Ad Ops and Cloud Computing

Depending on how much you work with Google applications last week, you probably found yourself unable to access them or at least hindered by how slow things operated. #googlefail as it has been called set off a flurry of articles about the potential dangers of cloud computing and reliance on online applications. I think many of us in the Ad Operations world are well aware of the problem with online applications. Who hasn’t suffered through a network outage and found that the only work that can be done is cleaning one’s desk, filing expense reports and finally listening to our voicemail. But what about cloud computing? How much do Ad Ops people need to know about cloud computing?

I think it is fair to say that until recently cloud computing wasn’t a concept that Ad Operations has had to know about. But over the past 12 months, a number of new companies and services have emerged promising to do things that I frankly had given up hope of ever seeing.  All these companies tout cloud computing as the means to quickly and cost effectively process data like never before. Yieldex is an example of one of these companies and has shown that cloud computing can move the bar on what we can expect for inventory analysis and reporting.

As more companies start to offer new capabilities powered by cloud computing, it becomes very important to understand the risk/rewards of cloud computing for your department. I think it’s a mistake to look at last week’s events and simply dismiss cloud computing. My recommendation would be to educate yourself about cloud computing and then ask the following questions:

Where does your IT department stand on  technologies like cloud computing? Consider this one of those opportunities to engage with your IT department about something they are probably quite passionate about – one way or the other. Most likely they’ve investigated whether or not cloud computing makes sense for their own usage. Realize they live and breathe this stuff and are held accountable for their ability to support it. Their concerns should help you establish your service level agreement with any vendor and hopefully prevent any future “I told you so” moments.

What is your tolerance for system/process failure? The simple answer of course is “no tolerance”, but this isn’t practical. As Ad Operations teams work with more systems and processes, it is important to establish an understanding of what can and can’t go down and for how long. Obviously if your ad server stops delivering ads, you are out of business. That’s mission critical. For other systems you should get a sense for how long you can go without. For example, you might be able to go without inventory reports for a day, but maybe not a week. Automated reports might be something that you can go without longer if you have a manual method of pulling the data. This analysis is going to help gauge your department’s ability to accept risk.

My final point is that these two recommendations are not specific to cloud computing and could apply to any technology. I can’t speak to the stability of cloud computing – but like everything else I would turn to the experts and I would measure the risk/reward of what a vendor can provide me. What I can speak to is that a number of solutions have started coming out recently and publishers and agencies who have long waited for solutions are quickly jumping on the bandwagon. Instead of waiting out the recession, companies are pressing forward. Ad Operations needs to help lead the way.

Rob Beeler is Vice President of Content and Media for AdMonsters and has worked in Ad Operations for over ten years. Rob started attending AdMonster events in 2004 as a member and will be in New York on May 21st for Ad Ops 360°