In this interview Jim Harris of OCDQ Blog talks with Jill Dyché, a partner and co-founder of Baseline Consulting, a management and technology consulting firm that provides business intelligence, data governance, and technology delivery services.
Jill is responsible for delivering industry and client advisory services, is a frequent lecturer and writer on the business value of information technology, and writes the excellent Inside the Biz blog.
Jill is the author of three acclaimed books: e-Data: Turning Data Into Information With Data Warehousing,The CRM Handbook: A Business Guide to Customer Relationship Management, and her latest book, written with Evan Levy, is Customer Data Integration: Reaching a Single Version of the Truth.
Update: Baseline Consulting was acquired by SAS DataFlux, Jill is now VP Thought Leadership at SAS.
Jim Harris: Two of the guiding principles of Baseline Consulting are "understand that business-driven IT projects are both science and art” and "let the team doing the work make its own rules.”
I wholeheartedly agree. However, it seems like many are looking for a complete framework to blindly follow and some are more than willing to sell it to them, often resulting in a lot of wasted time and money. Do your clients initially expect an "all science” and "here are all the rules you must follow” approach?
Jill Dyché: I love you for asking that question because part of what Baseline’s guiding principles ensure is that we "meet our clients where they’re at.”
To your point, this often involves having to undo certain assumptions that arise from some pretty broken corporate cultures. Assumptions like:
We already know which vendor product to buy—it’s the one our competitor bought
Come in and tell our executives why data is important
The hardest part of a data quality or governance initiative is often doing the arduous preparation slog to set up the program the right way. And that often involves more than just meetings and conversations — it means doing research on what’s worked, what hasn’t, and developing customized approaches for moving forward in a deliberate and business-driven way.
It's absolutely critical to discover a way that fits your culture, but drives change at the same time. It’s not for the feint of heart.
Jim Harris: When speaking at conferences, you often explain that the hardest part of data governance isthe data. I believe you are working on a new Data Governance book. Will you be writing about the role of data quality in a data governance initiative, and vice versa?
Jill Dyché: I ended up turning down the data governance book offer. There are really good books out there already—Tony Fisher,Gwen Thomas, and Steve Sarsfield are all flat-smart and I wouldn’t want to compete with any of them. I recommend their three books very highly. I’ve just contributed an MDM section to Paul Greenberg’s new CRM book and I’ll be writing the forward to Phil Simon’sf orthcoming book on emerging technologies, so I’ve got my hands full.
Frankly I’m more interested right now in applying data governance frameworks and best practices to actual client situations. The cool thing about being a consultant is that, while there are rules for a lot of this stuff—we’ve defined some of them—it’s never ends up the same from one company to another. So where data governance is concerned, I’m spending my time out in the field with our client teams. We’re helping companies design custom data governance programs. It’s fun.
Jim Harris: In a recent B-eye-Network article, When Data Can’t Be Trusted, Master Data Management Becomes "Plan B”, you explained that "master data management is not a new solution to an old problem, but rather a new solution to a new problem.” I think your article addressed a common misperception about MDM. Can you summarize the key points?
Jill Dyché: When MDM was first getting buzz, the predictable reaction was from its detractors was, "This isn’t new. We’ve been doing it for years.” But those people didn’t really get the holistic vision of MDM. They dismissed it as conformed dimensions or metadata or even enterprise data modeling. But if you look at what the MDM products actually do, they bake in functionality that was heretofore cobbled together. Identity resolution, real-time matching, data correction and enhancement, source system linkages—the MDM stack is varied and pretty unique. And it’s made a heck of a lot easier with SOA. MDM can provide—often for the first time—an authoritative source for data to not just users but other systems. Doing that in a homegrown way is no cakewalk, and it’s less necessary than it used to be.
Jim Harris: A recent post on the Gartner blog network asked: Can you "do” MDM without data quality? We have already discussed data governance, how does data quality relate to MDM?
Jill Dyché: If you’re down with my "authoritative source” definition, then you agree that data quality is baked into MDM.
Why "do” MDM if the data’s not going to conform to requirements and be accurate and meaningful?
I think Andrew was playing loose with Jim’s assertion about after-the-fact harmonization, which not only involves data correction but matching and linking back to heterogeneous source systems. That means combining some core data quality functionality with some complex processing logic to close the loop with operational systems. It includes data quality, but it also transcends it.
Jim Harris: When discussing business intelligence, you replace the "People, Process, and Technology” paradigm with the more comprehensive "Information, Strategy, Organization, Processes, and Technology.” Can you summarize these critical success factors?
Jill Dyché: It’s part of my devious campaign to get people to stop mindlessly chanting the "People, Process, and Technology” mantra. We heard People, Process, and Technology with data warehousing, with CRM, and now we’re hearing it with social media. By adding Information and Strategy into that mix, I’ve made it my secret mission is to get executives to think a little more. We want them to get that BI done right can drive accurate information into the business, and make it more consumable. Companies need to understand that data is more than just a by-product of the systems that generate it. It is—and here’s another mantra I’m going to have to take on pretty soon—an asset in its own right.
And as far as strategy is concerned, I’m a big believer in linking BI to a company’s strategic objectives. As BI programs evolve it’s increasingly likely that information that supports BI will be used strategically. Even organizations that start with some pretty basic operational queries see their data evolve to the point where they can ensure that customer feedback via social media makes its way to R&D to drive product refinements; or that historical purchase behaviors can distil competitive acquisitions; or that delivery analysis can drive preferred suppliers. Plus, this is the optimal way to bait the hook for executives, getting them to see BI’s long-term value.
Jim Harris: IT-business alignment is a common category for posts on your excellent Inside the Biz blog, including Dear IT: A Letter from Your Business Users and A Letter to Our Users, From Your BI Team. What recommendations can you make for facilitating the collaboration of business and technical stakeholders on enterprise-wide initiatives?
Jill Dyché: Well, speaking of strategy, using corporate goals and objectives to prioritize BI initiatives is a great way to get everyone to the table (Baseline invented the concept of the BI Portfolio™ just for this purpose). IT needs the business to communicate and explain those business objectives, and the business needs IT to understand the "hows” of enabling them. And—this is the squishy stuff, but here goes—people like collaborating around a larger purpose. Aligning teams around corporate strategy is easier than arguing over counts and amounts in last month’s revenue spreadsheet. Plus, the wins are bigger when they happen.
Another way to foster business and IT collaboration is to share success stories from other industries. One of our clients, CheckFree, won the TDWI Best Practice Leadership award a few years ago, and I recently told their story to a healthcare organization. It didn’t matter that CheckFree is a financial services company (now owned by FiServ). The healthcare CIO saw the parallels between CheckFree’s journey and what his HMO needed to do. Take a company like Harrah’s, which I profiled in my CRM book. What a great example for using customer data to engage customers and drive traffic to the properties. That goal certainly isn’t exclusive to the gaming industry—it applies everywhere!
Jim Harris: Your blogging style is a natural extension of your personality, which engages readers and makes them feel like they’re having an in-person conversation with you. As someone who is also an accomplished author and public speaker, can you share your insights on how blogging compares and contrasts with writing and speaking?
Jill Dyché: Wow. Now I’m reviewing my answers to your interview questions to make sure I’ve been engaging and conversational. I’m giving myself a B minus. Let’s talk about our favorite wines. I just got turned on to a heady little Sancerre...
Seriously, my writing and speaking don’t really vary much until I find myself using words like "penumbra” and "jejune” and then I have to go back and check myself. I think if your personality comes through, the concepts you’re communicating are also elucidated, er...I mean clearer. It’s not something you can fake.
Jim Harris: Have any of your clients started using social media as an additional customer contact mechanism? Do you think your clients will ever view social media (e.g. blog posts and comments, status updates from Twitter and Facebook) as a customer data source?
Jill Dyché: Most companies are adopting social media in one form or another. The question is: how are they measuring its value? Most of my clients see social media as either an extension of their brand, or as a means of low-cost customer support. Both are valid uses, of course. But the question is how to communicate, plan, and realize value from social media. Companies like Comcast and Ford have high-profile social media executives, and they’ve been celebrated.
But a lot of the stories are still very anecdotal: a customer was ticked about spotty service and we got her back on-line in 30 minutes. Someone on Facebook threatened to sue us and we were able to find him and talk him out of it. Someone said something disparaging about our brown sugar cinnamon frosted toaster pastries, so we sent her a free box. (Okay, that one was me and it was a transparent ruse but it worked! Bwaaa haa haa! Chomp chomp...).
[Dabs mouth] But to answer your question, yes our clients are already seeing social media as a data source, but until they measure it to distinguish the useful information from the rest of the noise, they risk overinvesting. Let’s face it—the majority of companies have different customer counts across departments and continue to struggle integrate customer profile data. So most aren’t ready for social media analytics.
Jim Harris: You often quote the venerable wisdom of Homer Simpson: "You don’t win friends with salad.” Please enlighten us.
Jill Dyché: Salad is a metaphor for not-the-main-course. When it comes to things like data quality, we have to get straight to the meat. Take things seriously. Bite off a chunk and chew, chew, chew. Be serious and earnest about it. So, no, you don’t win friends with salad.
However, in my experience, you do win friends with wine. Let me tell you about this little Sancerre...