Digital Data Analytics – Part 2

For this article I thought I would provide a view on the different types of digital analysis typically used and where some data sources can be found.I should state before I go any further, that a great deal of background information has come from some excellent Internet Advertising Bureau (IAB) presentations and information available online, as well as conversations with subject matter experts.

This article discusses:

  • Targeting
  • Data sources
  • New types of data


When spending money on advertising, it is important that the ad will be seen by the best audience to convert the ad into a sale. With the amount of inventory increasing every year (34% over last year) it is important to choose the right space otherwise you end up wasting a huge amount of money. To help with this process, data analytics is typically used as one of the tools to identify the following:

People You need to identify the right audience and their typical characteristics, such as their site viewing habits, social networks used and typical habits online. This is the key information for an analysis and it is not always easy to identify the correct audience from the huge volume of data.
Place You should identify where your audience go online, such as the sites they use, social networking habits and gaming.  But you also need to understand which online platform to use, such as mobile, general internet or perhaps connected TV to name only three.
Time What part of the day do they tend to be online and most receptive to your type of product? This might also affect the platform to target and the type of message being used.
Message As with all advertising, it is important to present the right message to generate interest.  In an online world you also need to decide on the format that message takes, such as Display, Search or Video.  Once you go beyond these broad three types, the options increase exponentially.

Whichever part you have to play in advertising, all the above factors are essential. When you are building a media campaign for an advertiser, you have to be able to make a sensible case as to why you have made your decisions on the sites and platforms to use. If you are selling a site as a publisher or network, you have to be able to match the demand side requirements with your typical audience and behaviour.

Data Sources

There are many sources of data for analysis and here I have tried to break down some examples of the various types of data suppliers you would expect to find. Very often you need to combine different data sources to provide all the information required to gain the right targeting answers.

Publisher Publishers probably hold most of the data that will be enable you to identify a more detailed analysis of how their sites are used, but they also hold other key information such as registration data, keyword searches and platforms used to access their site including mobile and gaming.
Advertiser The Advertiser will hold information on their customers and how their sites are used, but also have good conversion data assuming ads are tracked properly. Obviously this data is extremely valuable primarily to the advertiser and to the agency if made available.
3rd Parties There are many 3rd parties that provide information on online activity. Some of these are behavioural and audience related, whilst others cover all aspects of online activity.  There are also loyalty schemes such as Nectar and Tesco Club card to name two, which can be used to match against other data sources to enable a richer analysis of your audience.
Mobile Operators The mobile operators are in a pretty unique position to provide comprehensive and accurate data on usage of the internet and applications as data is tracked on personal devices with known profiles.  Currently, this is just traffic over the telephone network, but moves are being made to track data when the device is connected over wireless networks too.
Your own You may have your own data, which could relate to your business or trading activities and this all has value and can be used. It is important to find or create ways of merging your data to other sources, enabling a unique view that no one else can match.Most people now value digital data to such a high degree that they are actively trying to find ways of capturing more and having ownership of it in order to have a USP or to generate revenue.

Most people now value digital data to such a high degree that they are actively trying to find ways of capturing more and having ownership of it in order to have a USP or to generate revenue.

One of the main reasons data has to be brought together is fragmentation of the whole picture and different owners holding a piece of the puzzle. This is just accepted in the industry as a necessary evil as online content is so dynamic; differentiating it from linear media types. Many are trying to build this overall picture and time will tell if this is really possible and can be 100% accurate.

New types of data

There are new suppliers of data coming to the market and one of these is Experian who have recently made their Mosaic data available online. They have an industry standard set of 155 person types that boil down to 15 groups and 67 household types. These categories have been used for 30 years in an offline world and now they are available for use in digital analytics. Microsoft is working with Experian and they have found that there are nearly 33m addressable audiences that create new avenues for analysis with these extra data elements available. For more information on this please refer the Microsoft Advertising information on the Experian data.

There is no doubt that more companies perceived as providing offline reports will be turning this into a chargeable online source of information to help advertisers target the right people, at the right time on the right site/service with the right message.


It is very important to understand that with any of these data sources it is not possible to go down to an individual person or address for privacy reasons. I discussed some of this in a previous article Digital Data Analysis – Part 1 covering the risks and issues with digital data.

This is by no means an exhaustive coverage of this topic and I am sure many will disagree with some of my points. However, it touches on some of the key areas to generate a discussion. I am always happy for comments – positive or negative – to gain other peoples’ perspectives and to educate me more on this area. I am also always willing to discuss matters on the phone.

Telephone: +44 (0)20 7193 6879