An Interview with Judah Phillips, Senior Director of Global Site Analytics at Monster Worldwide

1.  I have had a whole bunch of conversations lately with analytics executives that complain that they only have number crunchers and report generators on their teams. What they all say they need are analysts… people that generate insights from their numbers and reports. The jobs that they are looking to fill are Senior Web Analyst jobs… 3-5 years of web analytics experience. Is it asking too much of a person with three years of work experience to understand the tools, understand the metrics and understand the business well enough to helicopter up and down between the numbers and the big picture? In the old days that was the job of C-level, or at least VP-level executives…It depends on the business, its complexity, and the individual analyst. I've worked at some relatively big companies in my life, and, in my experience, large companies with multiple lines of business and revenue streams, various levels of decentralization/silo-isms, legacy and new tools, emerging and historic business challenges and requirements, and difficult data can't expect to hire an analyst to "helicopter up and down" and be successful out of the gate. It can take months for a person just to understand complex data, the web software development lifecycle, and the intricacies of the business, its competition, and its unique challenges. On other hand, I've also worked at or with smaller companies that had very focused business models and a blank data and tools slate. It's much easier in those smaller companies with less "baggage" for a person to look at simple, elementary data, like bounce rate, and say "you need to do X, in order to improve Y" and, thus, help optimize the site and drive the online business. At the end of the day though, an analyst can't have much of an impact unless the business defines the "business questions" they want to answer, identifies why the site exists, and determines what their goals are for each product, experience, flow, and the site overall in order to give context to an analysis.I do think it's fairly insane for a company to expect to hire a person with 3-5 years of just analytics experience and expect them to tell a line-of-business executive, who typically has more relevant business experience, more education, and historic knowledge of their company and current challenges, the exact correct decision they should make. In fact, I think the best thing an analyst can do is make solid, data-driven recommendations that can help provide guidance and act as one of the many critical inputs that an executive needs to consult and consider when making a choice. The ultimate decision on what to do has to be made by the person - with the positional power, political clout, organizational influence, and adequate resources -who can actualize it. That's one of the reasons why I think the idea of incentivizing analytics professionals on conversion rate improvements makes little sense. Analytics has a very limited ability (if any at all) to actually implement the optimizations they recommend. Sure, if they are using a site optimization tool, like SiteSpect, they have a much better chance to do so because the output of a powerful tool like that is the recommendation and the recipe for success. But if you're just talking about behavioral data and applying technique to make it meaningful and insightful, and the business doesn't follow the recommendations, there's little analytics can really do. We'll just continually point to the data and the recommendation based on the data-driven insight until the person with the power decides to take a chance and listen.2.   If you are running an analytics team in an ecommerce business, how important is it to hire people onto your team out of an ecommerce environment (vs. a media company, a B2B site, an agency, etc.)?I'm a fan of different perspectives and giving people room to grow. If a company already has strong e-commerce talent and that's it, then hire outside of e-commerce and have the talent train the noob. The alternate perspectives the new hire from an agency, b2b, or media company can apply to the data evaluation, based on their past experience, may be eye-opening. The reality is that many of the same web analytics concepts apply whether you’re talking about a content-driven site or an e-commerce site. You have funnels, forms, abandonment, fallout, pathing, keywords, referrers, conversions, events, campaigns, and revenue whether you are selling books, registering people to download whitepapers, or signing up subscribers for newsletters. Search is search. Advertising is advertising. What's different on e-commerce sites from media sites is the shopping cart, transactional elements, promotions, and the specific data nuances of the particular company and its products.3.   A lot of the jobs that we work on require specific experience with the toolset of the company. Is that a mistake? Would they be better off, in general, hiring the “best athlete” and giving them a few weeks to transition to their tools?It depends on the tool and what you want the person to do with the tool. And it depends on the person and what you want that person to do for the business. If you are talking about hiring an analyst to pull data and tell stories to the business in PowerPoint, I don't think it matters that much if they have WebTrends experience or Unica NetInsight experience. They'll be able to figure it out. But if you are hiring an analyst to tag and configure a tool with all its glorious complexities - let's say, Omniture SiteCatalyst - then yeah, you probably want a person who has some previous experience with the tool doing what you need them to do - especially if you expect quick results. However, that really depends on the person too. Some people pick things up faster or slower than others do. Some tools are just more complex than others are too. For example, Omniture Insight compared to Google Analytics. I wouldn't hire a person who has just played around with utm_this and utm_that in simple JavaScript page tags (or has just written a tagging specification and handed it off to development) to architect the schema for Omniture Insight and do a bunch of data transformations. You just couldn’t expect anything useful in a "few weeks" especially if the business expectation is that the new hire will integrate data from other sources whether internal or third-party into Omniture. In most cases, it will probably make sense to hire someone with specific tool experience. But smart people are just that. They can learn new stuff if the business gives them time to do so. The larger question to answer when taking a chance on someone with little relevant tool experience is whether the business has a culture of learning and investing time (and thus money) into a new hire. Or is the culture all go-go-go-do-do-do with no time for much else but project delivery. Learning a new tool takes extended periods with no-interruptions and focused training. At the end of the day, it's different horses for different courses.4.   I know a few analytics executives that have moved to line management roles within their companies with great success. Once an organization becomes truly “data driven” this seems to be a natural progression. Could you see yourself moving in this direction in the future? Or are do you see yourself in analytics organizations for the rest of your career?I'd like to think I see myself moving in that direction. Web analytics is just one area in which I am experienced. In grad school, I was really, really good at derivatives and risk management (puts, calls, options strategies, futures, swaps and so on) as well as all things marketing. I have a Masters in Finance and an MBA. I’ve helped to run a small start-up and done a bunch of consulting. I like to think I am a good writer and public speaker, so I can certainly do more than web analytics and would enjoy seeing my role broaden in the future into a larger business-focused role with an analytics emphasis. I'd also be a happy camper if I ended up in an analytics organizations until I retire.Now allow me to digress... The concept of a "data-driven" organization is easy to say, but it may be a bit polyannaish to believe they actually exist. And yes, I've read Davenport. Lines of business can certainly be data-driven and so can entire executive management teams, but the entire business across all functional units - well maybe. Decisions will always be made based on experience, intuition, and feelings, not just exclusively on data. That's just human nature and reality. I think a lot of organizations may believe themselves to be "data driven," but the reality is that they are probably also "data reactive," "data fearful," and "data self-promoting" too. I have seen instances where people didn't care about the data until after the fact and then used it to plan the next site enhancement instead of looking at historic data to formulate a strategy before the first enhancement. I've seen people totally freaked out by the data because it promoted a level of transparency that was very uncomfortable. In other instances, I've seen executives who just want data that looks good so they can show how awesome they are at their jobs. My good friend Bob Page - who just landed a sweet gig as a VP at eBay - claims something smart. He says "all data is political." I believe that, which means all "data driven" organizations, wherever they exist, are actually, by corollary, "politically driven" too...5.    What would you rather have, a stud analyst with exactly the right skill set, mind set, and experience that works remotely or a B-level analyst that can sit in the office with your team and your clients?It depends on the company, what it needs the analyst to do, and what types of flexibility it allows. If I had my druthers, I would pick the remote "stud" or "studette" analyst and set the expectation that you'll be coming on-site every so often and as required to meet with the team and your business stakeholders. I remember in my first job at this awesome little startup there was this manager. He was once asked would you rather have a person who got the job done in 40 hours and went home or a person who worked 80 hours a week and was at the office all the time, but never got the job done? He chose the 80-hour a week person. I never understood his answer, and never asked him to explain, because I was too busy doing my job at work and enjoying my life at home. Your question feels similar.6.   There are some amazing things happening in the broader world of digital media measurement. The digital display world has become fascinating with video and ad networks and advances in behavioral targeting (or whatever the PC term is for that right now); social media and mobile measurement are evolving; search analytics has all kinds of new possibilities and implications because of social media… Which one do you look at or read about and say, damn, if I had time I would really like to become an expert in that…The stuff going on with social media right now is pretty cool. It's really mainstreaming and the shark (to jump) is nowhere in sight. I'm really looking forward to my pal Jim Sterne's new book, Social Media Metrics, which is coming out in early April. If it's like any of his other books, he's going to revolutionize the space, like he did with web metrics. In my spare time, I've been self-teaching myself a lot about social media listening platforms (and how none of them do everything a business needs and many of them just don't do a good job at all despite what the advocates say). I'm rather enamored with video measurement, but don't really have much of an opportunity to indulge in that interest as deeply as I would like. I know the smart guys and nice folks over at Visible Measures, and, boy, are they doing some really, really cool stuff with video measurement and engagement. The mobile space is interesting to me too, but it's very much like traditional web analytics on smaller screens with some absolutely crazy data collection, sessionization, and visitorization challenges. Targeting is interesting, but it's really just (and I use the word "just" very, very, very lightly) an exercise in advanced audience segmentation and automation serving ads to cookies. Quantcast has the potential to rock the world with their targeting platform. Ad networks are interesting to me, but a more compelling topic these days in the media space are ad exchanges and how you optimize your media spend with them. I guess if I had to pick one, I'd say mobile social video targeting across ad networks and exchanges. Just kidding. I will admit one thing I'd love to be an "expert" in is predictive modeling, which would apply to any of the data generated by any of these "happenings" in digital media measurement.