Personalization = Data + Decisioning: 3 Data Feeds You Should Prioritize

You can’t have a meeting these days without someone throwing out the platitude “data is key”, but all too often, we find that even with the vast amounts of data being collected, business leverage of this valuable data is a problem

– whether the long queue of other IT projects, or merely confusion of where to start. In fact, in the world of making every interaction into a personal experience, data is a key component for amplifying smart AI-driven decisioning.  Unfortunately the reality is that most retailers struggle to make the data available, in the right place and format, let alone make use of it … 

… if not getting stuck trying to decide how to prioritize what data to gather.

To help, below are 3 areas you should consider prioritizing and why.

1.  Online plus Offline Data Paints The Full Picture

While online data often the easiest to use and gather for real-time personalization, for retailers with stores, it is just the start. Of course, most new customers we work with have had trouble making effective use of in-store transactional data within their online channels.  Personalization simplifies the leverage of cross-channel data and shows meaningful results – even using simple strategies that surfaces “popular product instore” decreases online purchase friction, and our customers have seen an increase of their average order values over 2% just by leveraging offline data in their online channel.  More importantly, as customers expect seamless omnichannel experiences, retailers are beginning to use merged data to leverage in-store / offline purchases (e.g., “dresses” and “shoes” categories) to effectively recommend cross category offers and content, email follow up, as well as filter out offers for products that have been returned – and to do this at scale leveraging AI that requires little of manual merchandising.

2. Leverage Back Office Data To Drive Margins

What has been clear in the last few years – retailers that have been able to get even a little of their data back office data for decisioning realise substantial benefits.  Some simple examples which yield great results, yet uses only a single data point include: 

• Margin data – You can guide customers to higher margin products with boosted search results, content / promotions, or offers / product recommendation blocks.  We certainly have customers lifting customer profitability, not to mention store profitability by combining loss leader products to attract shoppers with intelligent high propensity, high margin cross sales.  
• Brand – So often, buyers brand affinity is not leveraged.  By being able to pair or link up similar brands within search results, or suppress budget brands within content, offers, and recommendations not only drives larger basket sizes, but perceived relevance and subsequent repeat visits by your customers. 
• Regional store / inventory – Even as retailers look to leverage the power of physical stores with growing online presence, adding regional inventory data can provide personalized / regionalized offers for products that can both shape demand, minimize both out of stock disappointment, margin destroying end of season markdowns, as well as drive profitable store visits.

3. Addressing The Unstructured Data Problem To Solve Unsolved Selling Problems

A common problem we hear is poor product feed data – whether quality or depth of data.  That said, multiple approaches can be applied to help expand data to boost sales:

• Simple rules can address basic data actionability.  If your data is not always available in the format you’d like then there can be ways round this that might enable a use case deliver an all important bit of lift.  For example, if you have products that are have a common word within a product name such a furniture range like “Oxford Bedside Table”, “Oxford Wardrobe” etc, then whilst the “range” value might be available as an individual value, the ability to set up rules that work on the basis of ‘if the product name contains the word “Oxford”, then boost other products containing the word “Oxford” from the same category.
• Deep Learning AI (see our NLP data sheet) connects how your products with how people research and buy.  Data like descriptions, comments, social posts were not easily actionable are now invaluable in both helping customers find what they are looking for, as well as connecting them to their “intent” of what they want with the power of AI.  Being able to link “similar” products together based on common concepts (e.g., comfortable, adventurous, modern, plush, etc.) and begin to understand customer intent (e.g., plush / soft clothes and products good for children who like unicorns) has opened up a range of possibilities that enable say new products to be paired to older so as to help surface more of the catalogue to customers more quickly.  

Bottom line

So, while it is true that “data is the new oil”, the reality is that there are quick wins to be had from gathering small pieces of data and leveraging AI to quickly find connections to how it drives consumer behavior.