Personalization – Digital - A Practical Implementation Perspective – Part-1

Personalization – Digital - A Practical Implementation Perspective – Part-1

This Article has been broken down into 3 parts. I describe everything here from an Engineering Leaders perspective, so it gives an actual implementation guide as opposed to a marketing or a generic talk about Personalization. The latter can be found in much better articles online expressed more eloquently than my simple writing. 

As we begin to talk about personalization, lets define it so that there is no ambiguity. According to the dictionary it means “to design or tailor to meet an individual's specifications, needs, or preferences”. This is not something new, a lot of merchants have been doing it for almost two decades in a smaller sense. For e.g., when you go to a retail website, it typically says your name in the top right corner (assuming your cookies are turned on and you shopped there before), it might also bring up your account and past-orders etc. (if you have an account there and are signed in). 

As revealed during the COVID pandemic, with shut-in shoppers online shopping for everything, including things that previously needed to felt or smelled, as well as price-checking within seconds, where retail sites have to have better in order to provide a more meaningful experience that does not just replace an in-store experience. Basic digital “personalization” is not enough. Consumers want more personalized content that helps them make easier shopping decisions. Although many online shopping trips eventually lead to the 800lb gorilla of online marketplaces i.e Amazon, there is plenty of room for competition as many consumers prefer a hybrid in-store-online shopping experience, remain loyal to certain brands for intrinsic reasons, or desire a shopping experience that can’t be generated by a quick Amazon one-click purchase.

Here are some of the ways that you can personalize the Digital websites that will greatly affect your traffic and more importantly, your conversion.

·       Personalizing the home page with recommendations based on past purchases which brings in a bit of community to a solitary shopping experience. 

o   For e.g. based on your purchase, other people who bought the same item also purchased these items

o   People who browsed these items, also looked at these items

·      Showing personalized search recommendations

·      Use Data-science algorithms like Affinity models, propensity models to drive your digital email campaigns and print-media campaigns. 

o   Without marketing a personalized experience via email promotions, 80-90% of these emails end up in the spam folder. For e.g. I get a lot of generic promotions from retails for stuff I never ever buy nor am I interested in.

o   Talk about junk-mail that your mailman delivers to you every week in the form of coupons, there is a reason it’s called junk mail because 90% of it has no relevance on your household. For e.g., imagine getting promotion coupons for re-meat when your household is vegan. Am sure you relate to it, right?

The aim of personalization is to get the “Costco effect” where you go to Costco to buy a couple of things and the next thing you know when you checkout is that your bill is $200, and you bought a whole lot more. Now, Costco is a physical store and hence there is a lot of eye-candy that pops out and digs into your instant gratification retail therapy mind-set. The same can be achieved on the digital front too. Ask Amazon who without physical stores give you same Costco-effect online. I can personally attest to filling my Amazon cart with recommendations of products I had no original intention of buying, never heard of beforehand, or knew that I needed. That is the power of personalization.

Initial Steps

There are a lot of product firms like Adobe, Salesforce and whole bunch of startups who will sell you products which they claim will solve your personalization problems. Folks, before you go and sign these multi-years, multi-million dollar, please consider building the foundation yourself before embarking on these deals. Don’t be the continuing alcoholic and wait for a magic pill to solve your issues. You already have the magic pill, you just need to harness it, hope you get the point. 

BigData Engineering Team 

I am not going to go into the details of BigData here but to process the main Vs of BigData namely volume, variety, velocity, veracity, value etc. you need to have a separate well established BigData team whose main focus is to ingest, digest and produce data into the single source of truth (discussed below). The success of your program depends on the skill of this team to rapidly consume, process and store the data

Single Source of Truth

One of the first and foremost things to do even if you don’t embark on a personalization program is to consolidate your enterprise data into a single repository. Too many decades old retails firms have their data sit in regional fiefdoms like separate repositories for merchandizing, marketing, finance, eCommerce and reporting from all these different systems is a nightmare when presenting it on Monday morning to your C-level execs. The reports produced from their local data stores will all be in a different grain, will have different aggregated values as the rate at which the data flows into these repositories from downstream to final place of resting varies. 

Embarking on establishing a single source of truth for your enterprise data in the Cloud should come as an agreement between your Chief Digital/Technology Officer, Chief Merchandizing Officer, Chief Marketing Officer, Chief Finance Officer and anyone else who keep their local copies of the data. This treaty has to be hammered in place and having a common place for the data in the cloud will give everyone the same view and latency. The aggregates can be built on top of this common data and this aggregates views/cubes can be different for each of the departments depending on their needs. 

Overall, it will reduce the data, analytics, big data, data science, personalization costs by having data consolidation. 

I will talk about the further steps in Part-2 & Part-3 of this article….

Good luck in your personalization journey, if you have any questions / clarifications, please feel free to ping me at Sudhir.ganti@gmail.com