comScore’s annual Super Bowl research, which included both pre- and post-game surveys, measured various aspects of media use and ad preference during the big game.
Denny’s had the third best brand improvement score of 39 percentage points. That should be good news, but the extra attention crashed their site. This article by Download Squad states, “Marketers need to learn to anticipate promotional effects on servers, end of story. Don’t write a check your host can’t cash, basically.”
Denny’s site wasn’t the only one overwhelmed with traffic, Vizio’s ad on the big game was specifically created to drive web traffic through a “free TV” contest. Yep, dead under the load.
So, the question that comes to my mind is, “Don’t big companies with big ad budgets also invest in good stress testing?”
Apparently the answer must be “no”. Either IT did not determine the maximum volume of users their sites could service, or the marketing department blew it by underestimating the traffic. While both could be true, my guess is that marketing forgot to tell IT. Too often marketing does not include IT in campaign planning or execution. And in this case, it backfired.
Transparent Uptime points out that QuickBooks online went down for several hours – leaving small businesses without access to their financial records and without the ability to process credit card transactions.
QB gave no reasons for the outage. But as web application developers or testers, can’t we easily speculate that stress testing was overlooked or not thoroughly executed? Stress testing gives us the ability to identify web app weaknesses before they affect the customers by:
- Generating virtual users until a site fails
- Determining breaking points of infrastructure
- Disabling components such as network devices
- Creating bottlenecks to assist in performance tuning
I submit to you that these companies did not place enough value on stress testing. If they did, they would not have had these very visible failures that almost certainly cost them quite a bit of money. Additionally, no one can accurately calculate the true future value of lost customers.
I was seriously considering moving to QB Online, but now I’m probably going to stay with my current solution. Wonder how many other folks like me are steering clear because of the lack of priority for stress testing?