All the technical detail, expertise and advice

Back

Not all data is born equal – tagging for data capture

Published: Thursday, 14 August 2014 09:48 by Paul Siddall, Senior Pre-Sales Consultant
Big Data Analytics

Tag Management Systems (TMS) are typically in place as a means to solve the 'tagging' issues caused by ongoing requirements to add JavaScript tags to a website, usually for traditional web analytics solutions and online advertising. A TMS does simplify the tagging process by removing the requirements from the web development teams to add the tags to the web site directly, but there is still a need to build and maintain a tag management strategy which is an extremely time consuming activity.

Since TMSs are already being used to assist with data capture, they have recently been extended to also capture this data directly themselves. For an organisation already using a TMS to assist with the tagging for their web analytics solution, the initial implementation process to capture data via the TMS is just as easy as deploying Celebrus. However, the quality of the data captured is entirely dependent on the tag management strategy defining exactly what data is to be captured: if it hasn’t been tagged it simply won’t be collected. This strategy must therefore be scrupulously maintained as and when new items of interest arise and website changes occur, otherwise data will be missed or incorrectly analysed.

Another important aspect to consider is the ease of data loading. Many TMSs offer a feed of event level data from their data cloud, however it’s important for an organisation to understand whether it will have to build a data schema and set up an ETL process to get that data into the database. The company must also consider whether an hourly feed is quick enough – definitely not if it wants to engage in real-time or event-driven marketing.

Moreover, this event level data has little to no data modelling or processing being applied, meaning that for detailed web analytics queries a large amount of data preparation will be involved by the company which must be performed by someone who is familiar with the tag management strategy as that’s the only way to understand the event data. In addition, individual level analysis would only be possible with even more in-depth work on the tag management strategy and huge levels of data preparation on top of the event data feed.

By contrast, Celebrus offers either an in-house or cloud-based solution which provides absolutely complete, contextualised and fully structured individual-level data loaded directly into a pre-defined schema in the organisation’s database in near real-time. The Celebrus data has had business context applied so is immediately ready for use to drive marketing actions or to power customer analytics.

As well as the application of context such as clicks, page loads, basket adds, form submits etc, all interaction processes are passed through the Celebrus Scenario Engine, allowing complex multi-page or compound event scenarios to be matched and processed in real-time. For example, ‘individual searched for a product, browsed it, added it to their basket and then left the site’ would be set up as an ‘abandonment scenario’ which could automatically trigger another system to send an abandonment email. As well as frequently used scenarios such as that one, businesses can build very specific scenarios to meet their own needs.

Overall TMSs have a very good and clear role to play in a digital marketing technology stack. However they are not a panacea to cure all evils as they still require considerable time and effort to capture and deliver good, meaningful data in the time frames needed by today’s marketers.

To read what leading industry analyst Gary Angel, President of Semphonic (now E&Y) feels about the role of TMSs and why he believes that using web analytics data as the key digital data source is often the wrong decision download his white paper “The Future of Digital Measurement & Personalization” here >