements gathering, selection, development, implementation and adoption — a full lifecycle that is often measured in years (certainly never weeks), even for the theoretically shorter duration best-of-breed point solutions.
New customer needs require new solutions that exhibit the following characteristics and attributes:
This class of applications requires a fundamentally different basis than traditional solutions, a basis that can be broken down into three categories.
Many organizations are, of course, limited by the scope of their own knowledge and experience. Bringing in fresh ideas and concepts can be difficult as can be supporting the change management that is so often a vital component of fostering the adoption of new methods, processes and solutions.
Data scientists are crucial for developing AI/ML solutions, but they often lack the subject matter expertise needed to truly make such a solution work for a given industry or organization. At the same time, organizations have plenty of subject matter experts, but few that know what information data scientists need to build an effective AI model.
Bridging this gap is key to developing nimble and adaptive retail solutions. That’s why it’s important to build interdisciplinary knowledge. Engineers who are great at AI and can become knowledgeable about the retail world should be paired up with retail experts who can convey their knowledge in a manner that is useful to the work of developing AI models.
With the appropriate technology platforms, solutions can be assembled and reassembled from working parts, as opposed to coded and built from the ground up. Technology platforms can include orchestration engines, microservice-based assemblies and frameworks and, of course AI/ML or RPA (robotic process automation) modules. The speed of deployment improves dramatically when the entire development lifecycle can be shortcut based on these building blocks.
Just as satisfying customer expectations today and tomorrow requires retailers to learn, adapt and progress, business solutions can't be static, working from fixed rules and assumptions. Continuous learning and improvement for the business require solutions that learn and adapt as well. The promise of AI is not just the traditional one-time gains expected upon implementation, but the ongoing improvements that such solutions can offer as they ingest new data, and tune results to ongoing business conditions.
Satisfying the IWWIWWIWI generation of shoppers’ demands means introducing a new class of solution into the retail environment, solutions that allow the retailer to implement what they need when they need it, solutions that fit their needs today and that adapt and learn for the circumstances of tomorrow.