Fundamental change and improvements are needed to raise the maturity of digital analytics, a.k.a. web analytics. This article provides some background for this assertion and highlights 3 specific areas of need and opportunity. It is not intended as either a complete list, or even a perfectly defined list. Solutions are directionally indicated, but the article is intended to provoke thought; it is not a detailed roadmap.
Maturity Of Digital Analytics
Information technology and the Internet - therefore the associated analytics - are immature relative to centuries-old fields such as literature, mathematics, science, art, etc. Certainly digital analytics is maturing; mostly gone are the days where nuggets of knowledge about customers were culled from server logs. "Big data" and mobile metrics are in play; attribution, predictive models and other advanced techniques are now applied to digital intelligence.
Inefficiencies are a natural characteristic of an immature field, and this holds true for digital analytics. There is much hand-crafted, custom work, data quality is often poor, and standards are limited and few. Effectiveness also varies and the business value of digital analytics is often missed. Stakeholders lack necessary knowledge and analysts begin careers from widely varying backgrounds.
The premiums paid for digital analytics today will not last; businesses managers and marketers increasingly demand results with greater value and usefulness at lower cost.
Call To Action
Achieving significant change requires thinking and acting "above and beyond". For example, tag management is good, it removes much pain - but it is focused on the means, not the end. As Albert Einstein said: "We can't solve problems by using the same kind of thinking we used when we created them
President Kennedy's call to go to the moon galvanized the nation around a shared purpose, with incalculable benefits: economic prosperity, technological innovations, and advancements in education. Digital analytics needs a similar call, but envisioning and executing some monumental goal would be overkill.
A more relevant approach might be inspired by the German mathematician David Hilbert
(1862 – 1943). In 1900, he put forth a list of 23 unsolved problems at the International Congress of Mathematicians at the Sorbonne in 1902. He said: "Who among us would not be happy to lift the veil behind which is hidden the future; to gaze at the coming developments of our science and at the secrets of its development in the centuries to come?"
Hilbert's list was a catalyst, influencing mathematicians for over a century and propelling the subject to its modern state. Some problems were solved within a short time, others solved over the years since and others remain unsolved to this day. Likewise, identifying and then solving a list of major pain points and opportunities in digital analytics can achieve the far greater efficiencies and effectiveness necessary in the field.
" is not about how to get some new feature of Adobe Site Catalyst to work, or to integrate data from Pinterest, or some passing fad. It speaks to the heart of digital analytics: what it is, how it can be significantly improved, how the "veil" can be lifted "behind which is hidden the future". Trying to see a century from now makes no sense in this dynamic field, but the field can leap-frog some current pain points.
With inspiration from David Hilbert, below is an initial list of 3 opportunities for significant advancement in digital analytics.
Need/Opportunity #1: Data Collection
How can records of customer behaviors (visits, clicks, rollovers, touches, downloads, etc.) be collected with the least effort, least cost and greatest quality? How can data collection be as "future proof" as possible, anticipating forms of human/computer interaction that is not yet commercialized?
Let's face it - tags are a drag. They are a time-consuming and error-prone means to the goal of understanding customer behaviors and their experience with our apps, websites and advertising. "Data quality sucks, let's just get over it
" may be the reality today. But it's time to move significantly beyond the unnecessary burdens of data collection which lead to poor data quality, and ultimately to mistrust by the consumers of reports and analyses.
Tag management is good; it has made things better. But is it just a better buggy whip, when what is really needed is a car? Or a hybrid car, when an entirely new means of energy-efficient transportation is needed? Tag management should be today's state of the practice; but it is not state of the art
tools also solve some problems, but create others. Tags are not the end, but one of many means to the end, so it will be necessary to examine data standards, web languages, content management systems, processes, and more.
Only the field of digital analytics can solve this. Collecting customer interactions on websites, apps, media, etc. characterizes digital analytics and distinguishes it from other marketing analytics. The analytical side of things - attribution, testing, segmentation, visualization, etc. – is not specific to digital analytics and not new. The concept of a "conversion funnel" dates to 1898 [Strong Jr., E. K. (1925). The Psychology of Selling and Advertising. New York.] Digital analytics teams have the Sisyphean task of getting marketers, digital publishers, webmasters, IT, etc. to collect data for them. But those folks are in the business of "get the site (campaign, ad, email, etc.) launched ASAP" and see data collection as a necessary evil, not a shared goal.
The broader topic of data management—including storage, access, integration, classification and retention—is also ripe for improvement, but it all starts with data collection.
Need/Opportunity #2: Privacy
How can data on digital media be collected, managed and used in ways that fully and easily comply with the letter and spirit of privacy laws, protect customers and provide value to both businesses and customers?
Privacy is a complex and often poorly understood topic. Technical discussions - for example, about persistent cookies, sessionizing, etc. - are well-meaning, but lead to FUD (fear, uncertainty, doubt) for those outside the field. Regulations are distinct to each country or region, lead to inefficiencies at best, and room for accidental misuse of data at worst.
It is difficult to imagine that even those in the analytics field would be perfectly comfortable with any party that, without explicit consent, knows who you are, what you do, where you are, the names and ages of your children and so on. There is value to be gained, by both marketers and consumers, from collecting and analyzing data. But what is feasible today should scare everyone. This point was popularized, if you will, by the retailer Target knowing
a daughter was pregnant before her father knew. Nefarious websites can easily follow-up anonymous visits with a personal email to the visitor, after triangulating available data and using tools specifically designed for this purpose.
The issue is broader than "online" analytics. Each time you reach for your frequent shopper card at the mall, you help the retailer build a distinct, personal profile of you and your family - not just a demographic segment. Nevertheless, digital analytics needs "a seat at the table" on this topic, if not lead it, given the vast and growing volumes of digital analytics data.
Regardless of one's experience regarding privacy laws and requirements, or the level of understanding of the complexities of data management and the impact on privacy, it seems to be evident there are both ethical and practical reasons to address privacy fully and urgently.
Need/Opportunity #3: Future Talent Pool
How will the availability of digital analytics talent meet the exploding demand? How can university programs be expanded and incoming students encouraged to participate? What is the full curriculum needed to accelerate the productivity of graduates?
It is difficult to find digital analytics talent today for various reasons, among them:
- Demand has remained relatively high, even during the economic upheaval since 2008. This is due to several reasons including: the accelerating adoption of digital marketing; a maturing, continuously evolving industry; and complicated analytical tools, methods and data collection processes. The result is inefficient use of people's time. In the book Web Analytics Action Hero, Brent Dykes brings this home, describing how analysts can be "heroes" the more they are in "Action Land"—adding value and really doing analyses - rather than "Setup Land" correcting data and configuring tools.
- Supply is lacking and variable. A consistent standard of education for the field does not exist, as there is for subjects such as journalism, accounting or chemistry. Digital analytics is an inter-disciplinary field requiring understanding of computer science, marketing, and mathematics, plus "soft skills" such as ability to present, collaborate, and, well, analyze. Undergraduate degrees in web analytics are not commonplace and entry-level analysts emerge from a range of backgrounds such as marketing, IT, or liberal arts, filling in educational gaps on the job. On-the-job learning plays a vital role, but is not an efficient, consistent way to establish a standard plateau of entry-level knowledge.
There are educational needs for others as well:
- Consumers of reports and analyses. Clientele need to understand basic metrics just as they need to understand reports for accounts, inventory or budget. Daniel Waisberg highlighted this point in a recent post on this website, focusing on numeracy for clientele.
- Technical folks, webmasters, etc. need to understand their role in digital analytics, and are better contributors and advocates the more they understand.
- Senior managers need to understand the value of digital analytics to the business, and need help learning how they can be advocates for analytical teams.
But while there are needs and opportunities for all of those who hold a stake in digital analytics, education for "future web analysts" is essential for future effectiveness and efficiencies.
Significant advancement and fundamental change are likely to be found...
- Through partnerships of industry, academia, standard bodies, and government
- Collaborating with practitioners from related areas, from offline marketing to decision sciences to creative designers to you-name-it.
- Using analogs. Identifying fields that proceeded through similar maturity cycles, and examining events that spurred significant advancement in those fields, can inform and accelerate maturity in digital analytics.
- Leveraging global knowledge and points of view.
Significant change will NOT be found...
- In the latest feed on Twitter. Enough said.
- In the industry, as it exists today. Digital analytics agencies have extremely talented professionals with the right intentions. Gary Angel recently commented on his blog "For every client that hires Semphonic to analyze and improve their digital performance, there are three that hire us to audit their tags, build them [sic] reports, or help figure out a strategy. I think the percentages should be reversed." But tagging, custom implementations and proprietary education pay a lot of agency bills today, so incentives are not there for agencies to lead the way. In-house practitioners - in verticals such as telecom, retail, media, etc. - are paid by employers to sell their products and services, not to take on a mission of significantly advancing a field such as digital analytics. Tool suppliers are vital, but they naturally seek their own standards, and hope they become de facto standards. Google Analytics has simplified web analytics to a degree and educated marketers a bit through an easier-to-use-than-most interface. But Google would need to clearly demonstrate it is on a path to a "culture of complicance" before it could be considered a role model to address privacy.
- With the Digital Analytics Association (DAA), as it exists today. Leading the way requires facilitating, and perhaps sponsoring, research into technologies, standards, and methods which significantly advance the field. It will be critical to establish and maintain strong, transparent relationships and knowledge sharing with academia, marketing organizations, standards bodies and governments. While the DAA is an obvious candidate to lead, there is a gap today between its activities and these qualities. With no outreach with higher education, beyond a closed relationship with the University of British Colombia, its approach to education remains introverted and proprietary. The DAA should be applauded for a "code of ethics" and the DAA's European SIG recently published a whitepaper on privacy, albeit only available to paying members. Both are excellent starting steps for member awareness, but not broader awareness nor action.
Optimizing data collection and quality, ensuring customers' privacy and expanding the talent pool are among the many challenges and opportunities faced by digital analytics today. Businesses will not continue to pay premiums for results; key aspects of the field need to move along the continuum from custom toward commodity.
Advancing the field to digital analytics "2.0" or "whatever.0", will require fundamental advancement and change. Those who laid the groundwork and founded web/digital analytics are intelligent, well-intentioned folks and should be applauded for committing significant portions of their lives to the field, getting it to where it is today. But as writer Charles du Bos said: "The important thing is this: to be able at any moment to sacrifice what we are for what we could become
- Radically Rethinking Web Analytics
- The Ultimate Definition of Analytics
- How To Build Your Analytics Confidence