Actionable Insights From APT's Retail Practice
Header

Taking Your Customer Analytics Beyond Campaigns

July 31st, 2017 | Posted by Haley Jackson in Analytics | Financial Services | Hospitality & Travel | Insurance | Marketing & Media | Restaurants | Retail

Researching. Browsing. Transacting. These are all components of the customer journey, and now more than ever, they are taking place beyond physical channels. While the process of researching and browsing increasingly shifts to online and mobile platforms, the majority of consumers still prefer to transact in brick-and-mortar locations. As a result, while physical channels certainly remain relevant – to a varying extent – the way organizations across industries evaluate new programs and initiatives must be customer-centric.

Many best-in-class organizations have already established a standardized test vs. control approach with which to evaluate new initiatives. Business experiments at the customer level are common in areas such as marketing, where leading companies leverage customer analytics to measure the true impact of their campaigns. But as the question for executives shifts from “How do we collect more data?” to “How can we use our data more effectively?”, decision-makers must think outside of the box in their application of customer analytics.

Many initiatives across industries can be analyzed with a customer lens. Such programs include:

  • Brick-and-mortar openings and closures
  • New service or product introductions
  • Call center initiatives
  • Mass media investments
  • Supply chain initiatives (such as store delivery frequency)
  • Capital investments (e.g., remodels, new technologies)
  • Multichannel fulfillment programs (such as buy online, pick-up in-store)

While analyzing such initiatives with a customer lens provides a more granular and actionable perspective, there are two key challenges to doing so. The first is a limited ability to design tests upfront and establish representative control groups, and the second is aggregating disparate datasets across customer touchpoints for comprehensive measurement across channels.

The challenge of designing a conventional test at the customer level arises when the organization executing the test can’t actively choose which specific customers are exposed to a new product or in-store technology. One common, viable option to overcome this challenge is analysis of “natural experiments.” This method entails identifying instances of natural variation for analysis – cases where testing was not conducted intentionally, but a given action affected some customers and not others. By analyzing these historical changes, organizations can conduct analysis to inform the best possible future rollout of similar initiatives.

As a potential example of a natural experiment, consider an airline that wanted to introduce a new seat type on select aircrafts, and wanted to understand which passengers responded best. By comparing the subsequent behavior of passengers that experienced the new seat to that of similar customers that flew in existing seat types during the same period, the airline could pinpoint the true long-term impact of the new seat type introduction. This process would allow the airline to unlock these insights based on analyzing natural variation, rather than a designed test.

The second key challenge, aggregating disparate datasets across customer touchpoints, can also be conquered to reach customer-level insights. Organizations can solve this problem through leveraging a common data source or analytic platform that allows them to combine data sets for a full-picture view of the impact of an initiative.

As an example, imagine a bank trying to maximize new customer engagement – particularly for those that opened accounts through online or mobile platforms, who might be less engaged than customers who opened accounts in the branch. The bank could test a new proactive outreach strategy through its call center to determine if the program successfully drives customer engagement. By combining online and offline data sets, the bank could compare the performance of contacted customers to similar customers who were not. From there, analysts could segment results to identify which customers would benefit most from the outreach.

Or, consider a retailer that wanted to continue its network expansion, but wondered if new stores were significantly cannibalizing sales across other channels. To solve the challenge of aggregating distinct data sets for a holistic view, the retailer could incorporate customer data and map each store to a trade area of online sales. The retailer could then compare stores and trade areas that were affected by a new store opening to similar ones that were not to understand if added stores were generating enough new sales to make business sense.

Even though a customer may interact with a company in one channel, that does not mean customer strategy should be focused only in that channel. In an increasingly omnichannel world, the ability to understand customer behavior across channels and align analytics with strategy is quickly fading from being a competitive edge to a requirement for success. In order to achieve a holistic view of initiatives across the entire organization, it is critical to take customer analytics beyond marketing or promotional campaigns to new areas of the business.

Be Sociable, Share!

You can follow any responses to this entry through the RSS 2.0 Both comments and pings are currently closed.