Your big data audience insight big data show 24 apr 2013

Editor's Notes

1. Unstructured data is complex and hard to analyse which requires a new way of thinking
2. Second slide – what are the main reasons people are looking at Big Data Retail – customer insight and better targeted marketing campaigns Finance – get more value out of old data and achieve competitive advantage
3. Unstructured data is complex and hard to analyse which requires a new way of thinking
4. So as the industry uses 3 rd party data to scale audience, it is important to understand the implications of this, and how the process could be improved. Firstly there is lack of both transparency and common standards between providers. Data is typically bought in “user segments” rather than people. However how these segments are arrived at is unclear and not standardized across providers. An advertiser buying a user segment who is in-market for a Rolls Royce may not be getting the same segment from each provider. For example a segment could be based on users searching for Rolls Royce related kws. Another provider may chose to use data gained from the Financial times combined with Rolls Royce content viewed. They are clearly very different segments but could be sold as the same. Removing the noise is difficult – what types of data truly drive performance. Understanding what data types demonstrate causation rather than correlate with a behavior is difficult. And actually data can be very expensive and is available to all on the open market. This means that in many cases the insights gained, especially around prospecting, are rarely exclusive and used by many advertisers. This can push up CPMs in the auction which can hinder ROI – especially when the cost of data is factored in. Finally, the insight sits with the media buyers. But that insight could be more valuable. It could have an impact beyond your media buying and into your whole marketing efforts. So it raises a question? How good is 3 rd party data?
5. Exclusivity of data is key. Exclusivity delivers a unique value proposition. 1 st party data is the most relevant to your audience. It is something your competitors do not have. However, it needs to scale. The more first party data that can be combined with 3 rd party data to achieve scale the better. The future has to be about people rather than cookies. Only then are advertisers able to truly understand their audience. And finally, brands own their audience not media buyers. But when data plugs directly into the buying technology, the insights can stay there and not ripple throughout the business. This is where the move from data led Demand Side Platforms to brand managed Data Management Platforms leading Demand Side Platform execution becomes really exciting…
6. So as the industry uses 3 rd party data to scale audience, it is important to understand the implications of this, and how the process could be improved. Firstly there is lack of both transparency and common standards between providers. Data is typically bought in “user segments” rather than people. However how these segments are arrived at is unclear and not standardized across providers. An advertiser buying a user segment who is in-market for a Rolls Royce may not be getting the same segment from each provider. For example a segment could be based on users searching for Rolls Royce related kws. Another provider may chose to use data gained from the Financial times combined with Rolls Royce content viewed. They are clearly very different segments but could be sold as the same. Removing the noise is difficult – what types of data truly drive performance. Understanding what data types demonstrate causation rather than correlate with a behavior is difficult. And actually data can be very expensive and is available to all on the open market. This means that in many cases the insights gained, especially around prospecting, are rarely exclusive and used by many advertisers. This can push up CPMs in the auction which can hinder ROI – especially when the cost of data is factored in. Finally, the insight sits with the media buyers. But that insight could be more valuable. It could have an impact beyond your media buying and into your whole marketing efforts. So it raises a question? How good is 3 rd party data?
7. Hype cycle – this is not what Big Data looks like
8. This what Big Data looks like