Learning the Lessons of Data Analytics in the Dot-Com Bust “Classroom”

by | Aug 25, 2022 | 2022, Agile Data Governance, data architecture, Data catalogs, data culture, data ethics, data value, Data-driven cultures, DataOps

data.world CEO and Co-Founder Brett Hurt is sharing his thoughts on the theme of “People + Data” on a weekly basis in advance of our September 22 data.world fall summit.

Brett introduced this theme in Part One of this series, where he explored an idea implicit in data.world’s DNA.

 

Understanding the past and present through numbers, math, and analytics is a skill I pretty much picked up as a child, thanks largely to a brilliant mathematics professor who was also my grandfather. Using knowledge derived from numbers to navigate the future, however, was an aptitude that came later — born of hard times made manageable with data.

The sum of this skills evolution is my subject today because it’s never been as relevant to the challenges that businesses face as the winds of commerce blow ever colder. Let me explain.

My grandfather, a professor at the University of Texas at Austin, taught the most advanced mathematics classes his entire career and he regularly practiced with me. I fell in love with math before I could ride a bicycle. Sure, his ceaseless testing was sometimes annoying – he was always testing me way above my grade level. But my love for my grandfather and his lessons was to cement my lifelong love for logic and facts. 

I was coding at age 7, over 40 hours a week, and got my first degree in Management Information Systems at U.T. Austin, which had a strong bent in the practical application of data to solve real business problems, known today as data science.

Thus it was pretty natural when I was well on my way to my second degree, my MBA in High-Tech Entrepreneurship from The Wharton School, that I turned to data and analytics as my wife Debra and I launched our first eCommerce business before even graduating. This was a retail site named BodyMatrix, which sold sports nutrition products like PowerBars online. While I programmed the engine myself, good analytics tools were hard to come by in an era really limited to the products of WebTrends. So I invented my own. 

The Birth of Coremetrics

By focusing on analytics and customer behavior to constantly improve the customer experience, Debra and I were able to boost our conversion ratio of browsers into buyers from 2% to as high as 3.8% – effectively doubling our online sales with the same level of inputs.

As I told my parents, both of whom were retail and direct-marketing entrepreneurs: “I have a bigger brain than you did in your physical stores, because through the use of technology I can remember everything and look at almost any aspect of how our online business is performing!” My Wharton professors took notice too and helped me evolve these metrics to be better.

This early epiphany born of lessons at my grandfather’s knee blossomed into full blown devotion to data and analytics after I founded my first major company, Coremetrics, in 1999. The mission was to empower online retailers, the importance of which we learned as trial by fire when the unforeseen dot-com bust began in 2000, ultimately erasing $5 trillion from the market caps of U.S. companies.

Sadly, we were too late to help many of the 100 customers for our new Software as a Service (SaaS) based business, which was one of the first. Before the turmoil – compounded by recession and the trauma of 9/11 – was over, 97 of our 100 customers went out of business. Much like the times we are currently living in, the proverbial sh*t really hit the fan.

Learning to Optimize Paid Search ROI

Still, Coremetrics was able to win Walmart, mattress retailer Select Comfort, and pet food giant PETCO, along with hundreds of other traditional retailers. All were hyper-focused on online success and we were able to help them amplify their market and multichannel power compared with their struggling dot-com-only peers who didn’t have the power of Coremetrics early enough. The results were really extraordinary. 

For example, I remember that Select Comfort wanted to get a leg up on its competition, most of whom were outbidding them on Google’s paid search. But it was a conundrum because most of Select Comfort’s high-purchase-price transactions were fulfilled via their call center sales representatives, not via their eCommerce site. Remember, these were the earlier days of eCommerce, back when Amazon.com was only around 4% of US online retail, largely dismissed, and itself almost shuttered by the dot-com bust. There just weren’t as many people willing to shell out thousands of dollars online on a mattress without speaking to anyone live.

So, Select Comfort came up with something truly novel. They generated unique 1-800 numbers to attach to each keyword buy. They would then have us at Coremetrics capture these inbound numbers to each prospect visiting their site. This required a lot of data engineering with multiple Select Comfort data sources (wish we had a data catalog back then), but as soon as a prospect registered, we could tie their call-center identity to their website identity. Eureka! We could now measure the ROI (return on investment) of their paid search in driving on or offline sales! None of their competitors had this capability, and they quickly outbid their rivals. 

Data Analytics

“Select Comfort could confidently outbid their competitors because they knew the ROI of their efforts.”

I’m sure their competitors thought Select Comfort were idiots who would spend themselves into oblivion buying keywords. But Select Comfort was laughing all the way to the bank, with a 250% greater ROI measurement for their paid-search ads. Select Comfort could confidently outbid their competitors because they knew the ROI of their efforts, not just online but offline (i.e., what’s called “multichannel” measurement or impact). And our analytics showed that more than half of the total sales driven through paid search came through the call center. 

In addition, the AOV (average order value) for web-influenced call center transactions was 39.8% higher than orders transacted purely online, due to the efforts of cross-trained agents in the call center. For its day, that was truly novel – it was the beginning of a technique that would change eCommerce measurement forever, especially for multichannel retailers. 

Riding the crest of this mathematical and data wave, the Select Comfort team that developed this analytic technique were celebrated at virtually every eCommerce conference. And they were promoted inside of Select Comfort too! Boosts to their salaries and bonuses yielded real ROI for them personally.

Inventing the Future

The satisfaction of this type of data and analytics breakthrough, especially after the fallout of the dot-com bust, cannot be understated. And, to an analytics and math geek like me, it gave me and many on our team a real sense of purpose. The Select Comfort team members involved are friends of mine to this day (hello, Carol Ott!) – we were shaping history together. It was People + Data at its best.

We were not just inventing the future of how eCommerce would be accountably run. Along with others, we were inventing the future of the economy, society, and humanity.

Next week in part three, I’ll discuss what one of our customers did at Bazaarvoice, during the Great Recession, to really distance themselves from their competitors.

data.world summit fall '22 is Coming Up!

we’re excited to announce our fifth virtual summit on Thursday, September 22, 2022. This fall, we’ll focus on cutting through the noise to spotlight the role of social interaction and collaboration in the age of data. Register now!