Even well-run operations that generate strong analysis fail to capitalize on their insights.
Efforts fall short in the last mile, when it comes time to explain the stuff to decision makers. In my work lecturing and consulting with large organizations on data visualization dataviz and persuasive presentations, I hear both data scientists and executives vent their frustration. They say decision makers misunderstand or oversimplify their analysis and expect them to do magic, to provide the right answers to all their questions.
As the cathedral is to its foundation so is an effective presentation of facts to the data. How could this song remain the same for more than a century?
- The Presidency: The Hardest Job in the World - The Atlantic!
- Influence and Persuasion Skills: Leading Strategies for Getting Results.
- A Caregivers Guide to Lewy Body Dementia?
- Data Science and the Art of Persuasion;
- Article Two of the United States Constitution - Wikipedia.
For one, the tools used to do the science include visualization functionality. But in the rush to grab in-demand data scientists, organizations have been hiring the most technically oriented people they can find, ignoring their ability or desire or lack thereof to communicate with a lay audience. They still expect data scientists to wrangle data, analyze it in the context of knowing the business and its strategy, make charts, and present them to a lay audience.
To begin solving the last-mile problem, companies must stop looking for unicorns and rethink what kind of talent makes up a data science operation. It relies on cross-disciplinary teams composed of members with varying talents who work in close proximity. In the early 20th century, pioneers of modern management ran sophisticated operations for turning data into decisions through visual communication, and they did it with teams.
It was a cross-disciplinary effort that included gang punch operators, card sorters, managers, and draftsmen they were nearly always men. Railroad companies and large manufacturers were especially adept, learning the most efficient routes to send materials through factories, achieving targets for regional sales performances, and even optimizing vacation schedules.
In general, the stories I hear follow one of these scenarios. See if you recognize any of them. A data scientist with vanguard algorithms and great data develops a suite of insights and presents them to decision makers in great detail. She believes that her analysis is objective and unassailable. The language she uses in her presentation is unfamiliar to her listeners, who become confused and frustrated. Her analysis is dead-on, but her recommendation is not adopted. A business stakeholder wants to push through a pet project but has no data to back up his hypothesis.
He asks the data science team to produce the analysis and charts for his presentation. The team knows that his hypothesis is ill formed, and it offers helpful ideas about a better way to approach the analysis, but he wants only charts and speaking notes.
A top-notch information designer is inspired by some analysis from company data scientists and offers to help them create a beautiful presentation for the board, with on-brand colors and typography and engaging, easily accessible stories. But the scientists get nervous when the executives start to extract wrong ideas from the analysis. The team approach persisted through most of the century.
In her book Practical Charting Techniques, Mary Eleanor Spear details the ideal team—a communicator, a graphic analyst, and a draftsman still mostly men —and its responsibilities. In the s things started to split. Scientists flocked to new technology that allowed them to visualize data in the same space a computer program where they manipulated it. Visuals were crude but available fast and required no help from anyone else.
A crack opened in the dataviz world between computer-driven visualization and the more classic design-driven visualization produced by draftspeople finally. Suddenly anyone could instantly create a chart along with overwrought variations on it that made bars three-dimensional or turned a pie into a doughnut. It helped make charts a lingua franca for business. It fueled the use of data in operations and eventually allowed data science to exist, because it overcame the low limit on how much data human designers can process into visual communication.
Most crucially, it changed the structure of work. Designers—draftspeople—were devalued and eventually fell out of data analysis. Visualization became the job of those who managed data, most of whom were neither trained to visualize nor inclined to learn. The speed and convenience of pasting a Chart Wizard graphic into a presentation prevailed over slower, more resource-intensive, design-driven visuals, even if the latter were demonstrably more effective.
With the advent of data science, the expectations put on data scientists have remained the same—do the work and communicate it—even as the requisite skills have broadened to include coding, statistics, and algorithmic modeling. Think of him or her as a hybrid of data hacker, analyst, communicator, and trusted adviser.
Theodore Roosevelt, 25th Vice President (1901)
The combination is extremely powerful—and rare. Here are the ways that various talents are involved as a data science project proceeds from gathering data to developing insight to presenting to stakeholders. A rare combination of skills for the most sought-after jobs means that many organizations will be unable to recruit the talent they need. They will have to look for another way to succeed. The best way is to change the skill set they expect data scientists to have and rebuild teams with a combination of talents.
An effective data operation based on teamwork can borrow from Brinton and Spear but will account for the modern context, including the volume of data being processed, the automation of systems, and advances in visualization techniques. It will also account for a wide range of project types, from the reasonably simple reporting of standard analytics data say, financial results to the most sophisticated big data efforts that use cutting-edge machine learning algorithms. Not quite. Rather than assign people to roles, define the talents you need to be successful.
One person may have several talents; three people may be able to handle five talents. Project management.
Because your team is going to be agile and will shift according to the type of project and how far along it is, strong PM employing some scrumlike methodology will run under every facet of the operation. A good project manager will have great organizational abilities and strong diplomacy skills, helping to bridge cultural gaps by bringing disparate talents together at meetings and getting all team members to speak the same language.
Data wrangling. Skills that compose this talent include building systems; finding, cleaning, and structuring data; and creating and maintaining algorithms and other statistical engines. People with wrangling talent will look for opportunities to streamline operations—for example, by building repeatable processes for multiple projects and templates for solid, predictable visual output that will jump-start the information-design process.
Data analysis. The ability to set hypotheses and test them, find meaning in data, and apply that to a specific business context is crucial—and, surprisingly, not as well represented in many data science operations as one might think.
Some organizations are heavy on wranglers and rely on them to do the analysis as well. But good data analysis is separate from coding and math. Often this talent emerges not from computer science but from the liberal arts. The software company Tableau ranked the infusion of liberal arts into data analysis as one of the biggest trends in analytics in Critical thinking, context setting, and other aspects of learning in the humanities also happen to be core skills for analysis, data or otherwise. In an online lecture about the topic, the Tableau research scientist Michael Correll explained why he thinks infusing data science with liberal arts is crucial.
U.S. Senate: Theodore Roosevelt, 25th Vice President ()
Subject expertise. People with knowledge of the business and the strategy will inform project design and data analysis and keep the team focused on business outcomes, not just on building the best statistical models. Joaquin Candela, who runs applied machine learning at Facebook, has worked hard to focus his team on business outcomes and to reward decisions that favor those outcomes over improving data science.
This talent is widely misunderstood. Rather, people with design talent develop and execute systems for effective visual communication. In our context, they understand how to create and edit visuals to focus an audience and distill ideas. Information-design talent—which emphasizes understanding and manipulating data visualization—is ideal for a data science team. Nothing was more hateful to these interests than corporate taxes, especially on companies that were, in their eyes at least, providing a public service such as water or gas.
By forcing the franchise tax through the legislature, Roosevelt made powerful enemies who informed Senator Platt of their disapproval. The boss worried that his hold on the party was fading because of his inability to control his governor. He began reconsidering his relationship with Theodore Roosevelt. Getting rid of Governor Roosevelt did not promise to be easy. While the impetuous governor may have made enemies in the business community, he was immensely popular with the public. In fact, it was this popularity that made him such an effective governor.
One reason Senator Platt had acquiesced in Roosevelt's nomination was that the senator anticipated controlling the state assembly. As long as Platt's will was supreme in the legislature, the governor's most threatening schemes could be defeated. Roosevelt, however, had developed a weapon capable of changing the minds of wavering legislators. During his campaign for election, the governor had demonstrated the power of his personality; as one observer remarked, "Teddy. It was electrical, magnetic. As the most vigorous governor most New Yorkers had ever seen, Roosevelt used constant publicity to push for his programs.
He regularly held two press conferences a day and consulted experts of all kinds on complex issues. The growing media of the day feasted on this constant flow of information, and the public loved it. Under such intense public scrutiny, only the most intransigent of legislators cared to challenge Roosevelt. This method of public persuasion would serve Roosevelt well in the future, as it defined his political style and formed his most lasting contribution to the political process in the twentieth century.