» MDM, Fuzzy Matching and Deduplication Blog

Excited to be back with another interview for the Master Data Maestro series. Today we have Howard Diesel, current DAMA SA President and a certified data management professional. Howard really enjoys working with people excited about data and information and working with community organisations like DAMA and ISO to help address information management issues. He specialising in information governance advisory services and helps organisations establish effective information management programs that deliver business value by enabling capabilities as part of data project delivery.

We hope you enjoy his take on data management as much as we did.

Thank you for allowing us to interview you. Really appreciate your time. Can you please tell our readers a bit about yourself? Your background, education, work experience.

In 1986, I completed my BSc Comp Sci and started working in the Data Management industry as an Adabas / Natural Database Administrator for a major retailer. I have been working with Data ever since and have had the opportunity to spend good quality time in every area of the DAMA Wheel. Most of the time has been in the Data Warehousing, Business Intelligence & Data Science knowledge area. There is a desperate need for good quality Master Data that supports conformed dimensions.

Editor’s note: True that! There is surely a desperate need for good quality master data.

I still do Database Administration & Business Intelligence for several customers! I also continued with my Education and went on to complete a BSc Hons in Computer Science, a Graduate Diploma in Engineering and then a M.Sc. (Electrical engineering).

Editor’s note: We learnt that you are also currently preparing to begin a PhD in Data Handling Ethics. Good luck with that!

My primary focus as an advisor/consultant is to provide organisations with thought leadership in Data Governance and building Data and Information Strategies as well as establishing the Data Management practices.

Please tell us more about DAMA as well as your work as President of DAMA South Africa.

For me, Learning and Certification has never ended. Our focus is primarily on Data Management Education in Southern Africa. We are constantly engaging with fellow Africans in trying to setup more DAMA chapters within the African Content. It is bad now where DAMA SA is the only affiliated chapter in Africa. We are hoping to change that this year!

In the last year we have achieved a massive growth in the number Certified Data Management Professionals in Southern Africa. The chapter started in 2009 and I got my CDMP in 2015, at that stage I was one of approximately 25 professionals. In the last 2 years, our team, predominantly Veronica, has managed to increase that number to close on 100. We will certainly celebrate when we break the 100 mark.

Must congratulate DAMA International in establishing the online platforms and the focus on Data Management Certification

Whats the business problem you are solving currently and what impact does it have on the organisation?

My advisory focus is on developing Data and Information Strategies for Organisations. It is still early days, but it has a tremendous effect of getting organisations to focus on addressing the critical components of data management – Change Management, Human Capital, Organisational Memory, and Practice Establishment. The lockdown has certainly highlighted the lack of Organisational Memory and effective practice definition, operationalization, and oversight.

My DAMA focus is predominantly on training, certification, and learning expeditions. We have great success with Monthly “Helpful” Zoom Meetup sessions for Data Modelling and Data Governance. It has been a great privilege travelling to surrounding African countries, like Botswana, meeting some real data management professionals. It is really encouraging to see the development of data management professionals within the African Continent.

We are really focused on addressing the Human Capital Constraints that companies are facing and their ability to apply mature data management engineering.

Editor’s note: Right. Technology has to adapt itself so that data practitioners and stewards are aided through AI. We also need self service and agile data mastering techniques which allow the user to easily master their data. If you have an implementation running for years, by the time the data comes out of it, the business environment has already changed and there is need for a different view.

You focus on financial service providers in South Africa, experience in Banking, Central Banking & insurance. Please tell us what kind of challenges you see in the BFSI sector around data.

“Regulatory burden” is a challenge facing all companies, industries and countries and for me it is one of the biggest and most exciting issues facing the Prudential Industry. There are many components to this challenge but I believe the biggest data challenge is harmonization of data starting from “The Law”, Industry Regulation Partners, Internal Regulations & Policies, Economic Concepts, Supervision, Surveys, Transactions, Events, Submissions and Market Data.

We need to convert our regulations & policy data, currently sitting as documents (unstructured data) to Master Data into Semantic Knowledge Graphs so that we can associate and connect Economic Concept data to our policies and vice-versa.

It is so hard to establish consistent semantics (local & internationally) of regulations and policies – basically we do not have a core authoritative taxonomy & ontology within the industry (local and international). There are some great examples in Portugal, UK and Austria.

Part of the challenge is that the regulatory institutions collect their own data, instead of sharing data amongst the other regulatory partners to reduce the reporting entities burden and the administrative burden within the regulator.

According to you, what are three biggest reasons for the BFSI industry to focus on the data they have?

According to me the three biggest reasons for Data Strategies, relating to the regulatory information, are:

Connecting data and information along the entire business value chain, from the Acts to Financial Market data. Agree and aligning to the regulatory information value chain.
Ensure the harmonization of this information across all the industry stakeholders. Establish and Speak a common data language.
Dramatically improve the “Question to Value rate” thereby allowing Policy Decision makers to ask the data questions, receive insights, take actions, and produce sustainable value in the time required to ensure economic stability

How would you define Data Quality?

As, you may have noticed in my previous responses, quality for me, is about establishing common semantics of all the data and information involved in financial regulation (IFRS 19 and BCBS 239 are examples).

As an industry, both local and international, we need to establish:

Confidence in Data Management Human Capital
Well managed and trusted Organizational Data Management Memory that is accessible and the preferred source of the truth, not google
Sustainable, Harmonized and Accessible Regulatory data

It’s great to see the development of making open data accessible to the world and I believe we can apply the same principles across all industries, not just the regulatory industry. I am currently working on ISO working draft standards that focus on this area.

Your work revolves around highly regulated financial companies. With increasing concern for privacy, what is your advice to your customers?

I have been very impressed with the focus that most financial organisations are paying to data privacy through Cyber Security projects. Every big organisation, that I am aware of, is investing a lot of time and money in this field. Several Central Banks have identified data privacy and security risks.

The only element that I would like to add to this equation is ethical data handling and introducing Ethical Impact Assessments with the Privacy Impact Assessment.

You are based in Johannesburg. How do you see the local demand for MDM in your region compared to internationally.?

I believe the demand is the same! Our challenge comes with our awareness and master data management capabilities to establish trusted master data that can be used to connect all types of data and information.

What are some forums, websites and conferences you would recommend to update oneself about the MDM field?

Currently, I am engrossed in updating my competence in the areas of semantics, data architectures, like data-centricity and knowledge graphs. I believe that this knowledge, lays the platform for new thinking and approaches to managing our data.

Please name few data practitioners whom you admire and would love to hear from in this interview series.

I am glad, you have already, interviewed Gary Alleman, he has a very sober and practical view of data management. Always enjoy having long discussions about data with him.

Paul Grobler is also doing some great work in the field of Master Data.

On an international stage, Steve Hoberman, has been truly inspirational in the field of data modelling which I apply daily to work on taxonomies and master data. Steve has reinforced, the requirement for me to “feel the data”. We need to do this a lot more before we start to manage the data.

I have been excited to learn from several Europeans, like Christian Kaul, Lars Ronnback, Martijn Evers, and other people in the field of Data Vault & Anchor Modelling.

Editor’s note: Quite a few new names here, thanks for this. We will be reaching out to each of these maestros.

We hope you enjoyed reading the views of Howard. Howard is available for consultations and training on any area of information management. You can reach Howard at howard@modelwaresystems.com or president@dama.org.za. We wish him the best with his journey on data management.

Comments, questions, suggestions, feedback – contact us here. Thanks!

Posted on May 2nd, 2020

Master Data Maestros march on! Hope you have been enjoying the series so far. Today we talk with Nadim WARDÉ. Nadim is the founder of the European Leader Product Data Management Platform in the CPG. He is Lecturer at Berner FachhochschuleTechnik und Informatik, Supply Chain & Logistic Health Sector. Nadim’s course specialises in Product Data Governance and Data Quality. Nadim is also Lecturer at HEG – Fribourg, teaching GS1 Standards in E-commerce.

Thank you for agreeing for this interview Nadim. Really appreciate your time. Can you please tell our readers a bit about yourself?

Thanks very much for the invitation. My data journey started several years ago and I have had different experiences with data and specifically on product domain, covering different aspects of the product data management. I started in a Swiss retail group, having the responsibility of building a product MDM for analytics purpose. At this time, there was no multichannel, web… but I was already facing issues like data acquisition, data quality, data governance. After that, I built the first pan-european product data exchange solution based on GS1 standards. Later, I joined Heiler/Informatica to envagelize about product data, MDM & PIM in southern Europe. This is where I detected the needs for consultancy and support companies in their projects, and this is where I founded arounData consulting, 5 years ago.

What is your current role? Whats the business problem you are solving and what impact does it have on the organisation?

From my experience, what’s around data such as organization, processes, solution are much more important than data itself. The data (value of an attribute) is an output of an organization and processes. If these pieces are not correctly set and connected, data will always be a problem.

So I help companies from project inception, building the business case for their product data journey in order to make the most of their investment, till support in implementation.

Of course I go through the whole scoping, working on processes, data dictionary and glossary. The challenges are often:

time to market (data is consuming too much time to be collected and distributed)
direct and indirect cost of data quality
poor and inconsistent content
data published and available to customer

This type of problems generates a bad employee experience with data (we don’t speak about employee experience – but if you want good customer experience, you need, before all, good employee experience); people within organization spend lot of time on chasing and collecting the right data from the right source, creating stress and unsatisfaction.

And in so many cases, employee end up in storing data in local Excel files, in order to secure their own data source… generating endless data quality problems.

Your LinkedIn bio mentions that you are engaged in “what’s the role of the data, friction and pain points in acquiring data and publishing information”. Please tell us more about this.

For instance, on a manufacturer side, each service stores the data in its own way : software, excel….and it’s a complete headache to collect it from other service, to be sure that they have the latest version, qualify it and publish it to the retailer through GDSN, Retailer Excel file or supplier portal.

Often, the retailer is collecting the same data in different ways.

So there is absolutely no continuity in the Information product supply Chain from the source point till the distribution. We can compare this information supply chain to the physical product supply chain.

Data is transported, stored, transferred from a database to another one, manipulated by different systems. While lot of money is invested to ensure a correct physical supply chain, lot of money is lost due to the bad data circulating; lot of money is spent to compensate disruption of the physical supply chain due to this bad data.

The lack of data integration, between the different stakeholders in the information chain, generates lots of disruption and data transformation (and often, several transformation).

And you can finally end up with a complete different data than the source one. When you add, to this, lack of data governance and data quality, you end up with additional and useless costs.

I’ll give you the example of a retailer sending its trucks to pick ordered goods from the supplier warehouse. There was no shared understanding of the height of a pallet; so the retailer couldn’t stack the pallets in the trucks (due to the fact that height of the wood support that was not included in information stored in the retailer systems). He was obliged to send additional trucks to remove the goods that were not delivered on time to the shops.

Editor’s note: Its a very interesting case you mentioned. So many companies are struggling around collecting data in different ways and lack of data integration, losing money on supply chain due to bad data.

According to you, how has the MDM landscape changed since when you started to the present?

MDM is definitely recognized as a strategic discipline, being foundational in the customer experience and digital journey. Data is recognized, also, as a strategic asset (or oil).

This is supported by 2 major evolution, I have seen in the last 5 years in MDM

Improved and very user friendly UI
Low-code technology

These 2 combined elements are making MDM much more “sexy” for business and facilitating their ownership.

MDM is no more an IT “thing”. Business people recognize the value of having and mastering MDM solution.

I see more and more business application as PIM, CRM, Supplier Management Portals… encompassing MDM features and functionalities. And making a multi-domain MDM and management of reference data, mandatory.

How would you define Data Governance and Data Quality?

Data Quality and Data Governance are very tightly linked. Data Quality is the way to guarantee that your data is healthy and ready to be used/consumed. Data Governance is all processes, organisation and tools that are set in place to ensure and maintain the level of required data quality.

Data Quality without Data Governance is not sustainable and Data Governance without Data Quality is useless.

And the easiest way to build a Data Governance culture and organisation is to start working on Data Quality problems: identify the most critical and pain one, evaluate the cost of the non-quality, impact on the organisation and the benefits of correcting it, setting a plan and a price of implementation.

Follow up with KPI’s and reiterate (select a second case and go through).

Unfortunately, lot of organisation think they have reached the paradise when they have solved some DQ problems.

I still see, after more than 20 years, lot of retailers complaining about same DQ issues when exchanging and sharing data… due to lack of data governance and data maturity.

Maybe it requires to go through a real painful problem to understand what’s data quality and data governance:

One of my customers, a swiss watch company, has had a full load of very costly watches blocked by the customs due to missing data in the documents.

They understood the costs and accepted to enter the Data Governance journey.

Editor’s note: Makes a lot of sense – identify the most painful part and use the simplest way to solve it. We have been advocating the same to our customers with Reifier.

What are some common data quality issues you see in your day to day engagement with customers?

The main and first challenge in Data Quality is change management. Trying to solve DQ issues means, that, you need to change data ownership, stewardship… improve processes, impact the way, people works.

Other challenge is to put some cost in front of DQ problems. Especially, when you start on the first sessions: Collecting the cost of DQ is a headache and a problem itself : identify the incurred costs, search for invoices, calculate them…

I’ll not detail the different product data quality problems I met. But in the product domain, organisation are very sensitive when you tell them that the quality of their data may reflect the quality of their products.

Going back to your previous question, I found, that for lot of people, Data Governance is limited to setting some data quality rules, policies and ownership attributes.

So usually, I wait for 2 elements to start talking/educating about Data Governance

the go-live of the project
Starting data quality workshops and assessment, highlighting business impacts (and after the go-live)

From there, people start to see concretely, in their new working environment that is supposed to improve employee experience, the impact of bad quality, why data cleansing is mandatory, lack of processes; and all of these elements.

And here is, where top management is ready to listen to Data Governance music.

What advice would you like to give to a customer starting their MDM journey?

BUILD the BUSINESS CASE, to support the budgets needed to this journey; to better explain why you’re doing it and show the ROI of the investment (time and material).

As you said, it’s a journey and not a project. So, spend time to explain what’s MDM to CEO and board of directors, what are the benefits and impacts on the bottom line.

If you put some amounts and money in front of this, you can initiate a behavioral change.

Review your processes around data (if you have) or build new ones if needed

Communicate in a simple way to explain the journey and you’ll see how many frustrated people you have in a company, that were waiting on such initiative.

Don’t let IT manage the project. Lot of people associate MDM with IT. So Keep business in charge of the project and different milestones.

This should be a good way to build a data culture within a company.

But don’t talk about Data Governance, Data Onwership…. Unless the organization has some maturity to hear and listen to it.

And last (but not least), work with the right and good consultants…

Editor’s note: And of course the right tools! Shameless plug 🙂

What are some forums, websites and conferences you would recommend to update oneself about the MDM field?

I appreciate the communication done by Georges Firican and lightsondata, explaining basic concepts as data dictionnay, data quality, data glossary…BobSeiner with TDAN – very instructive. Henrik with the regular newsletter giving lot of useful insights about PIM/MDM area. Ben Rund with its interesting data talks. And of course LinkedIn, which is currently a good place to hear and learn about MDM.

Please name few data practitioners whom you admire and would love to hear from in this interview series.

Henrik named them in the last interview. All of them are doing great and lot of communication on MDM on LinkedIn: Lara Gureje, Scott Taylor, Daniel O’Connor, Kersten Wirth, Chris Jobse, Michael Fieg, Kate Koltunova, Sophie Angenot, Steve Jones, Salah Kamel, Ian Buttery, Kimmo Kontra, Julie Hunt, William McKnight, Julian Schwarzenbach, Ken O’Connor, Malcolm Chisholm, Mike Ferguson, Nicola Askham, Prash Chandramohan, Dylan Jones, Thomas Frisendal.

So nice to get in touch and talk with you Nadim, many more master data mastership for you!

Comments, questions, suggestions, feedback – contact us here. Thanks!

Posted on April 27th, 2020

Today we have the effervescent Scott Taylor as our Master Data Maestro. Hope you enjoy this interview which has a lot of wisdom on why master data is so essential to business. There are a ton of resources too for reference and further reading.

Thank you for allowing us to interview you. Really appreciate your time. Can you please tell our readers a bit about yourself? Your background, education, work experience.

Happy to be here. I’m known as The Data Whisperer and act as Principal Consultant for MetaMeta Consulting – A Data Story Boutique. We help people tell their data story by enlightening business executives to the strategic value of master data and proper data management. I focus on business alignment and the “strategic WHY” rather than system implementation and the “technical HOW.”

master data management

I have worked with global enterprises by representing some of the world’s most iconic business data brands including Dun & Bradstreet, Nielsen, Microsoft and WPP/Kantar in a variety of strategic marketing, GTM, innovation and consulting roles.

I live in Bridgeport, CT where I often kayak in Black Rock harbor. I can also juggle pins and blow a square bubble!

Editor’s note: Square bubble is interesting, we should have online classes for this 🙂

One of your monikers is “The Data Whisperer”. Would love to hear more about how it started and what’s the story behind it.

As The Data Whisperer, I help calm data down. That’s what we all do in the data management space is calm data down. Data is unruly, data is corrupted, data is big, data is unstructured, it needs to be calmed down. As people learn to improve data quality and grow their data management skills, they will become Data Whisperers too.

What according to you is master data and how does it help?

Master data is the most important data any organization has. Period.

According to Wikipedia: Master Data is a common source of basic business data used across multiple systems, applications. I hope you like that definition, because I put it there.

Master data is probably the oldest kind of data being this foundational structured list of the things that are most important for an organization. It’s often referred to as “the nouns” of the business: customers, vendors, partners, prospects, brands, services, offerings. The things you make, the parties you buy from, and sell to, and partner with.

Master data is highly structured and coded, tightly govern and widely shared. It really is the foundation that makes all other data relevant.

If you think about a sales number, a transaction, is it the intersection between a customer and a brand. You’ve got to have good well mastered-data to have trustworthy sales numbers.

Today everybody’s about digital, they’re about cloud, they’re about hybrid, they’re about analytics. They are about fabric and mesh and all these other cool super things. Those all need structured content. So master data is macro-trend agnostic. No matter where we’ve been, no matter where we’re going, no matter where we are now, whatever the future is, we’re always going to need structured content about the basic entities that business is all about.

That’s what Master Data is.

According to you, is Master Data Management more relevant to the Fortune 500/1000 or should other organisations also pay some attention to it?

How do you know you need MDM? If you’re in Commerce on Earth and you’re an Enterprise, you need Master Data. I haven’t run into a company that doesn’t need this somewhere. I look at the Strategic intentions of an Enterprise to see how master data can fit in. Part of what I do is help people tell their data story about why they need this foundational data within their Enterprise. All you have to do is look at an annual report or an investor day presentation, where the senior leadership talks about wanting to “improve customer experience” or “be the preferred partner in our relationships” or “move to as-a-service.” But most compaines don’t have the data to back that up. It’s not going to work. So yes, if you are an entpereise, you have enterprise data. And if you have data, you have data problems. So, every enterprise needs master data as far as I’m concerned.

You are soon Keynoting on Data Management as the Foundation of Digital Transformation. Would you like to share some thoughts on this topic with our readers here?

I’m very excited that DATAVERSITY, the event organizer, was able to reschedule Enterprise Data World for October 18-23 in Chicago. Here’s the link: https://edw2020chicago.dataversity.net/

Yes, I’ll be speaking about how Data Management and master data are the Foundations of Digital Transformation.

Digital transformation can mean a lot of things to a lot of people. But one thing it means to everybody, every organization, is that you’ve got more data now than you ever had before, you need new ways to manage it that you never thought of before, so you can do things you never imagined before.

I look at three phases of digital transformation. Most people start in a legacy state. In the state of multiple silos. Sales, marketing, finance, all these systems that don’t connect. You want to be focused, as an organization, around the most important things that drive your business. That means cutting across all those different silos. You want to move from a legacy state to an integrated enterprise and become relationship centric and create customer 360 or product 360.

Now when you can do that internally, that’s great. You’ve made that leap from legacy state to an integrated enterprise. But now you’ve got to work outside your organization. That’s when you become part of connected ecosystems and engage in a trust network with your partners, with your service providers, with your distributors,

You’ve got to move along from a legacy state of multiple silos to an integrated enterprise being relationship centric, then to this trust network of connected ecosystems. The universal requirements to succeed in this journey are 1 – authenticated identity and 2 -a common data structure. Master data provides both of those.

Master data, I would say to you, IS the foundation of digital transformation. It provides semantic consistency; it enables deeper inter-operation and drives greater value. The only way to do it at scale is to have good, solid, well- governed master data, reference data, metadata, whatever you want to call it.

Editor’s note: Agree, our experience also shows that many of our customers have MDM needs arising out of their digital transformation journeys

Analytics are a focus area for most enterprises, big and small. Can you please elaborate some examples you have seen where an MDM implementation fuelled the analytics journey?

It’s all about TRUTH and MEANING. Data management is determining the TRUTH. Business intelligence is deriving MEANING. Analytics at scale, doesn’t work without master data. Nor does ML, AI or Data science. All forms of business intelligence need content that is the result of some form of data management. As a terminology note, I consider data management to encompass the supporting organisational and functional efforts (data governance, RDM, data stewardship, data catalog and the overall category of data management) as well as the outputs (master data, reference data, metadata). Business intelligence refers to analytics, artificial intelligence, machine learning and data science.

Master data should feed every type of enterprise system. No matter what the persona at the C-level, no matter what the department, and the prevailing system that drives that department, all of it needs master data content, “Rubbish in, rubbish out”, “crap in, crap out”, “garbage in, garbage out.” It’s the same thing. I wish we could come up with a new way to say it, but it is The Golden Rule of data. What you put in is what you get out. Garbage into CRM? You get a bad customer experience; you get missed opportunities. Garbage into ERP? You got a lousy ERP. Garbage into ML? You get bad robots. Garbage into BI? BS!! Garbage into AI? AS! Artificial Stupidity! No matter how you count it, no matter how you slice it, “Garbage in, garbage out. “

We can elevate this old data cliché by suggesting GIGO also stands for Goodness in, goodness out.

What are some forums, websites and conferences you would recommend to update oneself about the MDM field?

I’m an avid business evangelist and original thinker, and continually share my passion for the strategic value of master data through industry events, public speaking opportunities, blogs, videos, whitepapers, podcasts, cartoons, puppets shows and all forms of thought leadership. So, you can start with my stuff!

YouTube Channel https://www.youtube.com/channel/UCVQ1YhjNqc77GVsb3Xs4tvw

LinkedIn Profile https://www.linkedin.com/in/scottmztaylor/

Thinkers360 Profile https://www.thinkers360.com/tl/ScottTaylorTheDataWhisperer

Also, a number of events I am involved with:

Data Storytelling Virtual Summit April 28-29 online https://rebrand.ly/5d7f3

Information Builders Summit 2020 https://www.informationbuilders.com/summit

Data Science Life Mastery July 8 online

Maximo World August 4-6 https://reliabilityweb.com/maximoworld-2020?publisher_status=open

Enterprise Data World https://edw2020chicago.dataversity.net/

ThinkLinkers MDMThinkLab Berlin and New York https://www.thinklinkers.com/events

Also check out work from

George Firican https://www.lightsondata.com/