Data Analytics for Customer Experience

Scott Duthie - Principal Consultant

Customer Experience is hot. An online shopping boom spurred on by Covid, combined with a fiercely competitive marketing and selling landscape, has retailers scrambling to understand their customers and plot their next move.

Helping retailers keep their finger on the pulse are specialized customer experience agencies that deploy surveys to customers and analyze their movements and retail habits. They extract insights that aim to improve operations, win the next customer, and boost loyalty.

Competition is heating up in this space and retailers are demanding more advanced analytics than ever before – AI driven churn modelling, multi-variant analysis, and prescriptive analytics have become the norm. Big players like Qualtrics and Medallia make a lot of noise, but smaller vendors like SMG are carving out a significant slice of the market with their powerful customizations, and a niche offering that caters to individual clients – after all, a customer lifetime value model or churn risk score means very different things to a Home Depot vs a Burger King vs a Gas Station.

Helping accelerate these analytic offerings is Pomerol. We partner with customer experience agencies to build specialized apps that complement the agency’s core solution – for example, a customer churn score predictor that uses a live trickle of customer surveys and real time geo-location pings to flag at-risk customers. Our apps sit around the periphery of the agency’s core platform without being fully integrated. This approach allows us spin up and deploy laser focused analytic solutions in minimal time. It allows customer experience agencies to move fast, respond to industry needs immediately, and cater to the individual needs of their clients without the usual lead time.

Two great opportunities exist for Customer Experience agencies:

Data integration: Integrate multiple feedback sources to help retailers navigate the road back to normality in a post-Covid world.

Digital channel focus: Create analytic tools targeted on the feedback and interactions through the digital channel, with an emphasis on text analytics and unstructured data. Think online and mobile ordering, BOPIS (buy online – pickup instore), delivery, and post purchase customer support.

Opportunity 1

Integrate multiple feedback sources – help retailers navigate the road back to post-Covid normality with a holistic view of the customer.

“CFM customers should look for providers that integrate multiple feedback sources to create a holistic view of the customer.”
Forrester Wave Q2 2021

The digital channel presents an abundance of new data sources – customers are interacting by mobile app and website, dissatisfied customers often opt to complain in the public domain via social media channels, chat bots are now the front line of customer support, and digital ordering and fulfillment requires and whole new set of logistic systems.

Naturally, the most common request we hear is “we want to see all our data in one place”. Customer Experience agencies that can crack this will rise to the top, as the analytics enabled by merging this sea of data sets is unparalleled.


With the customer story being dispersed across multiple data sets, it can be hard to build the linkages required to follow the customer journey. This is made more challenging by a collage of systems and platforms that were never intended to work together, not to mention the challenges of unstructured or anonymous data often provided through social media channels.

Typically, customer experience agencies offer a web platform in which retailers can view survey results and visit metrics of their own customers, however the platform model forces all users into the same mold, making it hard to cater to the unique needs of certain industries or specific customers. The result is often a ‘one size fits all’ solution catering to everyone and no one at the same time.

To give further context to the customer experience, retailers need to see transaction data alongside their survey results, however this requires one-off development and does not fit well with the platform approach.

“We have so many customer touchpoints, building a holistic view of the customer is our holy grail. The same customer that visited the store also researched online in advance, they might have also called our support center and posted on our twitter feed. These different data sets bring context, and this is where the customer story is found.”
US big box retailer


Pomerol is solving these challenges with a multi-pronged approach. First, we rapidly build the data engineering that stitches the different feedback sources together in the background. We do this in weeks, not months, and the result is a single thread for each unique customer containing all their touchpoints.

Next, we create a flexible front end that enables a mix of prescriptive insights and self-service discovery by the end user. This approach surfaces key insights for management users, while giving analysts and researchers the power to dig deeper. The user interface does not replace the CFM/CEM vendor’s core platform, it sits to the side and compliments it, dealing with all the new data sources and edge cases that appear.

And finally, we iterate at pace. Retail environments are a roller coaster of micro events. Huge value lies in seeing what happened yesterday or analyzing a promotion as it happens. We process feedback and new analytic requests at the pace of business.

Opportunity 2

Piece together the digital story – with an emphasis on text analytics and extracting a narrative from unstructured data.

Pre-Covid, never has the retail or QSR space undergone such transformational change in how they interact with their customers in such a short amount of time. Traditional bricks and mortar stores have created online shopping sites overnight, and those already operating in the online space have seen wild surges in demand. In normal times, retailers would ease customers into this type of transformation, piloting different approaches in different stores and smoothing out the wrinkles as they go. Without the luxury of time, these recent changes have been rushed through without the chance to consult customers and hone the new offering.


Digital touchpoints are very fragmented. Picture a customer placing an order through an indirect partner channel, that then gets fulfilled by your team, but delivered by yet another partner organization. The customer then contacts support with a complaint – however they’re really interacting with a chat-bot and automated system. We then deploy a survey to the customer hoping to get a response, and then we rely on a web off third party data to tell us if the customer has returned to one of our physical locations – whether they returned as a web customer we often have no idea. So how do we harmonize all these data points and paint the true customer journey? And when log customer service interactions by phone, or ingest comments left on social media and surveys how do we extract the true story?


The first challenge we tackle is identifying and bringing together all the digital channel sources. Where possible we connect data at the individual customer level, but where data is anonymous, we use proxies like location and demographic profiles.

Pomerol leverages both third party and homegrown tools for text mining and analytics. Third party tools allow scoring of huge data sets via API without the need to stand up infrastructure or build complex algorithms, while homegrown text analytic models help refine the results and add a touch of unique customer context. For example, a burger chain will want to hunt for specific product names or include slang and commonly used jargon in their searches.

Ingesting and analyzing text data in real time lets us produce ‘emerging insights’ reports where we identify key themes that are on the rise or decline right now. This empowers clients to act before it’s too late. For example, during an unseasonal patch of hot weather an outdoors equipment retailer sees customers are complaining about many out-of-stock beach items like umbrellas, spike ball and coolers – they identify this trend and the impacted stores and move surplus inventory to these stores overnight to remediate.

Our Approach

First and foremost, we are all about value. Business Intelligence solutions do not have to cost the earth. We deploy the same ethos to building analytics as Credit Unions do in their daily operations – focus on what the customer needs and exceed their expectations.

Our team of consultants and data engineers have a depth of experience across the full analytics supply chain – from initial data extract, through transformation and modelling, to dashboarding and operational alerting.

Together with your team we identify the greatest opportunities and use our industry experience and proven technology stack to deliver solutions that will differentiate you from your industry peers.