Are You a Web Technologist or a Web Analyst?

• Tweet
• Buffer

I took time out of the office two weeks ago to attend the London leg of Adobe's Digital Marketing Summit. It was a lot of fun, provided food for thought for my team's development, and was a great chance to hear stories from the data coalface.

Back in 2008, when I first started going to these shindigs, the attendee list was dominated by web analysts and e-commerce managers just trying to make sense of the digital data they collected. Omniture was trying but failing to sell itself as a digital marketing expert but was still impressing with its dominance in the web analytics world. Now in 2012 the summit definitely isn't just an analytics event. Adobe has made good on the promises of Omniture and showcased product suites that design, deliver, and measure digital experiences for your customers. Five years ago I used to only chat to other digital analysts, but this year I bumped into creative directors, heads of technology, and CMOs.

So although the summit is less focused purely on analytics, it is still a hugely valuable event for the measurement community. Here are three "P" summit takeaways to think about if you are an analyst:

1. Power user. After reviewing all the notes I took during the two days, the one that stood out for me was "Be good at your job, not just good with the software." I wrote this down while listening to a talk by First Choice, a U.K. travel firm that sells package holidays, around optimizing its paid search using Adobe's Efficient Frontier. The company shared wonderfully impressive uplift figures of +77 percent customers from a +35 percent increase in cost base. However, the big improvements it made seemed to be driven not by the tool that automated its search buying but the strategy developed by its business analysts. The company had a great mix of power users of software but also power users of PPC and SEO strategy.
   
   If you want to be a successful analyst you are going to need to learn more than just how your favorite tool works. A great web analyst doesn't know only how web analytics technologies work but how the web works. Start badgering the other digital teams in your company to give you a 101 on what they do. You'll get context on how your numbers are used in the wider business and become more valuable to future employers.
2. Privacy. One of the sessions at the summit was a panel discussion on "Privacy versus Personalisation." The great and the good of digital Britain discussed the fine line between using personal data to deliver a tailored experience and the risks in becoming seen as "Big Brother" spying on your customers. They talked about the responsibilities of both brands and governments around how personal data should be used in the future. I'm not being disingenuous when I say they could have subtitled this: "We really don't know the answer to this one." You only have to have a passing knowledge of how the governments of Europe have bungled privacy legislation to understand what a headache this is going to be for the online world.
   
   So what to do? This may be the most boring recommendation you are ever likely to read but you need to stay on top of the legislation in your region because the privacy decisions made this year are going to affect the way you analyze, target, and communicate to customers for the next decade. Get to know what data your company collects about customers and how that ties in with your government's recommendations. Use surveys to ask your customers their views. And with this knowledge you can target more effectively and safely.
3. Predictive. I'm a sucker for a break-out session that uses the words "maths" (and yes, America, I do means maths) or "algorithm," but I rarely come out of these sessions anything but disappointed. However, Adobe spent a lot of effort this year talking about its predictive analytics tools and that gets me excited. If we as digital analysts can't talk about what the key drivers are in the behaviors we see (by simple correlation analysis) or aren't able to prove that campaign A beat campaign B (using simple significance analysis), then what are we here for? We may as well be guessing.
   
   So what to do if you are a web analyst who hasn't done this type of work before? Take a stats course or face being left behind. If you don't know how to correlate, complete simple regression, or spell the word "cluster" then it's time to start calling yourself a web technologist and not a web analyst.

Finally, a bonus tip from Arianna Huffington's New Media keynote that we'd all do well to remember: "Get more sleep."