Case Study: Data Collection Tools

http://blog.vendoservices.com/vendo-blog/2015/09/14/case-study-data-collection-tools

Screen Shot 2015-09-04 at 10.03.50

(This is a business case study. It will be used to guide discussions during the session: “Data Collection Tools” at the Vendo Partner Conference in Barcelona on September 16th.)

Intuitive-driven-vs-analysis-driven

“Sorry…but your code is the problem,” replied the guy at the hosting company. “You know what you need are our new upgraded servers”. That wasn’t what wanted Mark to hear. He needed a solution and he wasn’t getting it. And time was running out.

His costs were rising fast. He didn’t know if he could keep growing. He was starting to feel like he worked for his hosting company…instead of the other way around. “Got a problem? It’s your fault. Fix it,” wasn’t the approach he wanted from his supplier. Mark smiled to himself and thought, “Ok…help isn’t coming from there…we’re going to have to wrestle this to the ground ourselves.” The first step? Getting the data he needed to pull out real insights.

This wasn’t Mark’s first time wrangling with hosting. He’d been through it all. He started serving content online in the 2000. He’d witnessed several distinct phases in the development of online content delivery. In the earliest days he could only provide a comparatively poor experience next to big brands like Disney, Warner and MGM. They had massive files and massive budgets. His business had much lower dollar value transactions, much larger audiences and a greater number of transactions.

For the better part of two decades Mark has been locked in a struggle with hosting companies to deliver content faster, more streamlined and at a lower cost. That has meant developing better processes, hardware and piping. “Every 4 or 5 years there seems to be a big improvement in rates, availability and speed,” Mark recalled. Major improvements followed specialization in DNS, IP routing, load balancing, CDNs and cloud based systems. Before these breakthroughs there would often be intense cost increases as demand outstripped then-current technology. “At one point we were paying $300k a month. Tech changed and a few months later we were paying a fraction of that amount,” said Mark. “Of course, the YouTube era changed everything. Demand skyrocketed and the market responded: more pipe was laid, hardware got cheaper…but it still couldn’t keep up with demand so new technologies and coding structures were being built.”

Today Mark’s advertising technology company sees peaks of 2 billion ad impressions per day. As he grew he added hardware until the costs got out of control. Usage wasn’t the problem, it was the amount of connections. Costs kept rising. He knew that there must be a “new normal” for hosting…but he didn’t have the data to know what it was.

“I felt lost with hosting. It’s not my expertise but I had to find a solution to optimise our servers and get my cost down,” Mark recalled. “Our developers collaborated with the developers at our hosting company. They combined the methods they had learned. Together they picked apart our code and began optimizing. That reduced our costs by 30%. However, I wasn’t happy there. Our internal developers kept on saying they could bring it down further, ‘Let’s change hosting companies, let’s get with this hosting company and our costs will be 70% less,’ but I wasn’t sold. They didn’t have the organizational skills to make a hosting move and we all know the risks.”

Mark had been through too many revolutions in hosting to believe that this was the best they could do. He knew he needed to optimize further. He sought out a new lead developer who had experience in cutting edge tools and access to real performance data from similar companies. “In less than a month the new developer had brought on and incorporated new database, coding, and hosting structures based off of those used in larger freemium based technology companies. Our hosting bill was cut yet again by 70%, and our speed and reliability increased by 30%,” added Mark.  

Mark had been getting data and insights from people stuck in the recent past. Had he trusted it he would have become obsolete. The technology is moving too fast. His experience of multiple hosting revolutions told him that he could do better so he sought out fresh data and fresh insights. Mark concluded, “I learned that sometimes I need to look outside our organization, push our internal team to constantly look outside themselves for new benchmarks and new technologies, strategies and ways to compete.”

Mark had to go outside to access to the data he needed to get real insights. Questions for discussion: How am I collecting my data and how does it help me pull insights out? When do I need to change the way I’m looking at my data?

Leave a Reply

Your email address will not be published. Required fields are marked *