Inside the Biz with Jill Dyche

Understanding Social Personas

In which Jill reveals her secret. (Okay, one of her secrets.)

There are some things that only a few good friends know about me. For instance, I like running on the beach. I watch Animal Planet religiously. I enjoy napping in the back of the station wagon, even when it’s parked in the garage. And—I’m not proud of this, but here it goes—on occasion I eat my own poop. The secret is this:  I’m really a dog. And I have 800 Twitter followers.

Don’t go looking for these factoids at www.twitter.com/jilldyche. That’s someone else. Or IS it?

You see, when I started using Twitter I had to reconcile different areas of my life. I like to tweet about professional stuff, including customer relationship management (CRM), social media, and information as a strategic enabler. I also like to chat and share info with my non-work friends, many of whom are hiking buddies, dog rescue advocates, or fellow charity volunteers. Sometimes my work tweeps aren’t necessarily aligned with all my goodwill activities, nor do they always care about my doggie sagas. (Amy is blind, Lu is 14 with vestibular disease, and Colby has epilepsy.)  Likewise my non-work buddies think MDM is simply the acronym for the new slogan, “Mutt Dogs Matter!”

Since Colby comes the closest of my three dogs to having opposable thumbs, I set her up on Twitter. And thus I adopted a new social persona that lets me communicate with a subset of my friends in a more meaningful way, without bothering the rest with humdrum observations or irrelevant news.

[Hiking and dog buddies: go ahead and skip the following paragraph.]

Sound familiar? The definition of CRM (Customer Relationship Management) is the ability to differentiate customers in order to distinguish their treatment. That applies to our sales force, our customer service reps, and yes, to marketing. Why e-mail a customer a discount code for a product he already has? Why ship a catalog aimed at newly-marrieds to a single grandmother? Why pitch cat food to… You get the idea.

Your social persona means more than just having the answer to “Which Superhero are you?” or “Which Barbie are you?” (No, I’m not kidding.) It represents a circumscribed collection of behaviors and preferences, not necessarily your entire set of behaviors and preferences. Social personas—alright work friends, personae—encompass similar attributes and are more narrowly defined. Our social personas help us refine our networks, participate in new on-line communities, and invite fresh collaborations. Social personas essentially allow customers to segment themselves. And they’ll segment themselves in a far more accurate and nuanced way than a marketing department ever could. The implications on sales uplift and return on marketing investment could be staggering.

Indeed some savvy companies have discovered Colby, and are pitching her products according to who she is and what she, er, says. Colby buys bully sticks from Value Pet Supplies (@valuepet); attends fundraisers on behalf of Best Friends Animal Sanctuary (@BFAS); browses the aisles of Red Barn (@redbarnfeed); and has even been known to watch Vet TV (@veterinarytelevision). Colby is a Coonhound, but she’s also a consumer. Those with a stake in the $45 billion pet care industry are cultivating relationships with her.

[Hikers and dog friends: go straight to the end.]

This adds a new wrinkle to social CRM, especially to the already challenging practice of customer segmentation. Most companies still struggle to transcend the well-worn practice of demographic segmentation, seeking to enrich and integrate their customer data to enable more sophisticated behavioral, preference, and usage-based segmentation and with them more intelligent ways to engage and interact with customers and prospects. Persona-based segmentation should be next on their lists.

The hard part of social personas, of course, is establishing the connections between social identities. The very hard part is collecting and structuring that data, then integrating it with the rest of the customer’s profile in a sustainable way.  One could argue that companies have yet to accomplish this successfully with their on-premise structured data, never mind social media data.  But the time to start planning is now.

Among Colby’s followers are @BuckBasset, @Otis_the_Pug, @WinterKitten, and Governor Arnold Schwarzenneger (@Schwarzenegger). You can follow Colby on Twitter at www.twitter.com/Colby_the_Dog. She’ll follow you back just as soon as she finishes her bully stick.

Ode to Initiate Systems

In which Jill pays homage to a class act.

The rumors of IBM’s acquisition of MDM vendor Initiate Systems were confirmed yesterday in an IBM press release.  Herewith, a brief homage.

A few years ago Initiate Chief Scientist Scott Schumacher dropped by our L.A. offices to introduce his fledgling company and to share some MDM successes. If you don’t know Scott, this is like saying that Robert Deniro showed up to block a couple scenes, or that Jimmy Page needed to borrow your guitar to try out a new riff. Needless to say, I took notes.

Upon learning of our forthcoming book on CDI, Initiate referred Evan and me to several of its customers. One was Marty Moseley, at the time Intuit’s Chief Architect. These were the days when people still couldn’t articulate the difference between an MDM hub and a data warehouse. I asked Marty how he made the internal pitch for MDM at Intuit, and his answer was refreshingly plainspoken: “We have SOA here at Intuit. It helps us make sure that our systems do what they’re good at. Why would I use my data warehouse to do MDM? That’s like saying SAP can do everything. We needed to operationally integrate our customer data. So we bought a hub.” Marty went on to become Initiate’s CTO and a good friend.

Where other vendors hoard contacts and micro-manage their messaging, Initiate made transparency part of its culture. When our book was published CEO Bill Conroy personally offered his support and that of his management staff, and would follow up with personal notes of congratulation and regular company updates. Bill knows how to build a good management team too, and Gina Sandon and her marketing team fell over backwards to provide me with industry research. Ian Stahl, Andrea Zivalich, Michelle Woolfolk, Rick Clements, Jim Cushman, Lorraine Fernandes, and Greg Shaw have all been similarly supportive and impressive.

In addition to being an acknowledged leader in healthcare (most notably, master patient index) and government (person-of-interest recognition), Initiate has blazed a trail across a range of industries, geographies, and value propositions. Its matching algorithm is arguably second-to-none, it led the way in defining the concept of exchanges, and its data governance and stewardship offerings were first out of the gate.

Initiate’s acquisition means that its new parent company can now make good on those Sunday-morning network commercials that flaunt its leadership in health care and information integration. I do predict that Initiate will be more difficult to absorb into IBM than its rival Siperian will with Informatica. I’ve spoken to two Initiate Customer Advisory Board gatherings and the consensus among heavy-hitter customers is that they invested in Initiate precisely because it wasn’t a one of the big enterprise app vendors.

In the dark of night most owners of MDM programs will admit that they ultimately chose their MDM solution because the vendor was already incumbent, or because the incumbent vendor was willing to provide a steep discount (read, free) in exchange for a reference. But Initiate customers tell a different story. “We bought Initiate because of [insert Initiate employee name here]. And because they weren’t [insert large vendor name here].”

Don’t eff it up, IBM.

The Good News about MDM Market Consolidation

In which Jill stays high on MDM.

I usually root for the little guy, so when I first heard the rumor that there were multiple bidders for Siperian, I figured that SAP would take it home and that it really was a three-vendor world. With revenues passing the half-billion mark, it seems that Informatica is joining the ranks of the big guys after all.

I agree with those in the analyst community who say that Informatica’s purchase of Siperian will be good for both companies. (Check out great posts from Rob Karel and Merv Adrian.) Last year, Informatica’s MDM story verged on the schizophrenic as the company simultaneously advocated a “roll your own” approach to MDM using various software components while at the same time making investments in both Siperian and rival Initiate Systems.

Siperian fills in some significant voids in Informatica’s MDM capabilities, most notably hierarchy management and transaction integration—updating the golden record in real time. While Informatica supports message-based data passing in real time, with Siperian, real-time data requests will soon be a reality. In return, Siperian will now have the power of the Informatica sales force and a new set of prospects in the data integration company’s fiercely loyal customer base.

When Evan and I were writing our book on Customer Data Integration and MDM, we knew we had to write about “the stack” and we knew—despite comprehensive chapters on data governance, data quality, and implementation methodologies—a lot of text would smack of back-office infrastructure. Let’s be honest. Companies that invested in software from Purisma (before D&B acquired them) and Stratature (before they were snapped up by Microsoft) were early adopters. IT organizations that bought these toolsets were laser-focused on narrow, yet high-impact, business problems, and were willing to invest in targeted solutions from smaller niche players.

So we wrote a chapter on making the business case in which we presented a litany of industry-specific MDM value propositions. The truth is that MDM is a business solution, and vendors should sell it accordingly.

Whether you’re tackling state reporting compliance in the pharmaceutical industry, account consolidation in banking, master patient index in healthcare, channel obliteration in retail—the lists go on and on—MDM can solve the problem more quickly and effectively than custom code. Or—as now Informatica would admit—than an ETL tool. As Siperian’s multiple pharmaceutical customers will attest, the ROI for MDM is in the industry business problem.

Informatica CMO Chris Boorman’s spirits were high when I talked to him on Friday. “We’ve thought a lot about MDM as a key piece of our overall data integration strategy,” he said. “And we’ve embraced Siperian’s concept of a ‘trust framework,’ for data. The ability to assign a confidence level to data is a differentiator.”

My confidence level that other MDM acquisitions are on the horizon? High.

Photo by Sreejith K via Flickr (Creative Commons License)

Could Data Governance Help the War on Terror?

In which Jill wags her finger at the silos—ours, theirs, and yours.

“Our intelligence community failed to connect those dots, which would have placed the suspect on the no-fly list."
-- President Barack Obama

President Obama addressed the Christmas day terror attempt in his White House briefing this afternoon. He acknowledged that the U.S. government had botched terrorist tracking and that the system had failed. He admitted that each agency took responsibility for its own stuff. He also confessed that there was a lack of information sharing, that various government agencies had failed to pull it together. The subtext was that current practices and the best of intentions aren’t enough when it comes to the war on terror.

Obama sounds like a lot of corporate executives I’ve met. They admit that the individual departments are working hard and may be meeting their individual goals. But then something stupid happens. The company loses its largest account or a competitor beats them to the punch on a new product. The hard fact emerges that the company could have used its information better. And that siloed organizations beget siloed data.

The alleged terrorist, Umar Farouk Abdul Mutallab, had been in Nigeria, Yemen, the United Kingdom, and The Netherlands before flying to the U.S. Each of these governments had a slice of information about him. The UK had enough information to put him on a watch list. But the information was never integrated into a rich enough profile of Abdul Mutallab for the FBI to mark him as a person of interest. As Informatica marketing vice president Chris Boorman wrote in his excellent blog post, this would have been an apt use of identity resolution software. Various government agencies could have matched partial information to form a holistic view of Abdul Mutallab, one that could have been shared, thus pinpointing him as a suspect. With the right collaboration, this could happen at both federal and international levels.

There are over a million names on the U.S. terrorist watch list, and in fact the intelligence community had information that an attack on the U.S. might be imminent. White House Press Secretary Robert Gibbs insisted that since apprehending Abdul Mutallab, they had gleaned “usable, actionable intelligence” from him. Given their track record, you can’t help but wonder how American intelligence agencies will use this information. 

“It is increasingly clear that intelligence was not fully analyzed or fully leveraged,” Obama conceded. “That’s not acceptable, and I will not tolerate it.” As much as you believe him, you’ve got to wonder what he intends to do. Sure, he inherited a bunch of crap, but it’s crap we hired, er, elected him to fix.

The conversation in the press has been the same conversation that occurs in corporate conference rooms across the world after a major screw-up: “Who did the oversight?” and “Who should get fired?” “Growing questions about if and when heads will roll!” trumpeted CNN’s Wolf Blitzer. In a crisis, attention seems to shift to the people issues.

Instead, the government should be addressing process issues. Indeed, the real conversation should be how to move forward. These questions should be asked now: “How should we bring identifying data together? What are the key sources? How should integration, access, and usage policies be formulated? What would a sustainable process look like?” Those questions aren’t “who” questions, they’re “how” questions, and they should be front-and-center in the national security conversation.

Maybe they should also be part of your next conversation with executives or business users? The Christmas day terror attempt is one of many, many examples that the opportunity cost of not having data governance could be high, and the consequences nothing short of catastrophic.

Are You Really Ready for Data Governance?

In which Jill makes excuses for people who make excuses.

Fall conference season has (mercifully) come to an end, and I can finally reflect on what I learned on my most recent tour of industry and vendor events.

(Editorial aside: I don’t know about you but I like Vegas, I love New Orleans, and if I see Orlando again it’ll be too soon. Note to conference organizers as you finalize your 2010 event planning: The Big Easy still needs our cash!  Let’s go back. The jambalaya’s on me!)

Anyway, I’ve been speaking and writing a lot about data governance, and presentations at TDWI, SAS, Teradata, and various clients have placed me squarely in the crosshairs of people aiming for effective data governance programs. It’s interesting to hear their questions, particularly as they voice concerns about data governance adoption in their organizations. One question I get in almost every session is a variation of this one:

Alright Miss Fancy Pants, why is it that every time a so-called expert talks about a major business initiative they bring up change management? Why can’t we just launch data governance without all the change management fanfare? Why can’t we just ‘get ‘er done?’

Answer: I don’t know, why can’t you? Or, more aptly, why haven’t you?  Not trying to be argumentative here, but 1) my pants are just black wool crepe, trouser-cut, which I think is more understated than fancy, and 2) it’s just that the most effective data governance programs are often also the most disruptive. And, simply put, disruption makes people nervous.

Professor Robert E. Quinn at the University of Michigan Ross School of Business writes about taking people out of their “normal states” (aka, their comfort zones) and painting a picture of what they need to do. In a seminal Harvard Business Review article in 2005 called “Moments of Greatness: Entering the Fundamental State of Leadership,” Quinn described true leaders as ready to “venture beyond familiar territory to pursue ambitious new outcomes.” If that’s not data governance, I don’t know what is.

After all, the “normal state” in most companies often means contradictory policies, multiple versions of the truth, absence of true data ownership, selective collaboration, and decision-making in a vacuum. Data governance is the opposite of all that.

“We couldn’t afford to experiment with data governance,” says Kevin Davis, Director of Research Initiatives at a major southeast retailer. “We retained a consulting firm to teach us the foundational stuff and we followed their advice. That included a lot of deliberate structure, education, evangelizing, and expectations management.” Davis and his team faced the additional challenge of coupling their company’s new data governance strategy with a major MDM technology acquisition. “We knew we only had one shot at getting it right,” he says.

True, it’s easier to avoid sticking your neck out. And in some cultures (particularly in high-tech companies, but that’s another blog post altogether) wholesale changes to policies and processes can actually begin from the bottom-up. It’s not only less disruptive, it demonstrates value to incent widespread adoption. Bottom-up data governance is a legitimate and proven approach. But at the end of the day, your strategy needs to include a plan to broaden governance beyond the initial domain. After all if you’re not sharing data across organizations and business processes you don’t need data governance.

According to Professor Quinn, change happens when someone is “jolted out of his comfort zone…driven to clarify the result he wanted to create, to act courageously from his core values, to surrender his self-interest to the collective good, and to open himself up to learning in real-time.”

Are you ready to get out of your comfort zone? Then maybe you are ready for data governance!

(P.S.: I’m discussing data governance yet again with Initiate Systems CTO Marty Moseley in a December 3rd webcast. Register here!)

Photo by p_x_g via Flickr (Creative Commons License)

The Netflix Prize: Customer Intelligence for Hire

By Jill Dyché

In which Jill wonders if more data could have rescued her from Mama Mia!

Movie rental by mail (and now by streaming video) company Netflix has been celebrated for its disruptive business model. Woe to Blockbuster, Hollywood Video and the other industry dinosaurs that have regressed to slow follower status, lumbering behind Netflix’s innovation.

That innovation has included the company’s lauded Cinematch function. Cinematch is the Netflix version of Amazon’s purchase circles, a recommendation engine that predicts what you might like to buy next. Cinematch is how Netflix knows that because I like the movie “Almost Famous” I will also like the movie “Lords of Dogtown.” (I did.) Somewhat counter-intuitively, it also knew that I would like the movie “Rabbitproof Fence.” (I did.) Cinematch doesn’t really care what people in my neighborhood are watching, which is fine by me since most of my neighbors are twice my age and tend toward the black-and-white whodunit genre. Of course, I’m guessing on that one, which is my point: Cinematch knows.

But according to Netflix, Cinematch could be even better. The Netflix Prize offers $1 million to the first person or team that can improve Cinematch’s current accuracy by 10 percent or better. The contest was announced on October 26, 2006, and no one has won yet. (So far a team called BellKor in BigChaos is in the lead, delivering a 9.63 percent improvement.)

Why is the problem so vexing? We’ll leave discussions of collaborative filtering, linear regression algorithms and singular value decomposition for another time. One of the most buzzed-about reasons is the so-called Napoleon Dynamite problem. The movie is an outlier: it’s hard to categorize into a single genre. Moreover people either love it or hate it—“Napoleon Dynamite” gets a disproportionate number of 1-star and 5-star ratings. Such a wide preference disparity for a film with such a singular concept can flummox the cleverest predictive algorithms.

The fact is that Cinematch is bound by a singular limitation: a dearth of heterogeneous and comprehensive data. Cinematch doesn’t rely heavily on geography, age, or gender to predict what movies you’ll love, indeed it strips out personal details. Instead it uses information about other movies you’ve liked in the past. The smaller the number of descriptive attributes, the smarter the data mining engine has to be.

Our efforts to understand customer behaviors and preferences are easier. Or at least, they should be if we’re collecting a diverse set of granular-level data about our customers. And we need to standardize and cleanse that data so that it’s consistent. The more information we have about our customer and our products, the more we’ll know about them. That is, the better we’ll understand clear determinants of behaviors and preferences, and craft messages and conversations around them. .

Do the math and it’s clear that upping Cinematch scores is worth vastly more than $1 million to Netflix. What’s the ROI for better customer knowledge at your company? If you don’t know, then you don’t know what you should be spending, and odds are you’re probably not spending enough. Of course I don’t know the precise likelihood that you’re not spending enough, that would call for some analysis. Or we could just go to the movies.