In the 2019 Big Data and AI Executive Survey, 92% of respondents reported increasing the pace of their investments in big data and AI. Then COVID-19 happened, and that thing that was already at the top of their to-do lists got underlined and circled in red… with a star next to it.
Perhaps no business has done more to harness the power of big data than Google, and they suggest that a six-month time period has resulted in approximately 10 years’ worth of change. Habits have been broken. Priorities have shifted. Supply chains have been disrupted. Our movement patterns have become much more tightly local. Digital adoption has increased, but we can anticipate ongoing channel fluidity as local recovery progresses in a non-linear fashion. Brand loyalties have been called into question. Economic uncertainty is such that business leaders are being called upon to do more with less. For many businesses right now, accelerating toward the availability of real-time customer insights is no longer something that can be perceived just as a way to gain competitive advantages—rather, it is a matter at the heart of their survival.
Despite this increased urgency, the major forces that have always made becoming data-driven an exceedingly difficult proposition are still in play, if not amplified. Among them are lack of a cohesive data strategy, competing organizational priorities, the need to make fundamental changes to infrastructure, the complexity of integrating various systems of record with a universal customer identifier as a join key, a shortage of in-house talent to execute the work, and an inability to rally the company culture around the adoption of data. Not to mention the cost of overcoming all of these things.
Now is not the time to cling to purist notions of the perfect data architecture, the coveted 360-degree picture of the customer from across all channels, or the almost magical AI that will surface the insights you don’t even know you should be looking for. Instead, here are five ways to roll up your sleeves and start making some bold moves:
Choose one problem and build the MVP to solve it
What’s the one thing you wish you could know about your business right now, with rapid frequency of update? Maybe some of your customers have moved from retail to e-commerce—but do you know which customers? Do you know how your customers are feeling about the quality of service they are getting from you? Or how their fluctuating levels of engagement are impacting your revenue? Now is a good time to revisit KPIs and see if previously-validated relationships between data still hold.
Early in the pandemic, the US National Multifamily Housing Council released a tracker indicating the percentage of households that paid rent, with a weekly frequency of update. Surely it was no small feat to aggregate this data from multiple property management software providers, but it’s hard to imagine a single metric that would be more valuable to their mission at this time.
It is far easier to align stakeholders on how to solve one specific analysis problem than how to build something that will solve all foreseeable analysis problems. It’s also far easier to create a single-purpose data set, integrating only the sources of relevance.
As a business that places fundamental value on being fast-to-market with actionable insights, we have repeatedly seen the power of a proof of concept that can truly ignite action. Data culture tends to snowball from the point where the first outcomes are felt.
Prioritize what impacts the customer experience
Bain has identified which of the 30 Elements of Value are most important to customers during a pandemic: reducing risk, reducing anxiety, and promoting a sense of affiliation and belonging. As brands are pivoting to ensure they deliver on these values, customer data will play a key role in determining what a revised experience should look like.
One of my favorite, very straightforward examples of how a business could use customer data to reduce risk came from a chain of fitness clubs building real-time visitation analytics using the swipe of member cards at the entrance turnstiles. This could be used to broadcast real-time updates into member portals and help members avoid showing up to a busy gym. Applying predictive models and coupling this with a reservation system could help members plan their upcoming workouts during likely off-peak times that were closest to their own preferences.
Since the turnstile data is a simple set of “in and out” events attached to a customer, it was not a complex technical operation to make visitation counts publishable in real-time. However, the benefit to members was fundamental.
Consider the decentralization of some data operations
Within larger enterprises, there are many functional teams with differing needs for customer data. During times of extreme uncertainty, empowering these teams to make more of their own decisions on analysis priorities and how critical datasets get assembled may help foster data culture and improve the agility of the business.
Robust analytical capability depends on a multidisciplinary team. Consider which of these roles should be part of a center of excellence in your business, ensuring one point of governance, vs. which may live within functional teams and/or be outsourced. The answer may be different depending on whether the business is running under normal circumstances or in a state of emergency.
Incorporate data from beyond the enterprise
eMarketer reports that third-party research and social listening top the list of sources marketing leaders are using to get information about consumers during the pandemic. In part, this will be because these sources of insight are free of the challenges associated with wrangling first-party data. Most importantly, however, data from beyond the enterprise can give a much more robust picture of the market as a whole, and how customers are behaving, regardless of whether or not they are interacting with your business.
It may have been difficult for automotive dealers to understand how the pandemic could actually increase car buying interest, if they were looking only at what was happening on their lots. Data from Google showed that search interest in car buying was peaking, and that the majority of car buyers were interested in buying online, but only a few were given the option to even do a test-drive at home.
Ultimately, a successful pivot strategy requires an understanding of both the audience you’re reaching, and the opportunity being missed.
Adopt an iterative data maturity model
As we developed the Proove Intelligence Data Maturity Model, we recognized that the highest level of this—where the impact of data is artfully infused into every experience customers and employees have with your business—is more of a guiding vision than an achievable reality. When we talk about becoming a data-driven organization, this is work that is never truly done. The data we have and how we need to apply it change radically with the world around us. We need to think about creating the culture and the capabilities that will allow us to thrive in this state of constant change.
Enrolling the entire business in a data maturity model is a great way to have more honest conversations about where you are vs. where you want to go, what the limitations and risks are in the current state, and what the impact to the business will be at each successive level achieved. It’s also a great way to highlight that doing more with data ultimately requires more cross-functional participation, to carry the impact through to how your business fundamentally delivers value for customers. At Proove Intelligence, we have based our entire business around the idea that having great data isn’t what matters—but what you do with that data matters a lot.
Need help accelerating your customer insight plans? Get in touch!