Why Is Cross-Channel Marketing Still So Hard?

Twenty years after the introduction of interactive marketing, why is it still so hard to create and manage integrated cross-channel marketing campaigns? It's a question asked by many companies that want to increase marketing performance while decreasing marketing cycles and optimizing budgets.

One reason is that organizations have not fully grasped the potential of Internet technologies. The Internet is essentially one gigantic information-sharing platform and should be treated as such. However, data silos have sprung up in many places, fragmenting customer information between various channels and sometimes even between different Web sites and applications. At the same time, older information management approaches are often too slow to keep pace—it can take more than a year to integrate data into a centralized data warehouse and, by the time it's done, there will be new data, making the process an exercise in predictable obsolescence.

Between the inability to use data across Web sites and applications and the slow migration of data into centralized data warehouses, marketers have been unable to effectively create cross-channel marketing strategies using existing data. Here are five proven approaches to solving the data bottlenecks that hamper effective and integrated digital marketing:

1. Take stock.

We've found that very often companies do not quite realize just how much information they have locked away in different compartments. A useful exercise is to take a complete inventory of all the data you have and consider how it might be put together to enable better marketing.

Let's say that you have very granular data about site behavior in your Web analytics platform and transactional data from in-store sales. By merging this offline and online data, you may be able to gain insights into what sales are driven by digital campaigns.

2. Design for data flow.

Most well-run digital marketing projects pay a lot of attention to visual or usability elements, such as site navigation and functionality. But quite often, what to do with the resulting data and how it is related to the overall marketing enterprise are hardly considered or hastily added on at the end of a project.

From the first day of any digital development effort, whether it is for a Web site, a mobile app, or social channel, data flow should be part of the design effort. For instance, if decreasing shopping cart abandonment is a priority, capturing relevant events and getting that data quickly to a marketing system must be considered.

3. Develop application programming interfaces (APIs) for your data.

A key innovation of Internet technologies is the ability to wrap Web services around just about any data. In the past, accessing databases in real time was often an arduous programming task. The rise of Web services has made this many times easier, but it also requires something of a shift in thinking by IT professionals. Rather than creating extracts of data to support individual applications, it can be much quicker to develop simple APIs that can support multiple uses and channels. The same Web service that retrieves a customer model score for Web personalization can be called to inform an interaction in a call center.

4. Allow for laziness.

Because digital technologies need to operate in real time, tradeoffs need to be made between speed and efficiency as well as completeness and accuracy. From a data warehousing perspective, it might be preferable to have Web data arrive in perfectly clean condition, but extensive validation and error correction are not always feasible upfront. We're finding that some of the big data tools, notably HBase (the database behind Hadoop), allow us to be lazy about collecting data in real time and driving real- or near–real-time decisions, while supporting more rigorous data cleansing for data warehousing requirements.

Essentially, we can gather marketing events, such as shopping cart abandonments, and stuff them into HBase for immediate use. Then, when time permits, we can transform and move relevant data to a data warehouse. We can be lazy upfront and still capture data for both immediate and later use.

5. Preserve customer identity.

Every company has some way, and often multiple ways, of identifying customers—a customer ID, membership number, cookie, etc. As much as possible, companies should try to preserve customer identity over each leg of a digital marketing interaction. For example, URLs in emails can have query strings that identify each unique customer. These can be picked up by the referrer code on your Web site and logged in your Web analytics platform. If the customer then clicks a live chat link on the site, the identifier can in turn be logged in the chat session. Of course, such associations should be used not just for gathering information, but also for improving customer experience through more seamless navigation and more relevant information.

In summary, a vital strategy for improving cross-channel marketing is to pay attention to the information your company is capturing and how it is being shared. Doing this will lead to sharper insights, better marketing, and improved customer experience across your online and offline touch points.

Richard Vermillion is a software engineer, technologist, and executive with more than 20 years of experience in developing software and databases for customer management applications. As CEO of Fulcrum, he leads the executive team in setting and executing the company's strategy, communicating its vision, and establishing the corporate culture. Prior to joining Fulcrum, he worked as a specialist for McKinsey & Company and as a contractor at NASA.