Just in Time
Posted by Sastry at 3/12/2008 4:25 PM 
Categories: Business Web 2.0,Retail,Semantic Web,Master Data,Social Networks
Categories: Business Web 2.0,Retail,Semantic Web,Master Data,Social Networks
Back in 1999, which seems a lot longer to me than 9 years ago, I (Fred) was introduced to a “concept” that was being developed by the Uniform Code Council (now GS1-US) – the organization responsible for the administration of the numbers behind the bar codes (UPC) on the shelves of retailers – called data synchronization.
What if the semantic web concepts been well defined around then?
Back in the day – so to speak – we believed that there were about 10 to 15 attributes about an item that were required to identify an item uniquely. These attributes included dimensions, descriptions and the like. The original working group for the data synchronization effort defined about 60 attributes that would be synchronized between supplier and customer of the product in the supply chain.
Once the larger community, got involved there were hundreds of attributes needed or desired. Many of these attributes – like price – were specific to a trading relationship at a point in time. Many of them – retailer buyer name for instance were specific to the retailer and known only by the salesman calling on the account. In all cases, it was pushed back to the supplier to go find or create all of the data.
What has ensued in the fast moving consumer goods industry has been a very long effort to reach consensus on what items are required for the data synchronization effort to take place. To date, only a fraction of the items in the supply chain have been synchronized using the standardized data synchronization process called GDSN.
This process uses “data pools” and a global registry to send data between supplier and retailer. Despite this process, and all of the time and work spent in developing it, additional attributes are needed by retail systems that are specific to the trading relationship. These are often entered through retailer specific portals or other means. So the all encompassing solution of standardized data synchronization still does not provide a complete source of data for items it he supply chain – until the retail (or channel) specific items are included.
Simplistically GDSN looks like this:
We believe that if the semantic web or the “semantic plane” that we discussed last week (link) had been in place, that the whole process of data provisioning – whether through GDSN or specific point to point communications could have been (and still can be) much easier. While this technology is still being developed, we believe that the time to look at how synchronization of master data can be improved is now.
This semantic plane for the supplier needs to understand the knowledge about the specific requirements for a specific customer and where to find the different pieces of data (attributes) for each request. We like to think of this as Just In Time synchronization or JITS.
JUST IN TIME SYNCHRONIZATION (JITS) CONCEPTS
By dramatically reducing the cost of synchronization and making it easier to understand, the Small to Medium Enterprises (SME) can also participate in the JITS Process.
This JITS concept does not replace GDSN – nor does it require GDSN. It is important to remember that most retailers are not using GDSN and still need master data and need to have it available for “synchronization”. Many retailers only need to synchronize what they buy once because they only buy it once and the attributes are never updated. An example of this is Fashion Retailing, which in many cases is an “in and out” business, and has resisted any efforts in this area because of this fact. A skirt that my wife might buy this year – unless it is a classic timeless line – will be a different color and different style than what she bought last year. Why? Because the skirt she bought last year – or maybe last month – is not made any longer. Computers are the same – as are Digital Cameras – they are updated many times a year, with new capabilities and new specifications. Evergreen data synchronization is simply not needed for these items – simple data harmonization suffices. In a completely dynamic world such as ours, seeking true and complete data synchronization is an exercise in futility. What one should strive at is harmonization of relevant data using Semantic Web technologies.
An industry like fashion is actually a great example of how the semantic plane of knowledge can be used. The fashion buying decision starts with product specifications – those specifications grow in complexity until a product is ordered. Having the knowledge about the supply of items and the costs that could go into it and sharing that efficiently with a global manufacturing base would be much more interesting for the master data process for that industry. Other social networking tools can be used as well to drive to consensus on what can be built and at what price.
We think it is time for individual companies to look at how the processes of sharing master data with their suppliers and customers at all points in the product lifecycle can benefit from the concepts of JITS, Semantic constructs and social networking tools. At Dhara, we are open to using the social networking tools of “Web 2.0” to carry this conversation forward. If you are interested in this as well, please email us at semanticplane@dharacg.info.
Fred Geiger and Sastry Dhara
www.dharacg.com
What if the semantic web concepts been well defined around then?
Back in the day – so to speak – we believed that there were about 10 to 15 attributes about an item that were required to identify an item uniquely. These attributes included dimensions, descriptions and the like. The original working group for the data synchronization effort defined about 60 attributes that would be synchronized between supplier and customer of the product in the supply chain.
Once the larger community, got involved there were hundreds of attributes needed or desired. Many of these attributes – like price – were specific to a trading relationship at a point in time. Many of them – retailer buyer name for instance were specific to the retailer and known only by the salesman calling on the account. In all cases, it was pushed back to the supplier to go find or create all of the data.
What has ensued in the fast moving consumer goods industry has been a very long effort to reach consensus on what items are required for the data synchronization effort to take place. To date, only a fraction of the items in the supply chain have been synchronized using the standardized data synchronization process called GDSN.
This process uses “data pools” and a global registry to send data between supplier and retailer. Despite this process, and all of the time and work spent in developing it, additional attributes are needed by retail systems that are specific to the trading relationship. These are often entered through retailer specific portals or other means. So the all encompassing solution of standardized data synchronization still does not provide a complete source of data for items it he supply chain – until the retail (or channel) specific items are included.
Simplistically GDSN looks like this:
We believe that if the semantic web or the “semantic plane” that we discussed last week (link) had been in place, that the whole process of data provisioning – whether through GDSN or specific point to point communications could have been (and still can be) much easier. While this technology is still being developed, we believe that the time to look at how synchronization of master data can be improved is now.
This semantic plane for the supplier needs to understand the knowledge about the specific requirements for a specific customer and where to find the different pieces of data (attributes) for each request. We like to think of this as Just In Time synchronization or JITS.
JUST IN TIME SYNCHRONIZATION (JITS) CONCEPTS
Just In Time Synchronization (JITS) is a similar concept to GDSN, with a significant difference:
SYNCHRONIZE JUST IN TIME, AND ONLY THE SUB-SET OR SUPER-SET OF DATA THAT NEEDS TO BE SYNCHRONIZED.
By reducing the frequency and amount of data that is synchronized at a given time, the demands on IT infrastructure and the concomitant costs are dramatically reduced. This will drive better industry adoption.
In order to translate this vision into a reality, the existing GDSN Process needs to be reviewed – to reduce the complexity of item definition and to render sub-set updates possible.
Each supplier can implement their own Data Pool at minimal cost and Retailers can connect to Data Pools, by following open technology standards (OASIS, IETF, GS-1).
By dramatically reducing the cost of synchronization and making it easier to understand, the Small to Medium Enterprises (SME) can also participate in the JITS Process.
This JITS concept does not replace GDSN – nor does it require GDSN. It is important to remember that most retailers are not using GDSN and still need master data and need to have it available for “synchronization”. Many retailers only need to synchronize what they buy once because they only buy it once and the attributes are never updated. An example of this is Fashion Retailing, which in many cases is an “in and out” business, and has resisted any efforts in this area because of this fact. A skirt that my wife might buy this year – unless it is a classic timeless line – will be a different color and different style than what she bought last year. Why? Because the skirt she bought last year – or maybe last month – is not made any longer. Computers are the same – as are Digital Cameras – they are updated many times a year, with new capabilities and new specifications. Evergreen data synchronization is simply not needed for these items – simple data harmonization suffices. In a completely dynamic world such as ours, seeking true and complete data synchronization is an exercise in futility. What one should strive at is harmonization of relevant data using Semantic Web technologies.
An industry like fashion is actually a great example of how the semantic plane of knowledge can be used. The fashion buying decision starts with product specifications – those specifications grow in complexity until a product is ordered. Having the knowledge about the supply of items and the costs that could go into it and sharing that efficiently with a global manufacturing base would be much more interesting for the master data process for that industry. Other social networking tools can be used as well to drive to consensus on what can be built and at what price.
We think it is time for individual companies to look at how the processes of sharing master data with their suppliers and customers at all points in the product lifecycle can benefit from the concepts of JITS, Semantic constructs and social networking tools. At Dhara, we are open to using the social networking tools of “Web 2.0” to carry this conversation forward. If you are interested in this as well, please email us at semanticplane@dharacg.info.
Fred Geiger and Sastry Dhara
www.dharacg.com
Comments