MDM In 2012: What Was, What Will Be . . . And What Won’t Be

Happy New Year! As we kick off 2012, I’d like to reflect on what was accomplished during the past year in the “trusted data” areas of master data management (MDM), data quality (DQ), and data governance and consider what we might expect in the year to come. I also hear quite a bit of noise from vendors and analysts alike about what they want the MDM market to be in 2012, so I wanted to share my thoughts on what’s real and what (in my opinion) remains hype. 

I also just completed Forrester’s December 2011 Global MDM Survey of 274 MDM professionals. While the majority of those results will be shared in the annual MDM Trends research that I’ll be publishing later in Q1, here’s a taste of some of the intriguing results.

Let’s first reflect on what I’ve witnessed from my clients MDM journeys throughout 2011:

  • Data governance remained a challenge. In the abovementioned MDM survey, only 20% responded that they have a high or very high level of data governance maturity, indicating that significant work remains. But on the positive side, I’m witnessing increasing business sponsorship and prioritization, which has helped many organizations pilot programs to cut their teeth and build some repeatable processes, foundational policies, and early measurements to start building a case to increase data governance investment and momentum.
  • Multidomain MDM hit its stride. User interest in multidomain MDM strategies has finally caught up with vendors’ product capabilities and messaging. In Forrester’s MDM Survey, 47% responded that the scope of their MDM programs include more than two data domains to master, while another 9% are focused on dual-domain solutions (e.g, customer and product).
  • Business process and master data alignment became a no-brainer way to position data priorities. It’s been almost 2½ years since my colleague Clay Richardson and I first published research on our process data management approach, which aims to better coordinate data management and business process management (BPM) efforts. The fact that business processes and decisions require trustworthy data, and that data improvement initiatives should be framed in the context of the business value it generates is a no-brainer, right? If that’s the case, why do so few organizations actually coordinate these efforts? Well, this message has been very widely embraced; most MDM and BPM practitioners, vendors, and analysts alike now at least recognize that they need to improve their abilities to place master data efforts in the context of critical business processes to encourage business collaboration and sponsorship.
  • There was minimal MDM market movement in terms of M&A or consolidation. In 2010, we saw major movement in the MDM and DQ landscape with the acquisitions of Siver Creek Systems by Oracle, Siperian by Informatica, Initiate Systems by IBM, Netrics by Tibco, and Data Foundations by Software AG. In comparison, the MDM M&A market in 2011 was nonexistent. Beyond Oracle’s smallish acquisition of Datanomic to finally wrap customer DQ capabilities within its own portfolio, we haven’t seen any other significant activity of note. (I’m not implying there have been no other small acquisitions, OEMs, or other partnerships, but in my opinion nothing that significantly affected the market — let me know if you think I missed something!).

With 2011 wrapped up, here’s what I expect to see in 2012:

  • M&A acquisition activity to return. Macroeconomic conditions permitting, I specifically see a few M&A trends as a high potential in 2012. These include:
    • Governance-driven MDM solution consolidation.  Also known as analytical MDM, I’m referring to solutions like Kalido and Orchestra Networks who traditionally specialize in version-controlled hierarchy management with embedded stewardship workflows across multiple data domains. These solutions don’t compete directly with operational MDM hubs, but could be a nice complement for hub vendors, just as Oracle’s Data Relationship Manager (DRM, which came to Oracle via its Hyperion acquisition) has been a powerful complement for its MDM hubs. I could see IBM, Informatica, SAP, Software AG, or Tibco making a move here.
    • Product information management (PIM) consolidation. While many of today’s major multidomain MDM vendors began their MDM journey years ago with acquisitions of PIM vendors (e.g., IBM acquired Trigo, SAP acquired A2i, Teradata acquired MDM assets from i2, and Tibco acquired Velosel), there hasn’t been a significant PIM acquisition of note in years. With the multidomain wars significantly heated at this point, the ability to compete effectively across all data domains will be increasingly important. Potential acquirers over the next 12 to 24 months will include DataFlux, Informatica, Microsoft, and Software AG. Potential targets could include Heiler, Hybris, Liaison Technologies, Agility (formerly Pindar), and Riversand Technologies.
    • Industry-specific solutions. Surprisingly, the MDM market remains immature across industries when it comes to highly specialized vertical solutions that can provide deep functional and scalable MDM capabilities. But there are exceptions, especially in the financial services and securities markets where a number of solutions have reached high levels of credibility. These include Asset Control, Cadis, Eagle Investment Systems, Golden Source, and Kingland Systems. Potential acquirers here would most likely be limited to the Big Four: IBM, Informatica, Oracle, and SAP.
    • Complementary functional acquisitions. In addition to the pure MDM consolidation implied in the first three bullets above, I also wouldn’t be surprised to see MDM vendors acquiring technology to enhance capabilities in BPM/workflow management, enterprise search, data warehouse (appliances), and analytics.
  • Stewardship process automation. Whether this comes from MDM technology vendors or organizations leveraging their existing collaboration, workflow, and social technologies, expect to see more end user investment in enabling data governance functional responsibilities such as standards, rules and policy definitions, exception handling, and notification/approval workflows. In Forrester’s MDM Survey, 66% of MDM practitioners already use technology enablers to support their data governance processes, regardless of whether they use capabilities provided by MDM or BPM vendors or build custom solutions on their own. The market opportunity is here and growing.
  • Data quality-as-a-service (DQaaS) options increasing. I expect more and more traditional, on-premises data management software vendors (as well as smaller startups and pure-plays) with DQ capabilities to test the cloud waters by offering pay-as-you-go or subscription DQ services: tactical data cleansing, verification, and enrichment capabilities like those provided by D&B 360,’s, and StrikeIron. This will be a necessary precursor to true DQ and MDM platforms in the cloud (see below).

Equally as important, all hype aside, here’s what MDM will NOT be about in 2012:

  • MDM in the cloud? Not yet. Aside from classic hosted MDM scenarios (i.e., hosted database marketing service providers (MSPs) supporting direct marketing campaign management and analytics; hosted or BPO PIM solutions), most customer and multidomain MDM solutions remain on-premises. While 23% of respondents to Forrester’s December 2011 MDM survey said they would seriously consider or are very interested in cloud-based MDM solutions, 63% still said they are most likely or not at all interested. I expect 2012 and 2013 to be years of rational experimentation for cloud MDM, with a few early adopters paving the way for increased interest and momentum in 2014 and beyond.
  • Big data is not yet in play for most MDM programs. According to our MDM survey, only 9% have big data requirements or implications on their MDM road maps, while 72% have not gotten there yet or are not considering it. Big data is absolutely big, but right now the experimentation with it lives with most organization’s BI and advanced analytics function and has not yet evolved into the enterprise architecture-led cross-enterprise capability where many MDM initiatives reside.

I’d love to hear your thoughts, especially if you think there are any major trends I might have missed! And of course, may your 2012 be filled with cleaner data!


Quick Comment?

Excellent post Rob.

I fully agree with you on the section on cloud and big data. We are on a hype moment, but the market is not there yet.
I would have added in the trends an acceleration of the MDM projects velocity and a tendency for projects to get chunked into smaller iterations. In short, long projects (beyond 3-6 month) will no longer be acceptable as business users require to *see the value* sooner than before.

Agree completely

Thanks for your comment, and I absolutely agree with you that most MDM projects are now multi-phase, multi-year efforts for the exact reason you outlined. I don't see this as a new trend though, but more of a continuation of an existing trend. In the past 3 years of my MDM survey, almost 45-50% of respondents have shared that they planned for three or more phases in their MDM initiative (with a significant percentage of respondents that still haven't figured out their rollout strategy!).
Thanks again!

Don’t You See the March of MDM Democratization?

Rob, I agree with the points you highlight, but I think you missed the march of democracy and how it applies to MDM. See this post for details:

Can't agree with you here

Hi Yves, thanks for the comment but I think you're over-stating your case in your post. While I whole-heartedly agree that you folks at Talend have done a great job getting commercial open source into the enterprise ETL conversation (as have the good folks at Pentaho, Apatar and others), I have seen absolutely no real data to prove out your comment that "enterprise open source has democratized data integration & data quality - nobody is questioning this today". Well I guess I am :-).

There are only a small handful of firms claiming to have open source DQ, and only Talend has had the marketing muscle to even promote it. And yet I have not received one inquiry in the past 3 years from my clients asking about any of them. And from an MDM standpoint, I can share that less than 4% of the 274 respondents to my recent MDM survey say they are using Talend for MDM today. (fyi: these respondents are not just the $1bil+ large enterprises that Forrester often supports, 26% have 1,000-4,999 employees, 25% have 5k-20k employees, and 42% have over 20k employees). In addition, it's extremely rare for any of my clients to even inquire about open source MDM options.

And while I'm happy to hear about any midmarket success you're having, but honestly the number of vendors even targeting mid-market MDM is minimal since the likelihood for the heterogeneous complexity that makes MDM a challenge is almost non-existent. Doesn't mean there's not value to be had, but in my mind the midmarket MDM problem is an "Apple" to the enterprise MDM challenges "Orange". The requirements and resulting solutions will look significantly different. Same goes for department implementations at larger enterprises vs. cross-enterprise, global implementations.

So no, I don't believe there is an open source revolution in DQ and MDM that all the industry analysts, my friends Andy, John and Andrew at Gartner included, seem to be missing. What I do think is that open source will continue to be a viable, cost-effective option for small/medium businesses that have significantly lower complexity and could not afford the large costs of most packaged MDM offerings. But I don't believe open source MDM is or should be positioning itself today against the packaged software market leaders. Instead it should be providing alternatives to the almost 25% of MDM respondents from my survey who are relying on home-grown, custom built solutions.

Thanks again for your comments

Do we really disagree?

Hi Rob,
While we will have to agree to disagree on several points, I think our views on the democratization of MDM actually have some commonalities.
First, I am thrilled that your survey revealed 10 MDM projects using Talend. To put things in perspective, how many new clients does an average MDM vendor sign in a typical year? Can't be much more than a dozen or two, given their price tag. And you will probably agree that if you uncovered 10 projects, there are a lot more out there. Keep in mind we've only been out there for 2 years.
Second, I couldn't agree more when you state that "[open source MDM] should be providing alternatives to the almost 25% of MDM respondents from my survey who are relying on home-grown, custom built solutions". This is exactly what democratization is about: bring enterprise-grade MDM technology to organizations that couldn't afford it before, instead of forcing them to use duct tape and rubber bands (or to sell a kidney).
Where our views diverge is on the suitability of commercial open source for enterprise MDM and DQ projects. I know you are skeptical on this topic. And I can't blame you, it goes with the territory. But I am looking forward to providing you more proof points on this topic.

Enterprise-wide MDM

Hi Rob--thanks for sharing--couple of thoughts

There is substantial interest in enterprise-wide MDM in the cloud, but agree that most of the internal thought leadership and likely funding resides in the advanced analytics and BI, not EA, with good reason. One reason is that related practices are conjoined at the hip through data structure as well as enterprise software applications, certainly to include limitations of ERP and databases. The technical issues, however, have either been resolved or are in advanced stages of resolution to the point of pragmatic application. The larger problem I have seen for adoption in many organizations is internal protectionism -- in IT, EA, business analysts-- even the CIO. As some of the work coming out of leading analysts and consulting firms has recently confirmed, strategic alignment of IT is now very much on the minds and priority lists of CEOs and boards--the broader issue of IT ecosystems (internal and external working together often against the interests of the organizational mission). As you may (or may not) have noticed, I've been blogging on this issue recently. Perhaps more importantly, we're beginning to see enormous change in behavior and budgets in some of the largest organizations (and funded projects) surrounding these issues. I am seeing strong signs of convergence of enterprise acronyms, enabled of course by technology innovation.

Great point

Hi Mark, thanks for the comment. I do agree that the resistance to MDM in the Cloud is less so of a technical barrier than it once was and is now more of an IT cultural and sometimes political one within many organizatoins. That doesn't mean that there's a chock full of mature MDM solutions with the same level of configurability and usability in the Cloud today as on premise- there's not. But the crawl-to-market of these solutions is less a technical barrier than a wait and see attitude with the vendors to ensure that a viable market will exist for these solutions since the expense of getting there is steep.
Thanks again!

Well, mediocre on my side anyway...

But I think you make a good one here as well -- the nature of the beast with MDM makes it potentially very important to organizations, but much easier to provide in a holistic architecture--which goes back to my point of who is competing with whom... so agree that since cost of achieving is prohibitive for most stand alone efforts I can see (esp. since so few are even dealing with "big data" apparently--begs the question how anyone can achieve either M without the D?). It would seem much easier to layer on top of a rational architecture (or better yet foundational in establishing a rational architecture). Difficult to craft a business case without advanced analytics/BI -- esp. in the near term, but not impossible I suppose. On the other hand look at some of the investment still with basic integration in keeping the lights on... kind of tough to master anything when forced to ask permission of the conflicted gatekeeper for access. .02-MM

20% DG Maturity Won't Stand for Long


Back in 2008, when we met together at the Data Governance Council Meeting in Bachelor's Gulch, Data Governance Maturity was wishful thinking for almost all in the room and in any room for that matter. I remember someone saying that 95% of all companies weren't even ready for a Data Governance Maturity Model and they'd need some basic starting steps to get going.

Three years later, the rooms full of Data Governance practitioners are far larger and with so many new entrants into the market it isn't at all surprising that maturity surveys show only minimal growth in overall maturity. Since 2010, the Information Governance Community has been measuring Data Governance Maturity via the online Maturity Model Assessment and the average score is barely a 2 out of 5.

And I'm sure 2012 won't look much different.

Fact is, its getting easier and easier to find skilled help to start a good Data Governance program but its darned hard to keep it going and almost impossible to expand it and get better results.

We in the industry need to do far more helping organizations govern well than just govern at all.

Happy New Year and hope to see more of you this year.



I agree, maturity requires continuity

Hi Steve
Happy New Year to you as well, and thanks for sharing your insights from the Data Governance Council. I was actually the one who shared at that meeting that 95% of the companies weren't ready for the maturity model and am not too surprised at the score of 2 for average maturity. We're both clearly following similar trends and inhibitors.

As I mentioned in my post, while I'm not seeing higher maturity I am seeing higher levels of engagement from the business and business leaders. But to your point, that engagement is moot if the DG program can't measure success and prove its value pretty quickly.

Look forward to catching up soon
Best regards,

Solutions are evolving.. additional "cloud" perspectives

Hi Rob, thanks for the post.

I think one key aspect many organizations haven't looked at yet are the new "data sources" that in many cases can provide more up to date information than the traditional sources corporations rely upon, at least in the customer/partner space.

I recently shared my own IT owner/architect's perspectives on MDM evolution and MDM 3.0 concepts, that include how both social networks and cloud/data virtualization solutions can come together.

"MDM 3.0 - Virtual MDM and Distributed Data Governance: Advanced Hybrid Approaches in MDM’s future"

If you end up checking it out, I would be interested in your thoughts.


Virtual MDM

Hi Mehmet - Thanks for sharing your end user perspective and experience here, that's always extremely valuable to all.
While I agree that Virtual MDM is a viable and practical architectural approach for many use cases - and a fast maturing one at that, I view it just as that - an alternative approach, not a next step in MDM's evolution. In my MDM survey, only 8% of the respondents shared that they were using or considering a Federated or Virtualized MDM architecture, while the vast majority were still aiming at the various traditional and emerging persisted models (Hub, Index, Registry, EDW, etc). To your point, that doesn't mean that number won't increase as virtualization capabilities and cloud-based alternatives mature over time, but I really don't expect that to happen in the near future. But it'll be interesting to see how this plays out as more organizations such as yours educate others to the benefits and best practices for this approach.

Thanks again for sharing your comments.

Hi Rob, I am not surprised

Hi Rob,

I am not surprised most companies are still looking at traditional and (realizing the webinar was an hour in length) I do not propose virtual MDM as an alternative to traditional. Instead, I highlight the following points:

1. To effectively plan an MDM program, information and data quality problems need to exposed as early as possible. Data quality tools and data virtualization/EII technologies can help with this during planning cycles by showing data to users sooner rather than later

2. ETL development cycles across multiple data store layers are needed for stadnardization and trust rules, and there are other fields that are needed for reference/context purposes. Data virtualization in the MDM architecture can improve ease of pulling "supporting fields" for information consumers, including but not limited to data stewards, from other repositories, especially when this data does not have to be in MDM (though likely it would end up an in warehouse eventually)

3. Data virtualization tools also improve the value of an MDM solution over its life cycle by providing quick turn around information integration solutions through MDM Hub cross references/matching capability + additional fields from unintegrated sources, whether on the web or in another transactional system/spreadsheet.

Hope this provides further context on the ideas. Some scenarios/screenshots are in the presentation.


Thanks for clarifying

You're right, I didn't have time to watch the webinar so took my assumptions from the summary write-up. I agree with all your points here on how virtualization techniques can be an extremely valuable took in the MDM toolkit to support profiling, stewardship, monitoring and data access requirements.