Deloitte AI Institute
Inflation? Recession? Double-down on the customer experience, powered by AI
By Craig Brabec and Beena Ammanath
With lowered barriers to switching and increased sensitivity to digital-physical frictions, companies can gain advantage with a focus
on the customer experience by eliminating pain points. AI is an enabler
Open a news feed today and you'll be hit with reports of sustained, record-setting inflation rates coupled with continued prognostications of a coming recession. Companies are bracing for the impacts and responding by lowering inventories, tightening belts (including technology development) and overall lever pulls on corporate cost containment. Reducing operating hours, increasing customer service wait times and trimming services are occurring across industries in line with the traditional recession cost-containment playbook.
This comes with new risks, though, as the COVID-driven advancement of digital customer interactions has elevated consumer expectations for a seamless experience.
Despite the focus on digital, customer dissatisfaction with the integration of digital and physical services is becoming commonplace across industries. Informal discussions with colleagues provide numerous real experiences:
A fully app-enabled digital experience had one critical error—poor quality product delivered. Unpacking the produce at home reveals much of it is past the items' expiration date.
Digital/physical break: The customer contacts the store and is told, “To get a credit, you'll need to bring the receipt to the store. Don't bring the produce.” The customer's original intent was to not visit the store.
Customer orders and pays through the app—simple and seamless.
Digital/physical break: The restaurant is closed. Dissatisfied customer now needs to track down a refund, but no help is available through the app. There are multiple variants of this, including the menu item ordered is not available, coupon not accepted, app says store is closed when it is open, etc.
Online vehicle ordering:
Automobile manufacturer wins with a breakthrough vehicle and opens order reservations online.
Digital/physical break: The company cannot produce enough vehicles to cover all 2021 reservations and asks customers to come into a dealership to complete their order, resulting in re-orders (as 2022 model year configuration changes are made) as well as price increases. In addition, quality issues and supply chain shortages delay delivery, with some buyers having to wait another year. Boasting about a capability when it is not being executed destroys trust with customers.
The company was testing a new “virtual line” for one of their most popular attractions but did not notify their onsite customers.
Digital/physical break: A loyal customer that stayed at a premium on-property hotel to gain benefits of early park hours and express lines is unable to get in the ride line because the virtual queue was full for the entire day.
Digital done wrong can destroy customer value
So how did we arrive as this point? Digital has improved our customer experience with many companies, giving flexibility in choosing product delivery methods and timing. Telehealth has opened access to many who could not physically meet with their doctors. Financial transactions are easier than ever to execute and track, for both businesses and friends. While these technology developments were being planned, COVID quickened the pace.
According to a recent Deloitte study, companies have accelerated the digitization of their customer and supply chain interactions and of their internal operations, with three-quarters seeing the pandemic as fostering the formation of new partnerships and alliances. Companies have responded with speed and focus, although in many cases they've created pain points for the customer by not fully integrating the digital with the physical, at times increasing customer dissatisfaction with added inconveniences that overshadow the benefit of the digital.
AI can identify where the customer journey is not meeting expectations and help remedy it
This advance of digital interactions with customers brings an added asset to companies—more data about the details of their experience. By itself this data is valuable and drives insights, but the true monetization comes from joining this new source with other internal and external information sets.
In the case of the grocery order, anomaly detection at scale can help detect changes in a consumer's behavior compared to self, to similar customer profiles, and to the overall customer set. The negative experience can be detected though a customer's changes in ordering patterns even without capturing direct feedback. Algorithms can identify a change in buying patterns; for instance, reduction of sales in a product category (e.g., produce) can trigger alerts to an underlying problem.
Further machine learning can diagnose deeper connections: Is the anomaly connected with store sales pattern, day of week, supplier, weather, operational conditions like labor staffing, or a combination? AI can also power next-best-action recommendations for customers.
It all starts with the data, though
Data scientists are well equipped with current tools and platforms to attack these business issues. Often, though, they are limited by the quality of the data available. Data quality is determined by several factors, including completeness, accuracy, reliability, and timeliness. If information is not up to par, the data scientists will be limited in developing useful and actionable algorithms. Again, AI can help. AI can aid with data management with automatic data capture, synthetic data creation to fill in gaps in existing data, identification of erroneous data, and duplications.
In the example of the restaurant, detecting bad data in store operating hours can prevent the customer ordering from a closed restaurant, as well as multiple other failure modes, including helping restaurant operators detect variances in performance and potential root causes to address. Synthetic data can be a proxy for actual data, allowing scientists to run models even when source data capture failed.
AI can also be instrumental to predicting and avoiding future issues in the overall customer experience
To be effective, this mandates a well-defined customer journey map, with clear identification of failure modes or friction points along the map. In the example of the automobile manufacturer, this can be demonstrated in many ways. First, collecting customer communications can alert the original equipment manufacturer to common inquiries as well as ensure future responses and notifications to address key concerns. Combining production operations and supplier performance data with detailed order information can aid in better prediction of delivery dates for vehicles as well as optimize production lines, given order configurations.
In situations when orders must be converted to the next model year, AI can optimize options packages to align to customer orders as well as minimize options changes that the customer must accept. This also has the indirect effect of improving production operations.
The business cases for increasing the use of AI in the customer experience are clear. A study by Harvard Business Review found that customers who enjoy positive experiences are likely to remain customers for five years longer than customers who had negative experiences. In addition, delivering positive customer experiences can reduce your cost to serve customers by up to 33%.
AI can be the differentiator that helps eliminate the digital-physical frictions that are still common today.
Companies that keep the focus on preserving long-term customer value will better balance the immediate demands of cost containment while continuing to improve the customer experience.