Sow the seeds of good data and harvest later

Every website is facing the cross-device issue; further complicated when you throw marketing attribution into the mix. Equally, plenty of sites have pages numbering in the thousands or millions (or billions, if you’re very lucky/unlucky), so whilst it’s easy to see your popular content at a very granular level, it won’t help you to see the all-important bigger picture in terms of trends and patterns –fundamental concepts of good analysis. The same can be said of analysing product performance for big retailers– there’ll be thousands of SKUs to wade through. Finally many organisations face the problem of linking multiple data sources together, for example a CRM system with a web analytics platform. Again it’s difficult to get the complete view needed to understand what aspects of your website or marketing are successful and vice versa.

Plan, plan and plan some more

As with everything, there’s generally a more complicated solution and an easier one. In our instance, the more complex approach is to use database-geeks and powerful technology to run queries on data. In a web analytics sense, it could be integrating your web analytics and other data sources with Hadoop, querying that database using SQL and then visualising it in a (hopefully) beautiful way using Tableau so you understand what action to take from the data. There’s nothing inherently wrong with this solution but it does have some notable obstacles; i.e. having personnel with the required skill-set, budget for tools and all-important access to additional data-sources.

What’s the alternative?

Google has taken huge strides in this space to make Google Analytics (GA) the ideal tool to become a holistic business reporting tool. It’s possible to configure GA to track everything you are interested in, and then use the interface to run your analysis. Being able to carve up data sets using segmentation and filtered views, plus create custom reports and dashboards on the fly makes it a powerful analysis platform.

The core concepts behind Google’s move to make GA holistic reporting utopia can be classified into two simple areas; integration and customisation.

On the integration side, GA has long integrated with various marketing platforms in order to see the performance of those channels. More recently you’ve been able to build Audiences to market to via AdWords and DoubleClick Bid Manager.

On the customisation side there is the ability to pass Google a unique ID when a user logs-in to your site, tying together sessions across difference devices and opening up a realm of cross-device reporting. You can even pass this ID into GA to start to link directly with your CRM. There’s also the ability to upload a spreadsheet of data to create your own ‘dimensions’ to help better understand the performance across new attributes. For example, uploading the blog author and topic on publisher sites to understand reader engagement at those levels. This extends to uploading refund data from sales, and spend from marketing channels, to help with more accurate ROI calculations.

Finally, there’s the ability to group content on a website to make analysis easier. Rather than trying to wade through a list of half a million pages, you can tell Google how to categorise your pages.

Build your castle on stone, not sand

As the oldest of adages goes, build your data framework on a solid foundation and you won’t sink in your data. The insights you gather and the action you take as a result will be far more powerful than simply trying to look for trends amongst the top 5% of data in a given report. That is where you are most likely to see the efforts of your data analysis bear fruit and flourish.

Oliver Walker, Head of Conversion Rate Optimisation, Periscopix

Tel: 020 7234 0500

Email: enquiries@periscopix.com

Web: www.periscopix.co.uk

Twitter: @Periscopix