You are not alone — common enterprise data problems

I like the bumper sticker that goes: "Never forget that you are unique, like everyone else." Most of our clients believe they suffer from the ugliest pile of unmanageable numbers possible. Guess again; you’re probably no worse off than the next guy.

In an attempt to ease your fears of being alone with your data troubles, we’ve put together a list of common data-related issues we see in our client work:

  1. No unique identifier. Faced with numerous enterprise systems and any number of customers and employees entering data, many organizations are unable to maintain unique customer identifiers. With unique identifiers, you can match customers across their interactions with the organization; without them, it becomes very hard to get a full view of the customer experience.
  2. Blocked by the reporting front-end. Many enterprise systems (CRM, ERP, etc.) do you the "favor" of bolting on a reporting engine. Set up correctly, these tools can be modestly satisfying in their ability to spit out metrics and support basic slicing-and-dicing analysis. However, when you start to ask complex questions or want to dig into the raw data, you find that your reporting engine is more of a door than a window.
  3. Too many reports. A big stack of dashboards, key performance indicators, and success metrics is piling up on your desk—and yet you don’t feel like you understand the state of the business. As we pointed out here, too many metrics can mean you don’t fully understanding business drivers, but want to create that facade.
  4. Inconsistent data definitions. What does "customers" mean? For marketing, it is the number of people brought in the door. Operations only counts the number of active users. This leads to any number of  unproductive conversations trying to explain discrepancies in the numbers.
  5. Messy, unstructured data. Data is rarely arrives in an easy to use form.  Sometimes it is spread across tables in Excel (or worse, PDFs).   Dimensions and measures are poorly labeled and not defined.
  6. Access. Getting to the right data can mean a well-formed and informed request to the IT group. When IT and business folks struggle to communicate, data stays locked away. Our friend Dratz writes a great blog that offers nice perspective for business folk like me.

  7. Data shmata. Sometimes all the good analysis falls on deaf ears. "Some people in an enterprise apparently aren’t really looking for a single version of the truth. It can be easier for them to work with common assumptions and ’dance with numbers’ during management meetings", says Bill Hostmann of Gartner.
  8. The data warehouse is late to the party. This has to be our favorite. While working for AOL, I watched as on two separate occasions as expensive data warehouses were delivered just as the business changed direction.

Do any of these sound familiar? We’d love to hear your stories of data trouble. Or leave us a comment if you’re interested in our strategies for tackling these problems.

A few other resources on data and business intelligence troubles:

  • The Open Source Analytics blog discusses data marts, data warehouses and the related philosophical debates.
  • Gartner’s take on "BI’s seven fatal flaws"
  • has a host of articles about enterprise and customer data integration