Care about what happens after the analysis.
Data users known there are a lot of reasons to collect data, but for it to truly be used, they know the focus needs to be on what happens after it’s extracted, transformed, loaded, and analyzed. They focus the bulk of their energy on what happens after the analysis. Data users know the value of their work isn’t in their data warehouses or math but in the actions that they lead to.
Data users work hard to hide the actual mechanics of data use. Like the sugar in a really good cake, data users make the models, data sets, and technology invisible — they make the next step the star. They evaluate the success of their work according to how much movement takes place not the number of variables, processors used or creative visualizations.
Data users have a non-traditional approach to working with data. They practice the art of what they do as much as the science. If things end up being vague, vanilla and lukewarm after an analysis, they know it’s mostly from a lack of creativity not a lack of “good” data or technology.
Data users are a hybrid of trained mathematician, programmer, and designer. They merge analytical skills with industrial, graphic, user experience, and information design skills. They believe success after the analysis is dependent on the laws and art of these more experiential disciplines.
Data users want their work to end up as close to the transaction as possible. They know data being used at the cash register has much more value than data being used in corporate HQ.
Data users approach each project they take on with a business model mindset. They have an understanding of the core elements involved in their organization’s value generation and they don’t let the complexity of big enterprise cloud the basics that make the business work. They use these business model principles to simplify and hyper- focus the questions they ask data to answer.
Be really good at end products.
Data users use analytics to ask people to change and do. they’re not interested in work that measures egos. they’re not in the business of red- light green-light health-checks. Data users ask for action.
Data users make things that normal people can understand and act on. Unlike classic analysts, data users don’t let data-barfing and spreadsheet- slinging be the product of their work (even if it’s wrapped in pretty charts). They have the guts to work until all that is left to report is a crystal clear, one-sentence call to action.
Data users keep their action-asks simple and realistic. They ask for clicks not projects. They think in terms of next Tuesday not Q3.
Data users know that data’s value is only generated when something happens. They think in terms of to-do lists. To-do lists are the simplest yet most influential type of “report”. To-do lists are what data users want after an analysis. They curate all of the possible next steps their math points them to and then filter down to the most realistic and impact ones.
Data users know that the big ROI traditionally sought in data-projects can limit data’s power and that lots of little to-dos that are likely to get done out perform a few big to-dos that sit untouched. Data users look for little-bet to-dos with higher chances of return instead of big-bet to-dos with hope and prayer type of returns. They like the reality of finding 100 certain small things over one big, uncertain, magical thing.
Data users make stuff not just analyze it. They wrap their asks of people in products, not reports. They realize that well designed products are much better than pie charts at getting people to do things. People engage with things they can and want to use. They build products with the minimum feature set required to get the doers doing — often with single-feature, analog solutions that are as simple as they are well designed.
Data users operate in a continuous prototype-feedback-enhance loop which allows their work to be in the right place, in front of the right people, with the right action at the right time.
Take pride in the connections.
Data users master the connective characteristics of data. They’re tailors that pride themselves on the creative ways data gets stitched together. They don’t spend their resources on taming the bigness of data but focus on taming the way it connects. Data users believe most of data’s asset value is in the other data sets you can hook it to. They value column count over row count.
Data users don’t believe human experience and analytics are mutually exclusive. They believe they’re equals and work to connect them. They know skillfully mixing gut-feel and number crunching leads to things that actually make sense in the real world. Data users work with qualitative and quantitative data to tell the real story. Separate they’re just information, but together they can make what happens after the analysis realistic.
Data users know that the majority of data is connected to people in some way and that it’s people who bring transactions to life. They know finding and enabling these connections is one of the most powerful approaches to putting data to use. They use people driven data to inspire and start as many conversations as they can and then generate value by putting these data driven conversations to work internally and externally.
Data users believe everyone has data and that all data can be connected. They’re not lazy when it comes to creatively solving for what data and systems lack from a connections stand point. Creative data hooks are a data user’s badge of honor.
Know the role of humans.
Data users understand the many roles people play in data use. They’re fluent in the languages of organizational theory and politics, but spend little time managing and catering to them. Instead they leverage or short-circuit them to get things moving.
Data users refuse to let man-made information silos become excuses for not getting past the analysis. They figure out how to use data in spite of other teams’ attitudes and choices and then work to get these same teams collaborating with them.
Data users know that the majority of people are not in love with data, but that most of the actual day-to-day doers of things in a big organization are being asked to be analysts — whether they know it or not. Data users want to elevate the literacy of their colleauges and do so by obsessing over making the results of their work irresistible to the doers in an organization. They use design thinking to empathize with the doers not other analysts, strategists or the industry.
Data users don’t believe analytics in a big organization is a zero-sum game. They believe, like two artists painting the same scene, two analysts answering the same question employ their own unique approaches, yielding two distinctly valuable yet directionally similar results. Data users are not analytical competitors who have based their careers on having data or models that others don’t.
Data users spend very little time hunting for singular sources of data truth. They avoid the paralysis of analytics by committee and let others make the rules while they make action.
Data users tend to polarize. They don’t create things that everyone agrees with or likes. They’re pleased when something doesn’t work for one group because that means it’s likely working perfectly for another group.
Data users know that humans are the most important part of any data work. They’re not caught up in the trendy pursuit of a human-less, analytics factories. They know that good data use depends more on people than computers and algorithms (a dependence that deepens as the complexity of data and systems grows). They don’t waste time automating a computer to be a human. They let a computer remember things and find patterns, and let humans use their “gut” to ask the questions and then interpret and balance the answers with reality.
Don’t get distracted.
Data users are not distracted by technology. They acknowledge it has a huge role in data use, but not THE role. Data users can do more with SQL and Excel than most complexity advocates and vendors want to believe. They task technology to strengthen and enrich what happens after the analysis not make the management of data easier.
Data users fight against sexy-tech for the sake of sexy-tech. If there’s a vendor selling it, a data user is skeptical. They know bleeding-edge toys are powerful but that they can distract them. They evaluate all tools through the lens of hunting down action.
Data users are not interested in the academic game of defining Big Data, the debate over the effectiveness of visualization techniques, the arguments over data science’s role, or the fights over competing approaches to data architecture. They master the recipe of technology and analytics, using the appropriate mix of each to create the perfect action hunting ecosystem.
Data users pledge no allegiance to a specific tool, technology, or approach. SAS, R, Python, Amazon, Microsoft, an abacus . . . a data user would choose a stone tablet over any of them if it meant creating to-do lists faster. If a smoke signal gets people moving, they use it. Big or little data, structured or unstructured, distributed computing or not — data users are happy with each of them when they complement getting past the analysis, not distracting from it.
Data users tend to play with data more than they analyze it and avoid the distracting aspects of data environments that make playing hard. They think like hackers and prefer “development” environments where they can go against convention to answer tricky questions in rule-bending ways. They need the freedom to change the approach on the fly and see the impact quickly. The laws of traditional business intelligence frustrate them and the requirements of production level data environments, while necessary for the final product, hold them back when they’re playing. Data users build and work in data playgrounds.
Data users are not distracted by the data “unicorns” the Harvard Business Review writes about. They ignore the hype surrounding data in today’s popular press. They’re not interested in winning the analytical lottery. They don’t expect to cure cancer with every data set and analysis. They don’t demand or require PhD math backing from themselves or others, and instead lean on everyday analytics and techniques because of their effective simplicity — simplicity that makes for exciting actions not necessarily exciting journalism.
Data users are less passionate abut decimal places, confidence intervals and significant digits than they are about getting a to-do list in the right person’s hands. Unless they’re landing airplanes or selling pacemakers, they don’t let perfection become an excuse. They recognize most data sets and models can’t scientifically achieve absolute perfection.
Data users don’t wrestle with data incompleteness and are immune to outlier paralysis, not letting 3% of cases render data or systems useless. They accept and design for the norm of 80% accuracy, 80% population representation, and 80% completeness — they use what they’ve got and get going.
Data users work with what they have. They bootstrap. They hack. They do anything to keep from waiting for something better to come along before they use data. Budget plays little role in a their ability to use data. Data users don’t let one more variable, one more row of data, one more type of chart, one more byte of processing speed, or one more team member distract them.
Data users are comfortable with the wild nature of data and work with what the data gives them. They don’t try to force the data to be something it isn’t, spending very little time trying to teach it tricks or behave in certain ways.
Data users start as manually as possible. They learn to feel the rhythm of the data and the analysis before automating it. They don’t let the scalability of what they’re doing distract them too early.