Blog Host: Jill Dyche - Partner/Author, Baseline Consulting READ ENTIRE BIO

Since most of you are still in research mode on your CDI and CDM efforts, I thought I'd leave you with some final words of advice that will help pave the way to success:

1. Consider data as a service. Most of us associate our data with the databases, platforms and applications in which it lives. But, with the emergence of service-oriented architectures (SOA), our newly consolidated business processes should be leveraging certified versions of enterprise data, which means access to centralized CDI and PIM hubs.
2. Be willing to replace existing solutions. Adopting CDI, MDM, SOA, data warehouse appliances, and a host of emerging technology solutions might involve "sunsetting" existing platforms that might not be as efficient or as cost effective. Pride of ownership is the barrier here, but, with a good business case, it shouldn't stand in the way of innovation.
3. Understand what data governance means to you. A lot has been written about the different types of governance. Whether it's IT governance or data governance, simply convening a group of executives to make tough policy and investment decisions is only one component. Governance might have a clear definition, but its execution nuances are usually organization-specific.
4. Embark on MDM slowly. Master data management is a progressive concept and, likewise, an incremental effort. The meaning and location of clean, reconciled enterprise data should be well understood by the diverse services that need it. We've seen too many clients lately that want to eat the elephant in one bite.
5. Begin with the customer. A recent survey by the Business Performance Management Institute reported that evolving customer needs and preferences was cited by executives as the main reason for driving business process change and new applications. Clearly, customer-facing programs are not only getting mindshare, they're getting budget.

All that said, remember the goal: happier customers, more efficient front- and back-office operations and higher profitability. CDI and CDM can help propel you forward, and then some.

Jill Dyche is a partner and co-founder of Baseline Consulting, a management and technology services firm specializing in business analytics and data integration. Jill is responsible for delivering industry and key client advisory services, and is a frequent lecturer and writer on the business value of IT. She is a faculty member for The Data Warehousing Institute, a judge for several industry best practice awards, and a regular blogger. She is the author of three books on the business value of information, e-Data (2000); The CRM Handbook (2002); and her latest book, written with Evan Levy, Customer Data Integration: Reaching a Single Version of the Truth (John Wiley & Sons, 2006).
Posted by Jill Dyche Business-IT alignment: Advice, with a side of slaw -- Part 2
07 SEP 2006 12:04 EDT (16:04, GMT)
In yesterday's blog, I described the role of trusted advisor. We actually recommend this approach to some of our clients. It involves imparting your unique expertise, which is the easy part. It also requires changing entrenched behaviors, which is a bit harder. (That's where the food comes in, so stay tuned.)

For our business side clients, this might mean chipping away at IT's understanding of business requirements, and taking a role in communicating them clearly. It means keeping existing IT relationships warm. It's not about another requirements gathering session or BRD (Business Requirements Document) -- it's about regular, sustained conversations about the need for business information and how it maps to the company's strategic evolution. The business needs to educate IT in a structured and consistent way.

Likewise, we counsel our IT clients to educate the business on emerging technologies and buzz-worthy ideas. Just as the "big bang" method crashes and burns as a development approach, it's also dangerous when it comes to introducing a new technology or proposing a new project. Better to chip away at the business' understanding of key technologies in bite-sized chunks. "What is MDM and Why Should We Care?" -- note the collective "we" here -- and "Case Studies of CDI Success" have both been big draws recently. And they've been delivered by our clients, not us.

So called "Lunch and Learn" sessions are a good start here -- especially those that involve the proverbial free lunch -- but there are other ways to begin a conversation with your peers. If you're an IT person in a CDI status meeting, start planting the seeds for PIM if it's a recognized need. If you're a business manager mapping out your order-to-cash process on the white board, offer opinions about gaps in automation. Do have a point-counterpoint session on whether your ERP system's data warehouse platform offers enough flexibility for your end-users and spark a lively and friendly debate. Or simply get on the calendar of an executive you haven't talked to in a while.

The point is to have the conversation early and often. So put me down for the Turkey and Swiss on rye, and a side of you-know-what.
Posted by Jill Dyche Business-IT alignment: Advice, with a side of slaw -- Part 1
06 SEP 2006 18:24 EDT (22:24, GMT)
One of our core tenets at Baseline is that we want to be our clients' trusted advisors. Some of our clients have engaged us on multi-year projects. Others consider us their data integration and BI "go to" firm and hire us on an as-needed basis. Some engaged us on a single project and leveraged the knowledge transfer to become self-sufficient.

Whatever our client's current "status," we take our trusted advisor role seriously. Once you're a client, you can call us for advice regardless of whether or not we're currently engaged with you. Heck, we'll talk to you even if you didn't follow our original recommendations. You'll get our Client Insight newsletter even if we haven't seen you for a while. Our sales force automation system notwithstanding, this is how we do customer relationship management.

The reason I bring this up is because lately -- in the tension-rife climate of business versus IT -- we've been suggesting a similar approach to our clients. What I mean is that the business should begin acting as a trusted advisor to IT, and vice versa. We've recently watched IT practitioners teach their business counterparts about master data and why it's so critical to the business. And we've seen business managers formally lay out their information needs for the coming quarter and what they'll need in order to achieve their KPIs or conquer that market share mandate. These conversations are different, interesting, engaging and someone always learns something new.

In tomorrow's blog, I'll describe how these conversations take shape. (Hint: sandwich, anyone?)
Posted by Jill Dyche CDI versus CDM
05 SEP 2006 16:15 EDT (20:15, GMT)
Lately, there's been some confusion about the difference between "customer data management" and "customer data integration." I think it's because both CDI and CDM promise more accurate, timely, harmonized customer data available to a range of business processes. The knee-jerk differentiation is that CDI means automation and CDM means ongoing work effort.

If you assume that Master Data Management (MDM) is the umbrella for the management of enterprise master data across data domains, then CDM is the establishment of terms, value representation, and processes that determine the rules for customer data. CDM does not establish how customer data is integrated: That depends on the existing IT environment, incumbent technologies and business requirements.

For instance, when I need to move data between transaction systems because my Order system needs to inventory information from my Fulfillment system, some sort of Enterprise Application Integration (EAI) tool is usually in place to address this. Likewise in the BI world, Extract, Transformation, and Load (ETL) solutions usually migrate data from operational systems onto a data warehouse. And where there are operational integration needs -- for instance, when a CRM system needs the most accurate phone number from multiple systems with conflicting details -- CDI can serve the purpose. CDM is ultimately required in each of these circumstances.

CDM establishes standardized terminology, definitions, and details about the elements that represent a customer. Additionally, CDM is the means by which data stewards define business rules to address data conflict, data inconsistencies and error correction.

When most people talk about CDI they're usually referring to the combination of technologies and implementation methods for customer identification, data conflict resolution and data integration. The CDI hub is where all the magic happens. Essentially, CDI supports the idea of customer data as a "service" to the enterprise, ultimately serving as the enterprise system of record for customer data.

Ultimately, I see CDI and CDM as peers. CDM is methodological -- it involves the data practices for enterprise customer data management. Could you do CDI without CDM? No. If anything, CDI mandates additional rigor around customer data policies and practices. Could you do CDM without CDI? Yes. But it wouldn't be as much fun.

* For a complete discussion of the full "CDI stack" that describes the various CDI functions and how they work within a Service Oriented Architecture, see Chapter 7, "Making CDI Work," in Customer Data Integration: Reaching a Single Version of the Truth (John Wiley & Sons, 2006).
Posted by Jill Dyche A message to the early adopters of CDI: Yay, team!
04 SEP 2006 15:46 EDT (19:46, GMT)
Sometimes it feels like I'm solo-ing when it comes to customer data management and customer data integration. As background, Baseline Consulting, my company, has delivered data warehouse, business intelligence, and data integration projects for some pretty big, important companies. But, as I said in my September 1 blog, CDM and CDI are still confusing topics even to data people. There are some that don't want to adopt the management of customer data simply because they know what a big job it is. Others see incumbent technologies doing the job. Still others are amidst protracted MDM planning efforts that are already wearing thin with constituents.

So I was particularly encouraged by Aberdeen Group's recent report on CDI and CDM. As reported by Hannah Smalltree, intrepid TechTarget reporter, Aberdeen has just completed a large research study and found that companies were starting with small, controlled projects around customer data (my words) and "don't worry about MD" (their words). It all goes back to thinking big and starting small (everybody's words). Aberdeen's research finds that "above-average and average performers were twice as likely as below-average performers to use CDI and data quality tools." In other words, the early-adopters of new data management and integration technologies are starting with small, controlled projects and showing value quickly, while integrating this value into a larger plan over time.

"The takeaway for companies contemplating MDM is to do a pilot program with customer data or product data first, refine it -- then roll it out," Leslie Ament told Smalltree in the article. I couldn't have said it better myself.
Posted by Jill Dyche CDI and CDM in the crosshairs? Say it isn't so!
01 SEP 2006 15:33 EDT (19:33, GMT)
You'd think customer data integration and customer data management would be ideas whose time had come. I mean, who could possibly be anti-customer data management? After all, doesn't the need for de-duped, standardized, matched, integrated and certified customer data transcend size, gross revenues, vertical industry or market segment? Don't we all want a single version of the truth about our customers? Ask me about Baseline's old Christmas card list sometime.

But CDI and CDM have become controversial. Who knew? Like any "new, new thing," trends can be threatening. As they slog their way up the technology adoption curve, the hefty, robust trends will ultimately leap across the chasm and into the mainstream. In a lot of ways, CDI and CDM are still at base camp.

We've heard a lot of rationalizations for why companies aren't ready for CDM or won't adopt CDI technologies. The most worthy excuse is that a few visionary companies built CDI solutions from scratch before CDI was an acronym, and they are heavily invested in these sophisticated, homegrown systems and can't back out. Let's take a minute to honor these firms: they had a problem and they solved it before the solution became de rigueur.

But the excuse we hear most often by far is "we can already do that with 'x'" where x equals an incumbent technology or development process. These companies have hung their hat on adapting technologies that were never meant to do the job of CDI. "I'd love to consider CDI," one car company manager wistfully admitted recently. "But we're too far gone with our custom-developed ETL [extract, transform, and loading] code." Too far gone indeed.

I'll leave the Maslow's hierarchy of need discussion to someone with a more sociological bent. Sure, personalities and distinguishing cultures will always exist and human beings have historically competed for survival, so maybe the struggle is still in our DNA. But show me a corporate culture that's change averse, and I'll show one that's over-investing. And one that's likely spending more than money as it clings to the status quo.
Posted by Jill Dyche So many chapters, so little time
31 AUG 2006 20:41 EDT (00:41, GMT)
When you write a book, you ultimately don't get to say everything you want to say. Little things like scope, clarity, space limitations and the omnipresent deadline get in the way. Thank God for blogs on portals like TechTarget that let us continue the conversation.

There were a host of topics in Customer Data Integration: Reaching a Single Version of the Truth that Evan Levy, my co-author, and I wrote about that could have been books in their own right. We had a chapter on Data Governance that we ultimately had to cut short -- we had that much to say about it. Chapter 5, which focuses on top-down versus bottom-up development techniques for CDI, could apply to scads of different IT projects. Master Data Management (MDM) -- a topic that was woven throughout our CDI book, since CDI is a subset of MDM -- could be its own book, albeit with lots of different chapters, since master data by subject area looks very different and is beholden to wildly different algorithmic processes and access.

One topic we write about in the book that deserves more discussion is identity resolution. My feeling is that this is an underrated function in CDI, and one that can make or break, not only a CDI implementation, but can determine the success of a CRM effort or a compliance program. Identity resolution is more than just name searching and matching. It applies indexing and fuzzy logic to individuals who might not want to be recognized as one person, or to companies that are interrelated either actively or passively. Understanding these relationships -- and recognizing the entities involved in them -- has implications to customer relationships, financial transactions, chart of account detail, and even to homeland security.

Identity resolution is complex. The companies that know they need it often experience an "aha!" moment. One CMO we interviewed for the book told us, "When we realized we had only 80 percent of the number of customers we were reporting to The Street, we all just stopped and looked at each other, and our President looked positively stricken."

Indeed, most companies that need identity resolution don't really understand that they need it until there's a really big problem to solve. This could be a book in itself. Hey... Evan?
Posted by Jill Dyche Data, thy (thine?) time has come!
30 AUG 2006 10:31 EDT (14:31, GMT)
Thanks to those of you who have asked questions in the Experts Answer Center. The dialog has been great so far.

The subtext of some of the questions you've seen -- and that undoubtedly you'll be seeing -- is that people's knowledge of their companies' data is confined to whatever they've been working on, but not for long. The data warehousing people understand integrated data, often better than anyone else in their firms. Application developers are in the throes of actually integrating data for operational purposes, which can be quite different. And there's the data administration group, for whom the term "data management" becomes a catch-all phrase for any type of data "heavy lifting" for which incumbent skill sets are meager. Data modelers continue their toils, under-appreciated. Everyone agrees that data should be broader and more overarching than their little piece of the pie.

And management is starting to see the light. Data -- its importance, its strategic value, and its cost -- is more well-understood than ever in the C-level ranks. Data's not only out of the closet, it's prancing around from room to room.

So the phenomenon of data hoarding that I've written about in all three of my books and countless magazine articles and blog entries is giving way to new kind of corporate edict: Thou shalt share thy (thine?) data. It was so hard to get managers to even consider this a few years ago, and now they're happily laying down the law. What's changed?

In a word, a lot. (I know, it's technically two words, but I'm makin' a point here.)

The number of data sources has increased exponentially. Data volumes have exploded. New packaged applications are proliferating. Our data warehouses have hit capacity. New technologies like data warehouse appliances and CDI hubs have made data integration cheaper and easier. The line is blurring between operational and analytical systems, and compartmentalizing processing platforms is getting dangerous. The need for good, clean data transcends it all.

As the problems get hairier, our executives have begun wrapping their brains - -and their budgets -- around solving them. And if the lords of the realm can wrap their brains around the solutions, the laity (that's us) sure as hell can!
Posted by Jill Dyche MDM on the rise
29 AUG 2006 12:10 EDT (16:10, GMT)
It's funny. With all the talk about Master Data Management (MDM), it's still in its infancy at a lot of our clients. We've seen companies treat MDM almost like a science project. Let's combine these ingredients -- say a dash of executive management for data governance, a pinch of business process knowledge, a quarter cup of data modeling expertise, and a soupcon of new technology -- and see if it explodes.

The companies that are taking MDM seriously and acting on it are the ones that have critical operational applications -- like CRM and ERP systems -- that are dependent on integrated data and are finding that they've been investing time and resources in order to support one-off data integration efforts. Take a company like Amgen, which is a case study in our new CDI book. They have multiple relationships with a given physician -- and already had a powerful technology infrastructure to pass messages between systems. Their CDI solution gives them a way to reconcile and integrate the data between those systems and avail it to other systems. It prevented a lot of re-work, it ensured data that was clean and well-integrated, and it leveraged the company's existing IT infrastructure. It's the companies with a well-articulated business need for reconciled and integrated data who are the early adopters of CDI -- and, by extension, MDM.
Posted by Jill Dyche Why CDI?
28 AUG 2006 20:20 EDT (00:20, GMT)
First off, I'd like to say that I'm psyched to be a TechTarget expert. Hey, these things don't come around very often, and if this is a way to proselytize what's happening in the crazy world of Master Data Management and Customer Data Integration, then I'm down for it. I hope you all use this Expert Answer Center to head check your own project strategies, debate your colleagues about best-practice approaches, and get some thorny questions off your chest.

I thought I'd start the blog off by answering a question I get a lot: "Why did you and Evan (Evan Levy's my partner at Baseline and co-author of my new book, Customer Data Integration: Reaching a Single Version of the Truth) choose CDI for the topic of your new book?"

There are a couple answers. First, because it was the culmination of my two other books. My first book, e-Data, talked about the value of enterprise data to business, focusing on data warehousing, but ultimately presenting the benefits of business intelligence as a strategic program. My second book, The CRM Handbook, drilled in on CRM from different perspectives, from sales force automation to smart target marketing to predictive analytics.

After these two books, it was interesting to see what companies had mastered -- and what they still couldn't do. This new book is about breaking the barrier of data integration as a "data warehouse-only" capability, and making it an enterprise-wide function. CDI is bigger than integrated data on a single platform or application. When done right, it can establish consistent data management and avail a single version of the customer truth, not only across systems, but divisions and individuals, in a sustained and meaningful way.

I'm looking forward to discussing the "whats" and the "hows" of CDI and MDM in the Experts Answer Center. So bring 'em on!
Posted by Jill Dyche