Why has big data become such a hot topic?
Actually, if you walked into a meeting of IT professionals and tried to pass off ‘big data’ as a new concept, they’d laugh at you. Every commercial organization – as well as a host of government and agency bodies – has been housing and maintaining a variety of large databases for years.
So why is everyone in marketing so excited about big data?
What’s different now is that organizations are looking to integrate their offline databases with the online digital assets. Doing this gives them the power to fully understand and optimize the customer experience across all channels. In addition, being able to connect the dots across the channels opens up the predictive capabilities of analytics.
Why is this so difficult?
Integration usually means putting all the data together in a single database and that’s a huge task that creates a couple of key problems. The first challenge in creating a single database is that it raises the ugly question of internal politics. Each data set has an owner and each of those owners wants to own the integrated product.
The second challenge is that merging multiple databases is no easy task and it will take time, possibly years. Businesses might not be able to gain any short-term value from the process if they’ve locked the whole process up in a single ‘all or nothing’ project.
I’m feeling reluctant already…
Don’t forget that even before you can start to build a single database you need to know which bits of it are most important to the organization. You can’t create a scalable end product without this understanding and it needs to go well beyond the ability to report on the data. The plan needs to identify how you intend to operationalize the data within the organization to improve the user experience. The final challenge of this approach is, of course, money. There is a huge track record of big IT projects going wrong and every dollar spent will be heavily scrutinized.
Are there any advantages in this approach?
I honestly can’t come up with a single advantage to just trying to build the ideal integration (all data in one spot) right out of the gate. There are way too many risks and cons to this approach that I struggle with ever being able to recommend it. Clients trust us to add value and build programs, so we find the iterative approach works best. Each client matures at a different rate.
But is there an alternative?
There is, but it requires organizations to think differently about big data. Instead of integrating the information into a single database, they need to consider how they can get the right mix of data to the right people at the right time. The challenge then is not to build a giant database but to get the right combined data sets to the necessary stakeholders in a timely fashion. Long term, these individual databases may get you to the point where you have all of the data in one spot. However, if you take this iterative approach, you’re basically guaranteeing that the single database down the road will be built in a scalable way using the right data.
Can you give me an example?
One example would be for business stakeholders or agencies that are in charge of reach or acquisition campaigns. You may have cost data for those campaigns in one spot, conversion data in another spot, and impression data somewhere else. So, the program leader and his team might benefit from having those sources integrated to view in one spot and be able to understand how things are progressing in real time. That’s a pretty simple example, but it shows how you can get data ready for action quickly.
What are the benefits of this approach?
The goal for the data that you collect across your channels, both offline and online, has always been for you to be able to take action based on what that data is telling you. Taking an iterative approach to big data allows you to get some quick wins. The ‘Holy Grail’ is to build an optimized experience for each individual, or persona, that interacts with your business. That cannot happen overnight, so focus should be on getting the business stakeholders access to the data that will allow them to take actions to improve the customer experience one step at a time.
How does this work in practice?
The ideal approach to action is to crawl before you try to run. The reality is that not all data needs to be integrated. In fact, there’s a good chance that your stakeholders might only need to see the channel data side by side in order to draw some business-improving conclusions.
What are the first steps?
Here are four things you could do that would start the journey to enhanced customer understanding. You could implement a Voice of Customer tool such as OpinionLab. This will help give you real customer feedback about their experiences and help you understand more about the ‘why’ people do things in combination with the ‘what’ people are doing by analyzing your web analytics data.
You could also just take an hour to analyze some of your campaign results. That’s where you’re spending your money, and you’ll most likely learn something interesting. You can use a mix of A/B and Multi-Variate Testing. This allows you to test out different approaches and there are several cost- effective options that let you get your feet wet in the testing space. The returns can be incredible ROI. As you increase the data you have available for these tools to use, you can really start to target specific content to specific people based upon what you know about them.
Finally, you could test out some automated visualization tools that allow you to create ‘side-by-side’ visualizations of data silos without the formal integration of data. The key to success is to automate as many of the manual processes as possible so you can spend your time actually analyzing what’s being displayed.
Stratigent provides multi-channel analytics and consulting to enterprise level brands. For more information, please see www.stratigent.com