Proactive customer experience: How a CDP can help end bad CX

All the energy for the first 10 years of the customer data platform era was focused on marketing and advertising. Deep profile data and precise audience segmentation led to better performance in social ads, email campaigns, and customer journeys. Lately, we’ve seen CDPs also start to drive more connectivity between CRM applications for better customer experience. Now, we’re beginning to see the next phase of customer data management: Creating better, proactive customer experience by preventing bad experiences.

Bad customer experience: Why CX is getting worse and how brands can reverse the trend

Today, smart brands are unearthing interesting signals from customer data and applying them for various use cases. For example, why market to customers that have an unresolved issue in the call center? They probably don’t want to be sold anything new until their issues are resolved. By unifying marketing and service data, the CDP provides a way to execute this use case without making it a costly IT endeavor.

This is all great stuff, and exactly what customer experience managers should be doing. Leveraging loyalty data to drive e-mail campaign content (“one more flight gets you to platinum”), leveraging purchase data to drive next-best offers (“customers who bought this, loved this”), and using survey and preference data to personalize messages (“you’re seeing this, because you viewed that”). CDPs should be the glue that connects customer experience together, and it’s encouraging to see the technology finally start to cross the chasm from marketing to broader use cases. That said, perhaps we’re focusing on the wrong things. While these use cases are extremely valuable and certainly revenue generating if deployed correctly, maybe there’s an even more noble use for CDP?

Redefining customer identity for the cookie-less future

I’ve written about the concept of “data exhaust” – the notion that digital consumers continually give off thousands of signals as they move through the world. Advertising clicks, video views, interactions with IoT-enabled devices, pin pad transactions, GPS-guided car travel, and restaurant check-ins are just a fraction of the digital exhaust we emit daily. Collected at scale, and unified to a single people-based identity, they offer a rich snapshot of customers.

These data provide an analytical view of customers that’s also entirely contextual – the signals I give off on a Monday morning (commuting, consuming business content, sitting in an office, eating lunch salads) may entirely differ from my Friday night persona (fine dining, searching for live music, located near a beach). While much of the customer profile can remain static and change infrequently (home address, job, income, number of children in the household), many of our attributes change minute-by-minute, or day by day (location, time of day, and weather). Brands need to consider how they balance their customer view and leverage CDPs to help bring behavioral data into their systems, balancing the long term “truth” of their profiles with intent signals that offer needed context. This is powerful stuff and is key to unlocking better performance across sales, service, commerce and (of course) marketing and advertising. But, if you look at what we’ve just experienced during the pandemic, it doesn’t get us to completely proactive customer experience.

Real customers, engaged with your brand: Enter data authentication

Say you ran a popular truck rental service, and business was booming as people fled shuttered cities for remote working paradises in the suburbs and beyond. You set up effective marketing campaigns to likely movers, gave them a simple way to plan their move and book a truck. You offered them the ability to pre-pay, submit their insurance online, and add extras like boxes and moving blankets. You delivered an ideal customer journey from inception through purchase, but when they showed up for their pre-scheduled time…the truck was not there. (Cue sad trombone sound). All the data in the world didn’t prevent turning a delighted customer into a glaring red account with a negative NPS score. But what if it could?

What if all the rich “data exhaust” we collect and unify in the CDP profile could be sent into the key backend systems that mange the truck inventory, maintenance, and physical location of the fleet? This is known as the ERP system. As I noted in a previous column, commerce failures during the pandemic had nothing to do with a broken cart or check-out processes, but rather shortages due to the collapse of supply chains.

2022 commerce trends show CX needs a reboot

In our truck rental scenario above, the CDP would invisibly enrich customer experience – not by driving more demand and personalization, but by alerting the inventory management system via demand signals and throttling new bookings until supply issues ceased.Preventing a bad experience before it happens is deep, proactive customer experience – and one of the first steps on the way to enterprise CDP.Customer data pointed at the core systems that drive manufacturing and supply chain management is just one simple use case. There are hundreds more ways in which this powerful data can add intelligence and automation to backend of business processes. It’s an exciting new era!


Data Management is the Backbone of Enterprise Agility in CX

Moving from next-best-action to next-best-dollar requires richer data, actionable intelligence, and pervasive automation

As Heraclitus reminds us, “No man ever steps in the same river twice, for it’s not the same river and he’s not the same man.” This is arguably even more true since Heraclitus uttered the phrase, given the rapid and abrupt changes we have seen in the world lately and their impact on global business. The global pandemic humbled many retailers that were slow to adopt to digital, as digital-only interactions grew to 72% in a short period of time. Suddenly, “buy online, pick up in store” wasn’t just a nice-to-have convenience feature for customers – it was a requirement for doing business in the height of the pandemic.

Did anyone predict the how rapid and dramatic the rise in global inflation would be? Businesses across the world had to radically alter their pricing and delivery models; as input costs rose, global supply chains tightened – all while consumers were tightening their own belts due to the corresponding rise in prices.

The framework for managing through this is broadly called “enterprise agility,” or an organization’s ability to quickly adapt to market changes. Do you work in an agile business? If you are not sure, think of how well your company would do if your cost of goods surged by 25% in a single month, or how you would react if one of your key markets closed overnight due to a geopolitical conflict, or if the government’s response to a pandemic shut down your business. Of course, all of this has happened recently, and many businesses discovered just how agile they really are.

Amid rapid change, marketing always seems to lead the way. Whether it’s figuring out how to message a price increase to customers, changing a product roll out based on regional market changes, or reacting to the sudden unavailability of a product based on a supply chain shortage, companies need to account for how their customers will react to change, and manage customer experience delivery as appropriate. Product, Sales, and Marketing are the three-legged stool in most organizations – and only one can move fast enough to react immediately to crisis. Changing and introducing new products is time consuming. Changing a sales organization that has been trained, hired, and has set annual targets can’t happen overnight. What can change? Marketing budgets, campaigns, website messaging, webinar content, search keywords, etc. Marketing is the only part of the organization that can turn on a dime.

When times change, and the CEO picks up the bat phone, he’s calling the CMO first. So, what does the agile marketing organization require?

Let’s take an example: A popular outdoor retailer runs a promotion for a new hiking shoe that is a “collab” with a trendy brand — and it goes viral. Suddenly, a sneakerheads worldwide go crazy and start buying. This company, used to a steady and reliable seasonal buyer, is now flooded with orders, running out of stock, and confronting a flood of new customers. While most brands beg for such a moment, it is the ultimate test of business agility, and a critical moment in time. You can win lots of new loyalists – or quickly become a flash in the pan.

What are the three foundational elements needed to succeed?


It starts, of course with customer data management. First, you need the scaled ability to capture first party data with consent. Every new sneakerhead coming to the website and mobile app must be encouraged to authenticate and start engaging with the brand. This involves offering a give-to-get for new customers (free shipping on orders, or a discount) and, more importantly, a scaled mechanism to capture that user’s permission to message her in the future. The experience must be seamless and explicit, as well as completely transparent.

For returning customers, you must have the ability to unify everything you know about them on the surface (SKUs viewed, lifetime value for commerce, loyalty points and status) but also go a level deeper. What is the true value of a customer? How many times do they return an item, and by what method so they return items? How often are they willing to pay full price? This data is only accessible by connecting the backend (financial ledger and supply chain) data to the profile. With a limited amount of a new item, you want to sell out – but you also want to reward your most loyal and truly valuable customers. This uplift is only possible by connecting the backend of business data to the frontend of customer engagement – call it ERP to CDP.


If a company has managed to create a unified data model across the systems I described above, and has, using data science, created models that can predict true customer value, and is able react to changes in behavior and market conditions – you still need to scale intelligence. In other words, given the above requirements, every customer cannot be evaluated individually, and every decision cannot live with a data science team. How strong is your organization’s ability to implement a machine learning framework that updates customer segments based on new information? In the background, ML models need to be continually tuned to changes in engagement across channels, understand how pricing and availability for specific products change behavior, and overlap segments to understand how different buyers of the same product react to campaigns, creatives, and different outlets for marketing and advertising. Lifetime value scores need to be calculated against ever-changing baselines – LTV can change based on product and customer mix over time, making yesterdays big spenders tomorrow’s regular shoppers. Going beyond marketing and advertising, what type of intelligence is required to create success in the call center, or an ecommerce site, or a sales call? Models are only as valuable as their ability to create value in the endpoint of a specific application.


After intelligence comes automation. How do you action the insights you have aggregated? Low value customers need to be suppressed from the campaign for the popular shoes. When a certain colorway or size becomes unavailable, customers with those preferences must also be suppressed – or encouraged to pre-order. Customers who are the most loyal need to be notified to “buy now” or invited to use their loyalty status to get placed in the front of the line. They need to be put into the call center queue first and, when they visit the website, have a one-click option for putting the right-sized shoes into their shopping cart, with their shipping preferences already pre-filled. When loyal customers come into the store and can’t find what they are looking for, the point-of-sale system must give the retail associate a next-best-offer or action that has a high probability of success. This is the new battleground in marketing – the ability to utilize intelligence at scale to render the right decision across both offline and online channels, in near-realtime.

At some point, every organization comes across a situation that tests their agility, and the customer experience team is often on the front lines. Your customer profiles need to get progressively richer, starting with marketing and advertising interactions, including cross-CRM data from sales, service, and commerce touchpoints – but also go deeper to leverage insights that can only be derived from the backend: ERP data. Intelligence must go beyond data science-provisioned models and scale with machine learning, such that the customer profiles can be frequently updated as lifetime value and propensity scores change based on realtime inputs. To adapt to a fast-moving market, driving that intelligence into the action surfaces of business must be as automated as possible.

This next phase of customer data management, that brings the backend of agile business process together with the frontend of customer engagement, is not about the next-best action or offer. It’s about finding the next-best dollar.

[This post originally appeared in The Future of Customer Engagement on 12 April 2022]