At a recent Informatica Customer Experience (CX) VIP Seminar, an attendee remarked that “CX innovators use more data. Amazon, Google—they all use a lot more data than my company.” That’s true. Those companies understand their customers deeply and make personalized recommendations that usually hit the mark. The room tended to agree, and the general theme of the workshop was to use more data for CX decisions and processes. However, that sentiment bears closer scrutiny. It isn’t just that CX leaders use ‘more’ data to make them successful. Their success lies in ‘how’ they use data to impact CX, and that approach does indeed allow them to use ‘more’ data, successfully.
Although using data is the correct direction, rushing to get “more data” can often lead to bad results. Most organizations are thirsty for intelligence to improve CX. Lost in a desert of competing data silos and insights, an oasis was identified: the “data lake.” If all data could just be put into one repository, surely all their problems would be solved, right?
Unfortunately for most companies, the data lake is more mirage than oasis. Many companies complain of ‘data swamps,’ un-curated dumping grounds that are poorly governed and unusable by the average businessperson. Others have established curation and governance, so that data sets are cleaned, described, and usable. But even these are still a mirage because they aren’t connected. The data in the lake remains siloed and doesn’t quench their thirst for insight. In order to achieve game-changing decisions for CX, the data needs to be connected to give a much more complete understanding of each customer. This approach can have harmful repercussions, such as when businesspeople can lose confidence in data itself: (“I thought more data was the right approach and the data lake was the answer, but we don’t seem to be getting any insights at all from the lake.”) Sometimes a rush to a data mirage can create a culture of data mistrust. One thing that CX leaders have in common is a deep trust in their data.
CX leaders have focused on specific techniques to better use data to impact CX. Here are 5 best practices we’ve observed in leaders:
1. Connect Data – Leaders integrate data sets and deeply connect them into customer 360 data graphs. They break down silos and synthesize data sets together. They use identifiers, keys, and new matching techniques, such as contextual matching, to make more data connections. Leaders automate this process of data synthesis, so that data scientists, analytics users, and business decision-makers don’t have to waste time joining data, they simply use it to make better CX decisions.
2. Trust Data – CX leaders establish a deep trust in data by making the data management process transparent, so business users can not only see how data was connected and matched into a customer 360 graph but can examine the confidence levels of matches and the lineage of data. When data management is a black box, business users can develop a mistrust of data (“I don’t know how that was matched together”). CX leaders go a step further: Not only do they make the data management process accessible to business users, but they actually participate in it. The traditional way of making manual decisions on every data connection doesn’t scale, so modern approaches enable businesspeople to train machine learning (ML) algorithms that do the synthesis. By making a small number of link/non-link decisions during implementation, the ML algorithm quickly learns how to join data with the same confidence levels that humans deem acceptable. This approach has a dual benefit: one, the algorithm is very accurate and automates the same decisions humans would make: and two, humans trust the algorithm because they can audit it and they feel a sense of ownership and control because they trained it.
3. Experiment – More data doesn’t necessarily mean all data. CX leaders are able to ingest and experiment with data subsets to see if they are valuable for CX decisioning. Do emails contain clues for behavioral segmentation? If the answer is yes, then it makes sense to ingest one year’s worth of email history. CX leaders are also able to rapidly determine the suitability of data sets for specific use cases. This is because they don’t have to start with all data; they start with use cases and determine which data sets will need to be connected to improve the outcome. Eventually they end up with very well-connected 360 graphs, but they get there, one use case and a few data sets at a time.
4. Act on it – CX leaders drive deeper insights out of data, and make sure they are actionable. For a leader, everything is driven from a CX vision to a use case to data set(s) needed to create new insights for that use case. They never go the other way around (that is, they don’t connect all data and then find a use case for it—AKA “if you build it, they will come”). Leaders also prioritize iterative, quick deployments, preferring to deliver value for a use case in multiple iterations, before expanding to new use cases.
5. Adopt it – Technology is an enabler, but CX leaders understand that employees have to change processes and adopt new data and insights in order to be successful. Leaders use more data because they have created a data-driven culture. They drive adoption of new technology and ensure that marketers, analytics users, and business owners are willing to change their processes, behaviors, and even incentives to make the CX program a success.
Most companies want to improve their CX, and even become a leader in CX. Consider two pieces of advice. First, don’t copy a CX leader—real leaders forge their own path. Second, start with a realistic self-assessment of your current capabilities and your existing CX strategy. Remember, to be a leader you don’t have to be great at every aspect of CX. You do need to be great at the moments of truth—the interactions that matter most to your customers. Do you know what those are for your organization?
To learn more, please watch the on-demand webinar, "Evolving the Customer Experience to Win".