Month: December 2016

Everyday Digital Transformation

Henrik Gabs LiliendahlLeave a comment

Ben Rund of Informatica has a Youtube video running these days with the title/question: Enough Heard on Digital Transformation by Uber & AirBnB?

I share this sentiment with Ben. You don’t have to disrupt the whole world to take part in digital transformation and you don’t have to start something completely new. As an established enterprise you can transform your current business and combine the good things from the past with the new opportunities aroused from the digital evolution.

Forrester, the other analyst firm, some years ago devided digital transformation into a loop of:

  • Digital Customer Experience
  • Digital Operational Excellence

The below figure visualizes this landscape:

digital

What I would like to elaborate on related to this picture is the business ecosystem of your enterprise, which must be included in the everyday digital transformation.

Let’s take the example of product information management:

However, connect is better than collect. If you are dependent on receiving spreadsheets with product information from your trading partners or you let them put their spreadsheets into your supplier product data portal, you have an everyday digital transformation in front of you.

The solution for that is Product Data Lake.

digital2

Cross Border Master Data Management

Henrik Gabs Liliendahl8 Comments

One of the most intriguing sides of data quality and Master Data Management (MDM) is, in my eyes, how you can extend a national solution to an international solution.

Google EarthMany implementations starts with a national scope and we also see many tools and services built for a national scope. Success on a national scale does unfortunately not always guarantee success on an international scale.

Besides all the important stuff around different culture challenges and how to drive change management in an international environment, there are also some things about the master data itself that are challenging.

  • Location Master Data is probably the most obvious domain where we face issues when going international. Postal addresses are formatted differently around the world. Approximately half of the world puts the house number in front of the street name, approximately half of the world puts the house number after the street name and then in some places you don’t use house numbers on a street, but in blocks. City and postal code has the same issue. The worst solutions here tries to put the rest of the world into the first implemented national solution as told in the post Nationally International.
  • Party Master Data, also when looking beyond postal addresses, must encompass many national constraints and opportunities, not at least when it comes to exploiting third party data:
    • Utilizing business directories is one common way. Here you have to balance the use of many different best of breed national providers or taking it from a more harmonized provider of an international directory. Where I (also) work right now, we have chosen the latter solution as reported in the post Using a Business Entity Identifier from Day One.
    • If you, as I am, are coming from Scandinavia you are also amazed by the difficulties around the world there are in healthcare, elections and other areas when there is no public available national identifier for citizens as examined in the post Counting Citizens.
  • Product Master Data does in many ways look the same across countries. However, standards for product data often still are specific to a single or a specific range of countries. Also, if the national implementation was not in a country with multiple languages and the international scope includes more languages, you must encompass multilingual capacities for product information management.

What have you experienced when going from national to international?

Knowing what quality product data looks like

Henrik Gabs LiliendahlLeave a comment

Recently Daniel O’Connor blogged about Three Keys to a Successful Product Data Project BEFORE You Start the Project. Number one key suggested by Daniel is to know what quality product data looks like. I agree.

Besides Daniel’s very valid points on this matter, I would like to bring data quality dimensions into the game. Looking at data quality from a completeness, timeliness, conformity, consistency and accuracy point of view will help crafting tangible measures and identifying the root causes of where current culture, processes and technology lack the capabilities of meeting the desired state of product data quality.

QualityHere is my take on how to use data quality dimensions for product data:

Completeness of product data is essential for self-service sales approaches. A recent study revealed that 81 % of e-shoppers would leave a webshop with incomplete product information. The root cause of lacking product data is often a not working cross company data supply chain as reported in the post The Cure against Dysfunctional Product Data Sharing.

Timeliness, or currency if you like, of product data is again an issue often related to challenges in cross company supply chains. You can learn more about this subject in the post How to avoid Stale Product Data.

Conformity of product data is first and foremost achieved by adhering to a public standard for product data. However, there are different international, national and industry standards to choose from. These standards also comes in versions that changes over time. Also your variety of product groups may be best served by different standards.

Consistency of product data has to be solved in two scopes. First consistency has to be solved internally within your organisation by consolidating diverse silos of product master data. This is often done using a Product Information Management (PIM) solution. Secondly you have to share your consistent product data with your flock of trading partners as explained in the post What a PIM-2-PIM Solution Looks Like.

Accuracy is usually best at the root, meaning where the product is manufactured. Then accuracy may be challenged when passed along in the cross company supply chain as examined in the post Chinese Whispers and Data Quality. Again, the remedy is about creating transparency in business ecosystems by using a modern data management approach as proposed in the post Data Lakes in Business Ecosystems.

My 2017 PIM Clairvoyance

Henrik Gabs LiliendahlLeave a comment

When writing a blog post about predictions for next year a common way to start is to explain why your last year predictions in a way was true.

Well, my foreseeing for 2016 was called My 2016 MDM Clairvoyance. This included gut feelings about Master Data Management (MDM) including Product Information Management (PIM).

One hunch was about mergers. There is still two weeks left to see that coming true.

Bowl
Magic glass bowl

Another guess was about shortage of MDM people and more agile MDM implementations. Perhaps that was right.

The bet that certainly happened was about only one Gartner MDM magic quadrant. Though it is delayed – as reported in the post The Gartner Magic Quadrant for MDM 2016 – it will according to my sources land in 2016.

The burning issue for me in 2017 is how many companies that will abandon spreadsheets as the mean to exchange product information with trading partners and resist the temptation to set up a selfish supplier or customer product data portal? The potential numbers was examined in the post Alternatives to Product Data Lake.

Or put in another way: How many subscribers will we have at Product Data Lake by the end of 2017? My guess is 1111. In a year I will reveal if this number is expressed as a binary number, a decimal number or a hex number.

The Gartner Magic Quadrant for MDM 2016

Henrik Gabs Liliendahl20 Comments

The Gartner Magic Quadrant for Master Data Management Solutions 2016 is …… not out.

Though it can be hard for a person not coming from the United States to read those silly American dates, according to this screenshot from today, it should have been out the 19th November 2016.

gartner-mdm-2016

I guess no blue hyperlink means it has not be aired yet and I do not recall having seen any vendor bragging on social media yet either.

The plan that Gartner will retire the old two quadrants for Customer MDM and Product MDM was revealed by Andrew White of Gartner earlier this year in the post Update on our Magic Quadrant’s for Master Data Management 2016.

Well, MDM implementations are often delayed, so why not the Multidomain MDM quadrant too.

In the meantime, we can take a quiz. Please comment with your guess on who will be the leaders, visionaries, challengers and niche players. Closest guess will receive a Product Data Lake t-shirt in your company’s license level size (See here for options).

A Data Lake, Santa Style

Henrik Gabs LiliendahlLeave a comment

Following up on last years post on Big Data Quality, Santa Style (and previous years of Santa style posts) it is time to see how Santa may utilize a data lake.

birthday presentsI imagine that handling product information must be a big pain point at the Santa Corporation. All the product information from suppliers of present items comes in using different standards and various languages. In the same way the wish lists from boys and girls comes in many languages and using many different wordings.

Forcing the same standard on all suppliers (and boys and girls) is quite utopic – even for Santa.

So using a data lake for product information seems to be a good choice, not at least if that data lake encompasses the whole business ecosystem around the Santa Corporation.

By joining Product Data Lake the Santa Corporation will put their required product portfolio and the needed attributes for the products into Product Data Lake in all the languages operated at Santa’s site.

The suppliers of toys, electronics, books, clothes and heaps of other nice things will by joining Product Data Lake in the same way put their products and the attributes offered into Product Data Lake.

In here, the products and attributes will be linked to the ones used by Santa. But this is only the beginning of a joyful ride. The products and attributes can also be linked to all the other trading partners on Product Data Lake, so the manufacturer will only have to upload this information once.

Ho ho ho. This year it is not only nice boys and girls that gets a present – and this year the right one -from Santa. Smart suppliers will get a big present too.

Social PIM, Take 2

Henrik Gabs LiliendahlLeave a comment

My first blog post on Social PIM (Social Product Information Management) was over 4 years ago.

take-2Since then Product Data Lake has been launched. Product Data Lake resembles a social network as you connect with your trading partners from the real world in order to collaborate on getting complete and accurate product information from the manufacturer to the point-of-sales.

I would love to see you, my blog readers, become involved. The options are: