Business Intelligence (BI) & Digital Analytics (DA) serve overlapping objectives in many firms (such as understanding the results of marketing campaigns, or determining customer value), but because customers activate the insights from these systems differently, two separate fragmented vendor markets have developed.  The BI buyer’s use case revolves around business decision support, often based on the internal metrics of a business, while the DA buyer’s needs center largely around driving direct response marketing activities across the siloed universe of digital channels, content, devices and ad-tech stacks.

The different use cases have each required specific types of data sets, which lead to different underlying technologies.  BI grew up in the world of SQL-structured data, OLAP databases and on-premise ERP/CRM systems.  DA was built more recently in the cloud for the digital era of larger data sets of unstructured or semi-structured data and other difficult-to-integrate formats.  A huge new ecosystem of database and analytic technologies evolved to solve for the unique needs of DA as listed below –

  1. Data size.  Massive size of files and data transfers make it difficult for traditional databases and data transfer mechanisms to keep up.
  2. Data heterogeneity. Data extraction done by multiple vendors – each has their own collection format and data nomenclature, making them difficult to integrate.
  3. Data types.  Unstructured/semi-structured data.  Web log and JSON data provide flexibility but require new processing methods to make it useful.
  4. Aggregations.  Taking event data and aggregating it into formats consumable by marketing (e.g. online purchase paths – retargeting, click on ad, purchase) is not necessarily the strength of SQL based systems.
  5. Velocity.  Marketing analytics has demonstrated that reducing latency drives improved conversion, pushing the industry toward real time analysis.

Despite these differences, however, BI and DA serve many of the same purposes and require a high degree of integration if companies want to manage their businesses effectively.  As online efforts have grown, digital information plays an increasingly critical role in explaining the drivers behind BI metrics, especially in consumer marketing-driven firms (see our previous blog where we explore this use case for the eCommerce vertical HERE).  In addition, the rules that drive real time activation programs often require thoughtful analysis to create and refine.  BI offers a proven approach to facilitate those types of analysis.

Today’s Situation

With today’s solutions, companies find it painful to merge the BI & DA worlds in an elegant and scalable manner.  Customers are dealing with a custom built Babylonia of point solutions on the digital market and analytics side of the fence and, in many cases, in-house, legacy solutions on the BI side.

Companies often analyze clickstream data by channel in their AdTech tools (e.g. website, SEO, SEM, display) and when companies do attempt to integrate clickstream data into their BI systems, the data pipeline is often intermediated by AdTech vendors.  Anyone who has tried to integrate SiteCatalyst data with a search platform like SearchForce or Kenshoo knows that the definition of events can vary in nuanced ways that makes reconciliation tough and reduces user confidence in the results.  These issues can drives meaningful data integration to a standstill.

Latency often occurs when an AdTech vendor needs to process the data for its silo and the company then processes the data a second time to get the full view before taking action.  Too many steps almost always means too slow to get the right results.

Why BI & Analytics Went Their Different Ways_Diagram

The practical result?  Companies leave money on the table by not fully understanding customer behavior and by reacting to that behavior too slowly.  The former costs money in simple ways – like not turning off retargeting campaigns after a customer has purchased – and in more complex ways – like targeting a high-involvement considered purchase prospect with shallow media rather than a research oriented communication.  Latency can lead companies to ignore customers while they consider purchase and contacting them while they focus on other things.  A recent Ancestry.Com presentation (see it on slide 16 HERE) suggested that response differences, even over a period of a few days, can result in an order of magnitude change in conversion.

So, how do marketers address this Digital Analytics and BI gap?  Stay tuned for my next blog —
Why BI and Digital Analytics Will Integrate.