Gartner, the analyst firm, has a different view of the data quality tool market than of the Master Data Management (MDM) market. The MDM market has two qudrants (customer MDM and product MDM) as reported in the post The Second part of the Multi-Domain MDM Magic Quadrant is out. There is only one quadrant for data quality tools.
Well, actually it is difficult to see a quadrant for product data quality tools. Most data quality tools revolves around the customer (or rather party) domain, with data matching and postal address verification as main features.
For the party domain it makes sense to have these capabilities deployed outside the MDM solution in some cases as examined in the post The place for Data Matching in and around MDM. And of course data quality tools are used in heaps of organizations who hasn’t a MDM solution.
For the product domain it is hard to see a separate data quality tool if you have a Product Information Management (PIM) / Product MDM solution. Well, maybe if you are an Informatica fan. Here you may end up with a same branded PIM (Heiler), Product MDM (Siperian) and data quality tool (SSA Name3) environment as a consequence of the matters discussed in the post PIM, Product MDM and Multi-Domain MDM.
What should a data quality tool do in the product domain then? Address verification would be exotic (and ultimately belongs to the location domain). Data matching is a use case, but not usually something that eliminates main pain points with product data.
Some issues that have been touched on this blog are:
- Measuring and reacting on completeness as explained in the post Hierarchical Completeness withing Product Information Management
- Ensuring consistency as a result of increasing use of the approach of Sharing Product Master Data
- Capabilities of image data quality as mentioned in the post Digital Assets and Product MDM
Anyway the first vendor tweets about the data quality tools quadrant 2014 is turning up, and I guess some of the vendors will share the report for free soon.
Update 3rd December: I received 3 emails from Trillium Software today with a link to the report here.
Liliendahl on Data Quality 
Hi Henrik
A decent enterprise class data quality tool should be abble to handle any kind of data.
product and name and address data both havwe an ontology and can be standardised and matched – although name and address tends to be more standardised
But data quality must also handle transaction data, financial master data, account masters, and any more.
data quality is a much broader problem than MDM and must cope with much more variety
Thanks for commenting Gary. I agree. You may also argue that some DQ tools are good at some specific challenges. An example, from what I have seen, is that that some of the weaker positioned DQ tools are better at data matching than the leading enterprise DQ tools. As said, data matching is much more important in with party master data than with product master data – and with financial master data and transaction data.
Hello Henrik,
All good points, including the interaction with Gary ; it shows that our market is not so easy to segment both from an independent analyst, vendor or customer perspective.
An interesting outcome from this quadrant is that their customers survey show the highly growing need for multi domain and use case for data quality: although Party domain is highlighted as the most frequent priority, by far, three others domains are considered as a priority by more than 50% of their surveyed panel: financial/quantitative data, transaction data and product data.
I’m not sure I would fully agree with you on the fact that Product Domain wouldn’t take “enough” advantage of a strong best of breed data quality capabilities (no matter if they are provided out of the box in the MDM/PIM solution or as an add on either from the same or from another vendor). In fact, I found at Talend many example of customers that use our Data Integration/Data Quality platform together with PIM solutions. In that case, PIM is seen as an enterprise application, whereas the DI/DQ (and eventually the product MDM) is the cross application hub. Also, my feeling is that product data will have more and more to be sourced externally with Big Data and the Internet of Things, with bigger needs in terms of Data Quality and Integration.
Thanks Jean-Michel for adding in. I do agree about that product data will have more and more to be sourced externally.
@Henrik –
As was pointed out in the thread, Data Quality tools are used for both master and transaction data (and reference data, and metadata, and …). In fact, with Informatica MDM and Informatica PIM, we now embed our data quality capabilities within those solutions. This gives us the ability to standardize and cleanse data coming into the hub, translate it to the appropriate formats and “outbound” master data (heading to target systems), and also profile and identify data quality issues so that appropriate governance actions can be taken to proactively eliminate those problems in their sources.
PIM is *not* product data mastering, although it can include that capability. PIM is product data *authoring*, and modern PIM systems go beyond that capability to include related processes and data including:
* product data mastering,
* supplier on-boarding and collaboration (e.g., supplier product data on-boarding),
* location mastering,
* assigning relationships between products, locations, and suppliers,
* and much more.
One very interesting use case for data quality tools, when embedded in a PIM, is “auto-categorization.” With auto-categorization, data quality rules can be used to assign products to hierachies (or levels within one or more hierarchies), as well as to extract attributes from textual descriptions, translate manufacturer-specific reference data into the PIM customers’-specific reference data, and even to automatically derive cross- and up-sell (and other) relationships.
Great blog topic!
Thanks,
Dennis Moore
Informatica