“Nordstrom kicks its omnichannel strategy to the curb”
– George Anderson, RetailWire
Few topics are as hyped up as omnichannel in Retail – or as difficult to execute successfully. Part of the problem depends on where you sit within the organization. Marketers view this as a consumer issue, in-store this is a merchandising and fulfillment issue, online this is a website design and e-commerce execution issue, and of course there is also a supply chain and fulfillment angle. Next is your IT department that views this as a complex data integration issue with huge ETL implications and a multimillion-dollar price tag to get started. More often than not your retail omnichannel strategy ultimately gets kicked to the curb! However, as indicated above, Nordstrom is doing things differently and is currently testing a store redesign that would make in-store pick up of merchandise easier to manage for consumers and the retailer.
Having a consistent view of and delivering a seamless experience to your customers across all touchpoints in-store, online, via kiosks, mobile, digital and social media is the original premise of omnichannel and it has become an even more important strategy in today’s competitive market. Consumers want to be treated as individuals and responded to consistently – regardless of channel. And, given how easy it is to compare pricing and fulfillment options online, lack of an omnichannel strategy would increase the chance that consumers may abandon their shopping carts mid-transaction and kick your brand to the curb.
So why does Omnichannel Fail? Three Major Challenges Delivering on an omnichannel strategy requires linking consumer data with product and supply data, and delivering on a live transaction regardless of channel. You need a 360 view of the consumer, product and supply as well as the ability to manage live transactions that utilize heterogeneous data sources.
Let’s examine why this is so difficult:
Consumer 360: Who is Jen Dough? Simply put, most retailers don’t know who their consumers are! Consumer loyalty data for the retail store may sit in one database, while online transaction data and calls into the contact center are kept in numerous others. Each of these three data types has a specific format. To illustrate another case-in-point, consider what happens when a Convenience Store retailer’s loyalty program is not linked up to its pharmacy program nor patient visits to its in-store health clinic. If you don’t know who Jen Dough is – it’s really hard to understand why she buys, how frequently, what her preferences are, and what promotions will engage her. And you certainly aren’t going to be able to expertly advise her on nutritional information or potential drug interactions. Unfortunately for most retailers, Jen expects and requires a more consistent and meaningful interaction with you or she’ll abandon your brand for another.
Product 360:Can your consumers find what they’re looking for? Consumers really don’t buy products – they buy solutions. This may include recipes for dinner or a home entertainment system bought online, in-store or via mobile device. A product is a complex mix that includes the physical product and accessories, digital images, videos, recommendations and ratings, nutritional information or specs, locations, pricing and promotions – all of which need to be linked and communicated to the consumer. Consumers would also like to be able to compare and make choices across products. Whether the choice is between steaks or lasagna for dinner or an Ultra HD and Plasma TV set with the right service plan – the consumer’s path to purchase is individual and complex as they might research online, compare prices via mobile device, and finally purchase in-store. Regrettably, retailers are still stuck organizing category data around “products” i.e., an online search for wireless sound systems may yield 1,300 results that are not linked logically to enable better consumer purchasing decisions.
Supply 360: Making sure you’re operational: Or simply, knowing where the product is in the supply chain and fulfilling the transaction seamlessly. As this is retail, this process needs to be fully “Live” or “Operational.” If you can’t check availability, update inventory, price dynamically and suggest stocked, pick-up stores within driving distance from the consumer – you won’t succeed with omnichannel. Plus, there is also the issue that stores were not designed as distribution centers (DC) and you may want to consider redesigning yours to enable curbside pick-up just like Nordstrom is doing.
To make omnichannel work, you must be able to link consumer data with product and supply data in real-time to enable operational transactions.
Unfortunately today’s retail systems and databases just weren’t built for this challenge. Most retail database systems run on rigid, inflexible relational database (RDBMS) models first developed in 1970. Data stored in RDBMS’s is framed into a rigid schemas consisting of rows and columns prior to analysis. It’s really difficult to make changes to these schemas so RDBMS does not work well for the unstructured and constantly changing heterogeneous data associated with omnichannel. Next are the challenges inherent to complex data integration across multiple legacy systems – many of which are the multimillion-dollar dream projects for systems integrators and ETL software vendors. Finally, we’re looking at making this operational or live if it needs to work.
So, how can a retailer achieve success at omnichannel?
NoSQL Represents a Revolutionary Solution for Omnichannel Clearly traditional mainframe or RDBMS’s lack the flexibility and scalability to handle the data volume, velocity, and variability issues inherent to omnichannel retailing.
Trying to combine the vast scale of structured and un-structured consumer, product, and supply data for analysis into a relational database designed for structured “rows and columns” and then using SQL (Structured Query Language) to query it – just does not work. Part of the problem here results from the dominant position incumbent vendors like Oracle and IBM hold in Retail IT departments making a radical shift difficult even when the business value proposition is clear.
NoSQL (Not Only Structured Query Language) technology represents a transformational change in perspective. Instead of getting the schema just right before doing anything else, NoSQL advocates loading up the data first and then seeing where the problems lie. This problem-oriented approach focuses on how the data will be used (queried) rather than how the data must be structured to fit within a traditional RDBMS.
For omnichannel retail, the shift to NoSQL means you would not have to spend a year trying to figure out the right data model and perfect schema to analyze and store data on consumers, products, and supply. Instead, you can load the data, have it indexed automatically, and then search and query it for emerging trends and signals.
The retail industry has surprisingly lagged behind other industries like media, financial services, and even government when it comes to using NoSQL to solve the complex operational problems associated with Big Data. But with the growing importance of omnichannel it may finally be the retail industry’s chance to get it right.
Please be sure to read my other articles in this series: