As a data analyst and a researcher, we are always hungry for more data. Companies are struggling just to set the right expectations on KPIs, and understand what to do and how to take action on those data. That said, adding more data could be running a risk in creating another layer of noise, but if done strategically, more data to your metrics portfolio is still vital to creating new insightful stories for the business.
It is important for analyst to strike a balance between the reality and stepping up the underlying data for the business. Here are some ideas from me that I’ve used in the past to add more data points for insightful analysis, even if there are no needs or requests for the data at the time of implementation.
These added data could be leveraged to add additional insights and create new stories for your managers and executives.
Add more tracking implementation points into one request
I use Webtrends on daily basis at work, but custom event tags added for a specific feature could mean an opportunity for you to request another tag extending the request to add additional tracking point relating to that new feature.
Leverage QA process to come up with next action item
One of the time consuming yet important part of web analytics experts job is to QA ta implementations so that good data are coming into your web analytics tool.
Make sure you’re not only QAing one part of the implementation. Make sure you’re thinking about data nesting and its hierarchy so output of the data on to the report is consistent to the business expectation or needs. Think about new opportunities to track, implement, fix, optimize, and report as you click around firing off the tracking code on httpwatch, httpfox, WASP, etc.
Create multiple outputs from one tracking parameter
We shouldn’t be thinking about one tagging parameter equals one report anymore. Because we can create advanced filters and segmentations around one data combined with other data dimensions, we should be thinking about how the new data points would roll into other multi-dimensional data.
Apply one segment and another counter segment
When we create one segment, let’s think about creating another segment that does not represent that new segment. For example, when you look at a data segment around people who bounced, what about looking at a segment of data for people who did NOT bounced would be valuable, too. You get the idea right? There maybe something insightful by looking a pair of segments that could aid the analysis from two different points. Thinking like this upfront will save you some time in analysis down the line. Creating segments in pair is very nice.
As we are running tight on resource, time, and budget it is definitely helpful to think a little extra so analysts could scale. I believe little extra thoughts added to daily web analytics practices will save a lot of time, and add potential leverage in data insights where analysts could paint better stories and picture for the business.