Note: This post has moved from Leapthree.com to Ayima as part of the 2018 acquisition.
This post was partially inspired by Avinash’s post on the path to web analytics glory and partially as I have been frustrated by comments there and previously that you must have SQL skills, statistics skills, used these tools, done those courses, etc in order to be a real Web Analyst. I don’t believe there is a definite list of skills required. But I do believe there are attributes required, in line with my belief that web analytics, done properly, is an art not a science.
I strongly believe you will always be able to tell a true web analyst by their curiosity (the ninja as opposed to the reporting monkey). They have a desire to explore data, to check one more report in case that provides the answer. Good web analysts have a thirst for knowledge, displayed by constantly asking questions and striving to learn more. On the downside, they are occasionally at risk of getting lost in the data.
This is the ability to talk the language of the customers that you deal with. For the technical web analyst, they need to be able to speak the language of developers and designers to explain what code is needed and how to get it to work. The business web analyst needs to be able to speak to marketers, sales people and senior management, to make complex ideas simple and to persuade them to use data to make smarter decisions.
It is a bonus if one person can speak to all groups of stakeholders but, just as it is with actual languages, people who can speak more than 3 languages are rare.
Analytics is typically associated with very scientific and logical processes where there is a single answer for each question. That may be the case in the finance world but it definitely is not the case for web analytics. Instead web analytics requires a creative approach to problem solving to complement the logical thought processes.
Web Analytics data gives the “what”, web analysts need to be able to think of potential “whys”, to create hypothesises that explain the data and to make recommendations for improving business performance. It means that when the first three approaches to solving a problem have failed, possibly about how to evaluate a certain visitor behaviour, you are able to think outside of the box and give three more ideas.
Coincidently, Neil Mason has just written a post for ClickZ on the need for right brained creatives in the web analytics world.
There are numerous techniques and approaches available for turning raw web analytics data into actionable business intelligence. These range from statistics through to numerical reasoning & segmentation and further on to SQL and data mining. But at the end of the day, these are just tools. And tools do not solve business problems or improve business performance. They are just tools. It is how well the Craftsman (or Craftswoman) uses the tool that leads to success.
Faced with the same problem, each web analyst will use the tools they are most adept with. They don’t need to be able to use all of the tools, as long as they can get to the required end point. To explain the ‘C’, the tools and techniques that each web analyst uses defines their Craft.
I have very strong numerical reasoning and creative problem solving skills. I can’t write website code but I can conceptualise what should be captured and I do a great configuration in Google Analytics. No knowledge of SQL at all and I have forgotten more statistics than I remember. When my Craft is not sufficient to improve business performance, then I will learn how to use other tools.
This to me is the most underrated attribute and yet one of the most important – basic common sense. A good web analyst is able to apply common sense to their work, ensuring they are using their time where it is most valuable.
Examples of applications of common sense include:
- Ignoring a drop in traffic over a weekend as it was clearly affected by the first sunny weekend of the year.
- Identifying obvious issues with a website like internal search functionality where you have to click through to a new page before you can enter the search term.
- Knowing when data is too good to be true and therefore disbelieving it.
- Ignoring minor differences in data e.g. between web analytics revenue and actual revenue.
- Keeping things simple where appropriate e.g. not tracking every single navigational click on a website as this adds no value.