Like a tangled garden hose that slows a steady flow of water to a trickle, unruly and poorly managed business data produces substandard outcomes. And identifying the root cause of bad data can be as frustrating as trying to untangle a hose, unsure which end to pull to produce a clean, steady, trustworthy stream.
For marketers intent on creating an uninterrupted output of hyper-personalized customer experiences across all channels, in real time and at the cadence of an individual customer journey, any impediment has a deleterious, cascading effect. Bad data leads to poor segmentation, which leads to inferior campaign design and journeys, which in turn leads to shoddy execution.
Alas, many marketers are all too familiar with the problem. Consider the findings from independent research firm Vanson Bourne from earlier this year. In The State of Data Management – The Impact of Data Distrust, 91 percent of IT decision makers agree they need to improve the quality of data in the organization, with 77 percent saying they lack trust in their business data. The offshoot is predictable; 76 percent said they are missing out on revenue opportunities, and 72 percent see a negative impact on customer engagement due to a lack of timely data insights.
From an operational standpoint, data wrangling is also an enormous drain on creativity. The same report found that for every employee that works with data an average of four working hours is lost each week trying to resolve issues related to data prep. An older study from Harvard Business Review had similar findings, determining that employees who work with data waste up to 50 percent of their time hunting for data, identifying and correcting errors, and seeking confirmatory sources for data they do not trust.
Bad Data, Bad Outcomes
What predictably happens is that marketers reluctantly accept a trade-off; with looming deadlines for segments to build and campaigns to design and execute, they move forward with less-than-perfect data knowing full well that some experiences will miss the mark.
On the segmentation side, improperly matched data or substandard data governance will lead to inconsistent segment counts as different marketing teams may access different datasets. Segments will also be inaccurate, with customers improperly placed in segments that are not truly representative of what matters most to them.
In addition, poor data quality prevents marketers from adding attributes to segment on; without a quality view of the customer, it is difficult to build unsupervised machine learning models where the marketer provides data and asks the model to find meaningful patterns. The same holds true for a supervised model, where marketers provide historical data for a model to find a predetermined attribute such as customer lifetime value or churn probability. In both cases, if the data is off the outcome cannot be trusted.
That lack of trust become insidious; if a marketer can’t trust the accuracy of a churn probability, for instance, then what-if analysis that depends on the accuracy of the model becomes flawed all the way through to campaign design and execution.
If the accuracy is flawed, potential trigger points such as days since last purchase or 30-day average spend become skewed, leading to a customer who may incorrectly be identified as a churn candidate receiving an incorrect or poorly timed message.
Timing is Everything
For interactive journey design and other situations that call for real-time decisions, the consequences of working with bad data are just as harmful. Think of a typical online session for a customer browsing a brand’s website. Whether an anonymous or known journey, a marketer must have confidence that a unified customer profile is completely accurate and up-to-date in order to support relevant engagement while the session is ongoing. Data is being collected as the session continues, and unless it is cleansed, matched and merged within milliseconds of it being ingested a marketer will fall behind the customer journey. The longer the lag, the greater the distance and the greater the chance to deliver the wrong message, content or action.
For any type of situation that demands a real-time response – a call center engagement, a product recommendation engine, shopping cart abandonment, etc. – marketers do not have the luxury of playing out in advance all the permutations of a customer’s potential behaviors. Being perfectly responsive to a unique customer’s needs in the moment depends on trustworthy data. The alternative is to pull back on design options; instead of a personalized message or action based on the unique customer journey, a marketer works with a handful of options or paths, hoping that “good enough” carries the day.
Good Data, Better Decisions, Optimal Outcomes
With the cost of bad data becoming more clear, it stands to reason that perfect data elevates trust and confidence, saves wasted time and helps marketers tap into more creative and innovative design strategies, deeper analysis, better experimentation and more reliable execution.
On the segmentation side, when data quality processes are completed at ingestion, new attributes can be created on-the-fly because with a more granular view of the customer, automated machine learning models detect more meaningful patterns in the behavior of a customer or group of customers. Segmenting off more attributes instills more confidence in the accuracy of predictions. Whether it’s CLV, churn potential of another variable, having more attributes takes out the guesswork.
Flowing into campaign design, marketers can then base campaigns on any number of new triggers – in real time or otherwise – with full confidence that each customer’s path is optimized to a unique customer journey at the precise cadence. This will then of course translate to flawless execution, with a next-best action optimized for relevance for a specific customer, delivered at precisely the right moment.
Ultimately, ensuring data quality at the point of ingestion creates a closed-loop system that leads to higher levels of trust and higher data quality; segmentation, campaign design and execution that stem from high-quality data generate better responses and/or more intended outcomes, whether higher CLV, reduced churn, more purchases, higher average spend, etc. Higher-quality responses, in turn, lead to improved segmentation and the cycle continues, with the result being more perfect customer experiences for more customers.
The idea that comprehensive data quality and accurate identity resolution at the moment of data ingestion are the foundations of delivering superior customer experiences was the driving force behind the release of Redpoint In Situ, the first cloud-native, data quality-as-a-service (DQaaS) that delivers perfected data and resolved identities in real time, using exclusively first-party data, where the data resides.
Because when 77 percent of decision makers say they don’t trust their business data, and 72 percent see a negative impact on customer engagement due to a lack of timely data insights, there has to be a better way. Every minute a marketer spends resolving issues related to data prep is a minute lost to a marketer’s ability to experiment with creative new ways to engage with a customer. Perfect data removes the all too familiar obstructions, letting marketers cultivate continually improving, highly relevant experiences with trust in the data, trust in the process and trust in the result.
For more on In Situ, check out recent coverage featured in Tech Target.
Related Content
Can Your CDP Deliver Perfect Data? How a Need for Perfection is Shaping the Market
Why Data Quality Matters for IT and Business Stakeholders
How Does Data Governance Differ from Data Quality?
Be in-the-know with all the latest customer engagement, data management, and Redpoint Global news by following us on LinkedIn, Twitter, and Facebook.
