Data fragmentation in a multi-channel environment The relentless march of ecommerce has left in its wake a critical mass of data difficulties for multi-channel marketers. While the internet provides unparalleled access to services for customers, accelerated technological advances and the red tape that comes with them have also brought messy data fragmentation.

The core problem is that businesses often grow organically. Where they don’t, they may still grow awkwardly as a result of mergers and acquisitions. As new online marketing channels emerge and develop, a confusion of disparate data is dragged along behind:  a trail of valuable sales and marketing information spread across a mix of online and offline systems. Today’s problem is not how these various sets of data are stored - it’s how to go about retrieving, making sense of, and cross-referencing the data.

All this is at a time when, for the multi-channel marketer, every marketing penny counts. The benefits of having one easy way to access all of your marketing data are obvious. There’s less danger of duplicating customer contacts, or contacting them inappropriately, for one. There is also the administrative advantage: using an integrated data set, it’s easier to coordinate marketing activities across channels, gauge effectiveness, and allot budgets accordingly.

The importance of a holistic approach to data has never been greater, but according to research into multi-channel marketing strategies by E-Consultancy and Speed-Trap, achieving this goal remains distant: “Only 16% of organisations are currently able to combine online and offline marketing activity to maximise sales, but the majority are now actively seeking to capture and integrate the data required into an effective multi-channel strategy... Growing numbers now want greater integration.”

Stumbling blocks
Sadly, it’s far easier to identify these benefits than it is to enjoy them. Historically, two imperfect approaches have been employed to achieve data homogeneity. The first is to create a single “data warehouse” where all the disparate datasets are uploaded and stored. The second is to provide a common interface (essentially a virtual “friendly face”) that works with several custom “wrappers” - one for each dataset, each reformatting its repository of data.

" Only 16% of organisations are currently able to combine online and offline marketing activity to maximise sales "

There are many problems with both models. In data warehousing, the live data isn’t actually stored in the warehouse: essentially, a snapshot of each data source is taken, formatted correctly, and then uploaded into the repository. Maintaining accuracy is complex and costly and the data isn’t live.

Common interfaces don’t amalgamate data onto a single database; instead a bunch of different interpreters are bundled together inside one common interface, each accessing a different data source.  It sounds messy and it is.

The simplest way to integrate data is to use a homespun version of the first method. The end result is businesses copying information from various sources onto a dedicated marketing platform on a regular basis - resulting in the same problems and limitations of using duplicated data.

"ironically, the internet - which has done so much to damage data integrity -
offers a third, ultra low-cost method of joining the dots on data"

The third way
Reliable, fast broadband has blurred the boundaries between company networks. In fact, today more time is probably spent preventing network connections than enabling them. And ironically, the internet - which has done so much to damage data integrity - offers a third, ultra low-cost method of joining the dots on data.

This newly connected world has spawned a set of data interchange standards (such as SOAP) enabling web-connected systems to interact seamlessly and rapidly, as if they were based on a single machine. Online data providers enable businesses to publish their own data assets through secure web service interfaces - effectively offering a hassle-free virtual data warehouse. Reasonably well established in some specialist areas like address data management and business intelligence, these technologies are now ready to step into the mainstream.

The ability to access internal data using the “web services” model offers many benefits. Not only does the architecture provide real-time access to live data, but it also spans the online/offline divide, enabling both legacy and current data assets to be brought into play in any environment.

Web services offer the potential for the development of more highly integrated online services to increase sales and profitability, while enhancing customer experience. Creating the data feeds typically takes only minutes  and because servers rarely operate at capacity, infrastructure costs are minimal, while the services themselves are charged at a fraction of a penny per transaction.

Integration inevitability
It will take time for data managers to embrace the new technology. Protective attitudes towards data assets combined with concerns about the general security of the internet are short-term barriers to mass uptake. However, if data publishers like Royal Mail and Dun & Bradstreet can overcome similar fears, so can data managers – and the competitive advantage gained by the early adopters will seriously injure the laggards. It’s certainly fairly well understood that most security breaches are a direct result of doing something silly like leaving a CD on a train: it’s not rocket science.

"Ultimately, if we’ve learnt nothing else, it’s that having standards and easy access to organisation-wide data is a must"

In the meantime, however, some commentators are still missing the point. Aiming for one discrete, codified repository of data is a top-down and monolithic approach, effectively asking pre-internet technology to resolve what is a post-internet problem. The modern world of data is a shifting ecosystem, not a static model, and new marketing services, along with and mergers and acquisitions, are forcing change upon company data formats.

Adaptive, future-proof platforms will act as a babel fish for present and future data formats. By removing expensive middle-man format translation tools, web services create innovative new integration to increase sales and profitability while enhancing customer experience.

Ultimately, if we’ve learnt nothing else, it’s that having standards and easy access to organisation-wide data is a must.


Guy Mucklow
managing director
Postcode Anywhere