Last week, CMSWire posted a great article by Matt Stevens (IT Executive, Toyota Motor Europe) on how Toyota was forced to rethink their data management practices to keep clients delighted in the era of Big Data.
Matt outlined the goals and expectations of the customer experience in retail today – all oriented around the notion that “customers these days expect much more connected experiences” and in that spirit Toyota Motor Europe (TME) “began with the premise that our customers deserved to have seamless access to their information whenever they used apps, logged into websites, booked services or visited our retail locations — all with their data synced and updated in real time.”
Unfortunately, while all large organizations focused on customer growth and satisfaction would agree with this pursuit, most fall short in delivering on the promise. The primary reason: their data environment has become too large, fragmented, and complex to handle.
TME outlined their own problems in this regard when they attempted to master customer records across Europe, saying “while we found that [traditional data management] approaches could be effective with small amounts of relatively static data, none could be scaled effectively to meet our ever-increasing volume and data complexity requirements.”
What TME experienced in their data environment is no different from any large enterprise that has attempted to curate a high volume and variety of datasets. Traditional approaches – whether driven by legacy technologies or manual effort – are simply not fast enough and can not scale to tackle these large challenges. Managing data at scale needs a new model.
Machine Learning-Based Approaches Are Built For Large, Complex Projects
TME considered many alternative options for solving their customer data problem but “needed a solution that would consolidate customer data across the entire European continent – that was scalable, efficient and respected local needs”
This led TME to evaluate and select Tamr’s human-guided machine learning-based approach. Unlike traditional approaches to unifying data for analysis, Tamr connects and cleans datasets through the use of automation – essentially using sophisticated algorithms to recommend linkages within the data and ask internal experts for validation where needed to correct and further automate the modeling in the future.
TME’s willingness to adopt a new, bottom-up approach to preparing data for analysis is what sets them apart as thought leaders. They recognized that in order to satisfy customers and deliver a connected experience, they needed to prepare all of their data quickly – not just some of it over the course of months or years. Speaking to the benefits of a machine learning-based approach to managing their data, Matt highlighted:
- Scalability – “with one platform, we could map our many disparate data sources to a single view of the customer. Not only would this capability have been prohibitively cost- and time-intensive using traditional approaches, but the economies of scale generated by machine learning let us capitalize on our increased ability to add and integrate more sources of data.”
- Flexibility – “we were drawn to the enterprise data unification approach because we knew that consolidating our data would never be a one-off exercise, and enterprise data unification would let us accept entropy, or disorder, in our data as a fundamental property”
- Collaboration – “including expert feedback in the generation of our model ensured trust and accuracy by making those closest to our customers responsible for data quality.”
Follow The Leaders
TME’s use of Tamr’s enterprise data unification technology to connect and clean customer records across the entire continent will have a remarkable impact on customer satisfaction and growth. Consumers will have a connected experience across all channels and TME will gain a complete, clean view of each of their customers for marketing and other purposes.
This new model of data management is rolling out across leading companies in many major industries – including organizations such as General Electric, who has realized hundreds of millions of dollars in cost savings through the implementation of Tamr’s technology within their procurement data environment.
Toyota’s continued rollout of Tamr will further emphasize the importance of machine learning-based data management technology in today’s data environment. It will also demonstrate how embracing new approaches to working with data can pay huge dividends.
TME, General Electric, and others have embraced this new world and we hope to see more organizations follow the leaders.