The #Digital30 series is Sift Media's fortnightly, 30 minute session aimed at upping our digital skills & literacy in the ever complex world of publishing. We'll share the practical advice, tips, tricks and takeaways that you can apply day-to-day, in your job.
Our first #Digital30 session kicked off today on analytics, insights and audience. Here are my top 10 tips.
1. Define your objective before undertaking the analysis.
What are you trying to achieve with the piece of analysis? Be clear on what outcome you want, and the tools you’ll need to achieve it.
2. Trust the data.
Let the data do the talking. If you’re interrogating a dataset to confirm an idea and the data shows you something different, it is more than likely that you were wrong in your assumption.
3. Don’t trust the data!
Data comes from human interactions with a system and sometimes people do things to cut corners or use a system erroneously. Be prepared to factor this in where a dataset looks odd, or isn’t returning what you’d expect (but at the same time, remember point 2!)
4. Use online resources.
If you’re stuck or don’t know how to do something, search online. There are a lot of people out there doing all sorts of analytics work and most of them are more than happy to help. Depending on the ubiquity of the system you’re using there will be something out there to help.
EXCEL
Excel Forum: http://www.excelforum.com
Chandoo Blog: http://chandoo.org/wp/
Mr. Excel: http://www.mrexcel.com
GOOGLE ANALYTICS
Analytics Help: https://support.google.com/analytics
Google Analytics Forum: https://productforums.google.com/forum/#!forum/analytics
Lynda.com: http://www.lynda.com/Google-Analytics-tutorials/Google-Analytics-Essential-Training/160343-2.html
GENERAL ANALYTICS
Occam’s Razor: http://www.kaushik.net/avinash/
Analytics Talk: http://cutroni.com/
5. Break things!
You need to know the limitations of the tools you use, and the best way to do this is to push them to their limit – try pulling huge date ranges, pile up the metrics to the point of absurdity, mutilate your dataset! BUT – make a copy: if you’re using a dataset that you might need to use again it’s a whole lot easier to have a backup waiting as opposed to having to pull data all over again.
6. Customise for your job.
Most analytics tools have some sort of custom dashboard option. Use these to make the reports that matter to you – this will save time and make your job a lot easier.
7. Get as much data as possible.
The more data you have, the better and more accurate your analysis will be. As a rule of thumb, ten data points is the bare minimum.
8. Know the data source’s limitations.
All tools have their limitations and there’s no point trying to get data out of a system that isn’t going to give you what you want. Go back to point 5 and repeat.
9. Be creative.
A data table doesn’t know the context of your job, a pie chart wasn’t at that meeting you went to last week. You need to be creative and apply context, narrative and inference to your analytics to get the most out of them.
10. Look out for our audience insight blog posts
We'll share a regular blog post detailing interesting analytics and insights will hopefully keep you curious into what’s going on with Sift Media'a voluminous data sets and how they might be used.