In his emphasis of quality over short-term profitability, W. Edwards Deming’s management philosophy provides an excellent roadmap for analytics enablement best practices.

Dr. W. Edwards Deming is a popular figure amongst the data community—famously known for quotes such as “Without data, you’re just another person with an opinion” and the famous (potentially misattributed) quote, “In God we trust; all others bring data.” These quotes often surface in memes in data culture online, and are taped to the walls of our workstations.

But Deming was more than a famous and esteemed statistician. His career was something of a multifaceted object: Turn it one way, and he’s an academic. Turn it another way, and he’s a composer, writing liturgical choral arrangements. Turn it another, and he is a management consultant. Turn it yet another, and he’s the man who counseled Japan’s meteoric rise in economic power.

Deming was a data strategist to the core. He saw possibilities in data that no other person could, and produced world-changing insight in his lifetime. The most impactful of his influences, I would argue, is what he did for Quality Control, and how that helped Japan.

In 1950, Deming ventured to Tokyo to give a speech to a throng of business leaders who, at the time, reportedly represented nearly three-fourths of the industrial capital base of Japan.

As he was concluding his remarks, Deming professed, “I firmly believe that if product quality administration and market surveys…were used in a correct manner, you would be able to create a market for Japanese goods overseas, and the Japanese standard of living would greatly rise.”

History ended up proving Deming right. In fact, Deming’s core business management philosophy, “total quality management,” played such a central role in Japan’s massive economic boom of the 1970s and 1980s that, to this day, the Union of Japanese Scientists and Engineers still awards an annual Deming Prize to a Japanese organization that excels in company-wide quality.

Extending Deming’s Legacy to Data Strategy and Data Engineering

Deming is a wonderful person for analysts to learn from—the legacy he leaves behind teaches us principles we can use in many areas of our work. The strongest overlap is between his principles of quality control and our core discipline of tagging and data management.

Deep within our Analytics team, is a much-hidden discipline called Tag Ops. Tag Ops is the busy, buried engine of the team—it is the quiet engine that keeps the “intelligence machine” going. Analytics teams have a reputation for analyzing clients’ data, Tag Ops and Ad Ops teams are in the business of actually producing it.

The easiest example of Tag Ops comes from websites. While our clients and strategy teams see the websites we build as landing platforms for patients and HCPs to learn about a condition or treatment brand, as data strategist, we see a website as a place to gather important data about our audiences and our creative.

Web analytics can reveal what motivates consumers to visit a site, what they’re looking for, how old they are, where they live, what their interests are, and more. This is all very rich data that helps our clients understand their target audience. But, the richness of this data is contingent upon rigorous and strategic tagging enablement. Sure, all web analytics tools offer out-of-the-box codes for basic tracking, but the truly “good stuff” comes from the data you get from custom implementations.

Raw, clean data is the foundation of any analysis—indeed, it is the solid foundation upon which analytics maturity is built.

A Commitment to Built-In Quality Control to Get a Really Big Picture

Deming’s views on quality control centered on one basic principle—quality is the heart of the client-business relationship. He counseled companies to “build quality into the product in the first place,” meaning that any product is the sum of its component parts. And so it is with Tag Ops—the most complex and automated programs, such as trigger-based CRM for patient support programs and programmatic media, rely on good, solid datasets.

There are two major forces working against Quality Control in data. The first is the compressed typical lifespan of a digital campaign. With often just a handful of months to get creative into market, the rush to get things up and running can mean that proper tagging is ignored.

The second force is the fragmentation of teams. The responsibility for enabling end-to-end tagging, from digital promotion through to purchase and then into post-purchase activity, sometimes falls between all of the teams working on a client—and when this happens, the groundwork for that good robust data never gets lain.

And so, the forces would destroy good data—or never see it created in the first place—actually have nothing to do with the technical components of tagging. They have to do with process.

Deming was a huge, huge advocate for processes. “We should work on our process, not the outcome of our processes,” he said. Quality Control is at the heart of success, and process drives Quality Control.

An Industry that Builds to Learn—and Keeps It That Way

In the grand scheme of things, Analytics is still something of an emerging discipline. Over the years, our industry is learning how to build datasets, manage large-scale data, and how to use the data we’re creating.

In order to be good that this job, you have to know how to build the materials you use.  As Deming famously said, “Nobody should try to use data unless he has collected data.”

Written by Allegra Mira

I lead the Data Strategy practice at Saatchi & Saatchi Wellness, where I ensure strategic business insight for my clients by focusing on a balance of technical leadership and the subtle art of asking the right questions. As a data steward and a close partner to my clients, I am responsible for leading inter-agency efforts and building active teams. I pride myself in being able to bring out the best in the datasets *and* the technical teams I work with, to ensure a consultative level of insight which I bring to my clients. In the technical sense, my career path has taken me through leadership positions in digital analytics, media analytics (including paid search and SEO), business analytics, CRM & deployment rules, social media analytics and social listening, and primary research techniques based in survey design and harvesting customer-level data from the web.

One comment

  1. I think SSW has a gem of a genius on their hands. The article written by Ms. Mira is stunningly articulate. She delivers.

What do you think?

This site uses Akismet to reduce spam. Learn how your comment data is processed.