Automate your Analytics with a Virtual Web Analyst
A few months ago, I was invited to participate in the beta test of a new tool marketed as a virtual web analyst. Andrew Dempsey reached out to me and introduced me to Needl Analytics. I spent some time with the tool and had it analyze data for some of my clients, and I also spoke to Andrew about my experience with the tool and my suggestions for improvements. While Needl Analytics isn’t perfect, it’s very impressive for a tool in the beta stage.
What is a Virtual Web Analyst?
The ‘virtual web analyst’ is actually a pretty clever description. The thing is that most web analysts often find themselves doing the same things (or slight variations) again and again. Without getting into too much detail, a lot of the work with web analytics is often focused on a few things. One of those ‘things’ is to identify high or low performing parts of a website’s traffic. And we often do that by combining different audience, acquisition and behavior dimensions with conversion metrics. That allows us to identify segments (or dimension combinations) that need improvements.
So, if you’re objective is to improve your paid advertising targeting, you might want to look at traffic and conversion from female mobile users in specific age groups. Why? Because most advertising platforms allow you to target people using those specific dimensions.
But in order to find the most relevant and promising segments, you need time. Time to setup custom reports or data exports to Excel, so you can do the dimension combinations. And then you’re able to score them and prioritize them. In some cases, you have to do hundreds of combinations before you find the right ones.
As such, many tasks in web analytics can be time consuming. But what if you could automate that? Enter the virtual web analyst.
Meet Needl Analytics
This is exactly where Needl Analytics comes into the picture as the virtual web analyst. While Needl in fact is a machine, Needl is also a web analyst. Mostly because Needl does some of the things that ‘real’ (what does that even mean today?) web analysts do. Which is crunching your Google Analytics data by combining the many available dimensions with (mostly) your conversion metrics.
Needl proceeds to score each combination and even prioritizes them for you showing you the potential gain. For instance, Needl can actually tell me that I need to improve my bounce rate on a specific landing page for users arriving via my Facebook campaigns.
So think of Needl as an ‘insight machine’. It tells you where to put your focus - and to some degree also what you should do. But not how you should do it.
And Needl is easy to work with. Being a machine, you basically just signup for a free (so far) account, link it to your Google Analytics, select the goal you want to optimize and then tell Needl to begin the analysis. I even tried to ask Needl to analyse 12 months of data (~2.5 million sessions) and it literally took minutes to process it.
How good an analyst is Needl?
If you don’t have an in house web analyst and if you’re not a web analyst yourself, Needl will definitely give you quite a few insights. And if you’re an experienced web analyst, I’d dare to suggest that Needl will give you a few eye openers as well.
As of now, Needl performs a lot of dimension combinations. Maybe too many. The thing is that each identified focus area (or ‘needl’ as Needl calls them) must be evaluated with a lot of parameters that might be business specific. Which is not something an automated tool can do really well yet.
For instance, Needl suggested that I should improve the bounce rate for direct sessions on Windows 10 computers if they landed on a specific page. That’s all good. But in reality, very few companies and CMS platforms have the capabilities to ‘personalize’ their websites to that degree.
On the other hand, Needl does provide a set of filters. So it’s possible to ignore suggestions (needls) that you’re unable to anything about anyway. But in order to perform that kind of filtering, you would actually need to be a somewhat experienced web analyst in the first place.
Who should use Needl?
Well, everyone could use a virtual web analyst. I definitely want one. And Needl makes an excellent effort. Obviously, it has its strengths and weaknesses. But it’s still and beta and it will definitely see some improvements in future releases. My suggestion for you is to take it for a test drive. Spend a couple of hours with it. You might be surprised.
In total though, I would imagine that Needl can find its place as a virtual web analyst in two ways:
- For people and organisations without any web analysis capabilities, Needl can be the shortcut and the de facto in house web analyst. Needl will provide fast access to otherwise hidden data and insights, and will tell you where to focus your efforts.
- For more experienced Google Analytics users (who are not necessarily web analysts per se), Needl can be the web analytics assistant. Needl will be able to provide suggestions that you might otherwise have missed in your daily work.
Personally, I think I will use Needl for some projects. And probably use it as an initial analysis tool when the project has an open brief (i.e. check my Analytics and tell me what to do). Most often, I work with well defined focus areas, but from time to time I do get those very open briefs that demand an explorative approach. And I think Needl will be able to help me on those.
In conclusion, Needl is just really cool. At the least the idea is fantastic, and the execution and the results are already pretty good, but will probably get even better in future releases. It’s not perfect, but no tool is perfect. But Needl does a really good job at automating some of the time consuming tasks in Google Analytics. So go ahead and try it!