I hope you are all enjoying summer and getting out of the office a bit. It's definitely refreshing to get some warm weather out here in Chicago. This month I wanted to cover the following:
How do you define an "Enterprise" or "Best in Class" analytics program?
I've been asked this question quite a few times in recent months, and I think this is a result of a common problem in our industry. We try to generalize and put everything in a nice, little box and it simply doesn't work in the ever-changing environment we live in. This is the reason the vendors themselves haven't aligned with a set of standards for things like definitions and terminology.
Sure, it would sound pretty cool to walk in a room and announce to everyone: "We have a best in class analytics program." But, the fact of the matter is that anyone could say that and we have nothing to point to that can challenge the statement.
What does a "Best in Class" organization look like?
We don't put stock in a template roadmap for a best in class organization. Best in class means something different to every organization, which is why we spend our time focusing on the business goals and the business units themselves to help determine what "best in class" means for our client.
In our experience every roadmap has looked different, even for our clients who have been in the same vertical and view each other as competitors.
We believe in a sustainable competitive advantage, which can mean many things. At its foundation, that requires a balance across Strategy, Infrastructure, and Value Creation Tactics.
That being said, there are quite a few themes that we see commonly when compiling a roadmap for our clients:
Education: One of the primary inhibitors to a successful analytics program is education. This requires educating different user roles within the organization and focusing the education so that it's actually actionable for the business. This typically takes shape in both an upfront and ongoing training program with both external help in addition to internal best practices (brown bags, etc.).
User Roles: Education plays right into this, because every organization needs to design a system for "who does what?" Typically, there are several stages of users: people that simply consume reports, create reports, conduct analysis, ad-hoc business intelligence, power users etc.
Reporting: It's important for a successful program to not only be getting the right data to the right people, but to also be publicizing successes to the organization when it comes to things like testing and optimization. KPIs, dashboards,UI design within the analytics application and many other tactics come into play in this category.
Analysis: Analysis is going to change over time. Typically there are people that interact with the web analytics tool only, and others that focus on the larger dataset which typically includes a data warehouse. You'll see some standard analysis that gets done weekly, monthly, quarterly, or yearly. You'll also see ad-hoc analysis done when someone asks a question or needs more information on something they saw in a report. Then, you'll see deep-dive analyses being done to help feed testing and targeting initiatives that help guide the business. This typically involves data integration.
Integration: At a high lever there are two ways to go about integration: physical integration or reporting side by side. When you are just looking to be able to draw correlations in the data and not segment a dataset with another, then you can simply report on it. However, the actual value is in data integration. As I've mentioned, our team is very skilled at integrating datasets while minimizing the costs typically associated from the vendors. Typical strategies here end up focused on the quick win integrations first: the ones that take less time but provide very valuable insights. Then, the longer terms ones, such as with the data warehouse, can build off that momentum since they take longer to complete and start the analysis.
Business Processes: There will most likely be a revamp of processes for things like the development of new content, the releasing of a new campaign, the Q/A process for new data collection, and the integration of a new technology to name a few. These processes will be adjusted to allow for a solid foundation in data collection.
Governance: A successful analytics program will require a foundation in governance. I've covered a lot of this already in the above points,but it would include things like accountability, common terminology and understanding of the concepts and reports, accuracy and validation that you can be confident in the data. There will also be global adoption across the entire business to become a data- driven organization. Some of these will happen right away, others will take time as the program evolves.
A forty thousand foot view doesn't become very actionable for organizations, so Stratigent focuses on providing the actual tactical steps to achieve the goals. These are just some of the main themes we see show up in organizations that have successfully become data-driven.