Blog post

The Evolution of Marketing Measurement

By Christi Eubanks | September 18, 2015 | 0 Comments


“It would be nice if all of the data which sociologists require could be enumerated because then we could run them through IBM machines and draw charts as the economists do. However, not everything that can be counted counts, and not everything that counts can be counted.”

– William Bruce Cameron, 1963 (and Albert Einstein, according to some people)

If I hadn’t looked up the date, I’d insist that quote was about marketing analytics, a discipline that would not come to exist for another 50 years. Prescient, really.

Counting Hits

Depending on what you count as its inception, web analytics is around 25 years old. Web site measurement – and the need to justify marketing investment – is as old as the public web itself. For the uninitiated, the original method was server log files, which were used by the IT team to make sure the website was up and running. Those log files could also tell us how many visits the site was getting (how many times the server was called), which could help justify the cost of that web presence. The original web analytics vendors realized the market for this and began selling analysis of log files. That analysis was little more than counting “hits,” but it was a start.

Most of original web analytics vendors don’t exist anymore, but a few do. You can tell how tenured your web analyst is by asking her what her first tool was: Webtrends, Omniture and Web Side Story (now Adobe Analytics), Coremetrics and Unica (now IBM Digital Analytics and retired, respectively) warrant major analytics street cred [if you care about that sort of thing].

Around the same time web analytics was gestating, specialist firms and agencies began to commercialize techniques developed by econometricians to estimate the incremental impact of different marketing (TV, radio, print) tactics on sales. They called the practice Marketing Mix Modeling (MMM), and it would go on to be loved by CPG companies and adopted by many Fortune 500s, the only companies who could afford the artisan service and who had good historical sales records for inputs. These guys were doing the stuff that counts, the real math. Unfortunately they wouldn’t cross paths with our web analysts until much later.

Counting Visitors

In 1995, a developer at Netscape wrote Javascript, providing an alternative way to collect data from websites via tags (also known as web bugs, beacons, pixels) which could communicate through the browser with 3rd party servers. Throughout the ’90s IT people debated the merits of log files vs Javascript. Javascript tagging is not above criticism, but you know how this ends. Javascript won, transforming the web analytics landscape and laying the groundwork for modern marketing analytics. Why so transformational? (1) Vendors could collect data on the company’s behalf and marketers could perform analysis without needing access to their own servers. (2) User-level data “unique visitors” could be accurately captured and user preferences saved from one session to another. (3) The web page could display content (like ads) and share data with other 3rd parties.

Counting Impressions, Then Clicks

The internet had reached 16M users by ’95, and they were just dying to be monetized. Web portals and popular sites began offering ad placements directly to brands (AT&T was one of the first). Doubleclick launched their ad management service to help advertisers optimize their ads at scale based on performance (all enabled by – you guessed it – Javascript). By 2000, Google had moved out of the garage and decided to pay for their new building (and new chef) with contextually relevant text ads. Ads were originally priced on a CPM basis, but after the bubble burst in 2000, advertisers were looking to justify their spend in this new medium with tangible returns. Google didn’t invent PPC pricing (Yahoo was already doing it), but they perfected it, and in doing so, changed the course of web measurement again as we moved beyond eyeballs to actions.

Things were going really well, as you can imagine, for web analytics providers, right up until everything imploded and they weren’t. The bust of 2001 forced industry consolidation/natural selection, leaving only a handful of key players standing.

In 2005, Google acquired Urchin (that’s the “u” from “utm” tags) and leveraged its technology to launch Google Analytics…for FREE. The VP of product at Google said in the press release that “this technology will be a valuable addition to Google’s suite of advertising and publishing products.” Google brought web analytics to the masses and married it to digital advertising analytics. It’s no wonder that web analysts fell hard and fast for performance marketing. It was so easy, it was like crack.  Just give credit based on the last click before a purchase (or other desired conversion action). That last click was often search, which worked out pretty well for Google. We called this new measurement methodology attribution.

In 2007, Google acquired aforementioned Doubleclick, which was not only meaningful from an advertising industry perspective, but also from a measurement perspective.

Double Counting

2007 was also the year Apple released the first-gen iphone, changing consumption behaviors, challenging the state of state management, and throwing a big wrench in user-level tracking.

Meanwhile, brands were branching out from websites to “owned” (really rented) presences on social networks like Facebook, Twitter, and YouTube. Just as they were getting comfortable in their new social digs, the big three decided they weren’t in the brand charity business and launched ad products. Digital marketers needed data to beg for bigger budgets, so the web analyst’s job got a lot messier and a lot less web site focused. We were still counting, just counting a lot more things (sometimes by hand). And then we were adding them all up and showing results that were much too good to be true.

Counting Clouds

Google crushed the market, with around 80% share in the web analytics category. Omniture, one of the major enterprise analytics vendors (they had gobbled up competitor Web Side Story in 2007) sold to Adobe in 2009 and became a fundamental piece of what would come to be called the Adobe Marketing Cloud. In 2010, IBM purchased Coremetrics, another of the original web analytics providers. Pure-play web analytics solutions were a thing of the past.

Counting What Counts

All of this brings us to the not-so-distant disruption of the twenty-teens. Longstanding industry group, the Web Analytics Association, changed its name to the Digital Analytics Association in 2012, signaling to everyone, who hadn’t caught on yet, a new era of measurement that transcends web sites.

HBR told us that data science was sexy, so not to be left out, we analysts dusted off our college stats skills and added “learning R” to our vision boards. Vendors added advanced analytics capabilities to enable things like segmentation and regression analysis. We got smarter about attribution and started throwing around words like “incrementality.”

After growing up frienemies – they never gave each other any credit [get it?] –  digital (nee web) analytics and advanced modeling (MMM) finally got together and bonded over their shared interests, namely: predictive analytics, helping marketers measure ROI and trying to figure out what to do with social media. Google’s acquisition of Adometry and AOL’s acquisition of Convertro (now both part of Verizon) in 2014 made the relationship official. More traditional MMM vendors also began offering solutions to support digital inputs and leverage customer-level data.

With the addition of offline data and the focus on creating business value, calling it “digital analytics” just feels too limiting.

So here we are in 2015, introducing a new discipline, a new tool category, and a shiny new Magic Quadrant. 

Marketing Analytics is the practice of harvesting, managing and analyzing online and offline data to maximize marketing effectiveness and optimize return on investment. Marketing analytics drives marketing strategy through the full lifecycle – from planning, to activation, testing and measurement – emphasizing near real-time actionability and prediction over backward-looking description.

We’ve finally moved beyond counting and are focused on what really counts.

Clients can read the MQ here. You’ll see some familiar players and maybe some surprises.

The Gartner Blog Network provides an opportunity for Gartner analysts to test ideas and move research forward. Because the content posted by Gartner analysts on this site does not undergo our standard editorial review, all comments or opinions expressed hereunder are those of the individual contributors and do not represent the views of Gartner, Inc. or its management.

Leave a Comment