June 8, 2007
Braintrust Query: Are the goals of data synchronization realistic?
By Bill Bittner, President, BWH Consulting
At a session from this week’s UConnect Conference in Orlando, FL, a few retail tech bigwigs shared war stories on data synchronization and their efforts in pushing their organizations to get to one version of the truth.
The panel featured J. Alexander (Sandy) Douglas, president of Coca-Cola North America; Craig Herkert, president and CEO of Wal-Mart of the Americas; and Randy Mott, former CIO of Dell and now head of IT strategy for Hewlett-Packard.
Wal-Mart told how it took two years to reduce the apparel life cycle from design to store sales from 50 to six weeks. Coke emphasized the need for interoperability between corporate and all the bottlers who actually package the products. HP described their reduction of 750 internal data marts into one data warehouse.
Looking forward, Wal-Mart described how they are continually striving to improve operational efficiencies and their RFID initiative continues to expand, going from 1000 to 1400 stores this year. They are confident that RFID payback will be there as tag costs continue to decline (now at 8 cents) and the technology becomes better understood.
But officials also touched on how the internet and globalization can complicate progress. When asked to express their concerns for the future, Coke had an interesting perspective on the effect of internet anarchy. Big brands create big targets and it is getting harder for the brand owners to defend themselves against misinformation that can be broadcast instantly over the internet. Wal-Mart discussed how compressed timelines have become by describing the worldwide release of Pirates of the Caribbean on the same day, with all the promotional merchandise needing to be in place at once.
These are all great examples and they are fun to talk about at a meeting that is being held to emphasize data accuracy, but I think we may be expecting too much. Not every carton, jar or box will be exactly the same as the next. As you keep improving the measuring devices, it reaches a point where each unit is unique because of minor differences. Data synchronization is built on the assumption that there exists one version of the truth, but the fact is that the supply chain is never standing still. The parts are always moving and variations between units will always exist.
Discussion Questions: Are the goals of data synchronization realistic? Can there be a single version of the truth that applies universally? Is there a way to meet the spirit of data synchronization while accommodating the vagaries of the real world?
[Author’s commentary] In supply chain automation, we use a digital tool to manage an analog world. The computer can carry details to a much greater level of precision than is humanly possible or even necessary. We need to recognize in our system designs and our data capture that there needs to be provision for some “fuzzy logic.” We need to recognize that GTINs (Global Trade Item Numbers) and ‘Items’ are not always the same thing. One “Item” can have many GTINs as various net contents are produced to meet a price point and various promotional labels are produced to spur demand. We need to handle the flow of product variations as they pass from manufacturer to distributor to retailer and to consumer. There is never simply one version of the truth in our constantly moving supply chain.
This means there must be “data alignment” that supports the flow of authorized item variations through the supply chain. This kind of flexibility must be built into everyone’s applications so that the whole supply chain looks at items not as a single unit, but as a group of physical units addressing the same consumer needs. Assortment planning, replenishment, and pricing applications must recognize this relationship so they can affect changes across all the units of an item. This type of flexibility will make it possible for all members of the supply chain to better share item data.
BrainTrust
Recent Discussions
Discussion Questions
Poll
I am not sure that there will ever be 100% pure data synchronization throughout a given supply chain process, but I do think there are many processes that can get extended retail industry players closer than where they are now.
The first thing that comes to mind is better business intelligence practices at each point of the supply chain. The use of BI within retail has yet to reach its peak point, but if retailers where able to clean and organize their own data first, and weed out irrelevant information, it might make for better sharing externally.
Second, retailers and their trading partners must learn to play nicer with each other and trust each other. Lack of trust is one of the reasons CPFR has never really taken off (GREAT concept–but still makes IT and business staff jitter at the prospect of sharing data). Trading partners need a greater sense of teamwork before they coordinate their data is an organized fashion. Given the complexities of today’s trading environment referenced in the article, this is clearly a top-line problem that must be reconciled.
I wouldn’t go as far as to say that data synchronization involves the establishment of a universal truth, but it is highly possible for a community of businesses to choose to work together in a form of a network and to seek a high level of synchronization in data description, best practices and standards.
Kanban, a Japanese word for Community Practice, is a well accepted concept universally. Standards established by ISO, or specific industry standards are commonly embraced by businesses big and small. Large organisation like Coca Cola, Wal-Mart can specify standards which they required their vendors to comply with. Such standards, if they make sense–not punitive or anti-competition–can quickly become industry standards.
With globalisation, increasing international trade and relations, networks or communities of businesses will become larger.
I have no doubt that all these developments will stimulate greater incentives for synchronization and alignment. A universal truth is perhaps quite distant or perhaps a moving target given today’s dynamics.
The emergence of China as the “manufacturer of the world” has, however, impacted the dynamics. In order to exploit the low cost, major businesses–even those in the West–are beginning to adjust their rules to accommodate the generally lower standards of Chinese made products.
Ultimately, it is probably not about what is right or the truth but what is feasible, practical and pragmatic at a given point of time.
And if we can in our short term mindedness contemplate once a while on what’s universally good for the world in the longer term, it would certainly be a bonus.
Two factors appear to be hindering data synchronization. First, even in today’s digital world, few companies have the discipline for accurate, clean data. For most management, the cost is too high and the payback does not significantly increase from 95% accurate to 100% accurate. It is for this reason that modern computer systems simply don’t have the level of data editing required to achieve higher accuracy. Not letting any data into the computer until complete and passing all reasonable edits delays and increases the input time, thus cost.
Second factor has to do with data definition. Not the basics of case or pallet, but the meaning. For example, is it sales or set sales? What is included and what is excluded? Each data element in each company has an assumed definition. This variance among companies is greater than the data synchronization proponents would have you believe.
Great topic, Bill. If the biggest factors in the digital flow were data and the things the data refers to and triggers, I doubt we would be having this discussion at all. Instead, our studies show that among 1200 organizations, at least 40%, and an average of 53%, of the time usage in all data flows is organizational and human interruptions of the flow. And note this: in those flows, data directly triggered events, or was directly generated by events, less than 10% of the time — humans interacted in the flow and made decisions regarding the data 90% of the time. This is amazing to me — it suggests that data flows are so weak, or are perceived to be, that humans “have” to interact in the flow 90% of the time.
I think it is easy to see data systems as some kind of separate, isolated, stand-alone system in organizations. At the present, however, our information suggests they are so fused with the organizational error processes that they sometimes create more work and more problems than they save. We have a long way to go, in my view, to making data systems function the way we assume they do, want them to, and hope they do. Only after that will we be able to make them live up to the promises used to sell them.
As more and more companies use the same small group of supply chain and ERP software packages, it will be easier and easier to achieve data synchronization. If Coca-Cola and the majority of its highest-volume customers all used off-the-shelf systems, especially if those systems were hosted by their creators, the challenges would be much easier to solve at a lower price. As a comparison: the airline industry largely shares SABRE, a hosted software package for reservations and ticketing. SABRE was formed by American Airlines using IBM’s infrastructure in the 1950s, but was rolled out to include airlines and travel agents worldwide. SABRE is a successful supply chain data synchronization tool. Wal-Mart or P&G or a similar large company with excellent supply chain technology could leverage that service across the CPG and retailing industry, with the support of a major technology partner (SAP, IBM, etc.)
More Discussions
Has Same-Day Delivery Become an Expectation Across Retail?
Are expectations around retail delivery speed in flux?OpenAI Names the Infrastructure Standard for Agentic Commerce
Will OpenAI or Google (or another entity) set the gold standard for agentic commerce?Is Five Below Poised For Even Greater Growth?
Five Below is hitting serious growth numbers. Can this trend continue?
I am not sure that there will ever be 100% pure data synchronization throughout a given supply chain process, but I do think there are many processes that can get extended retail industry players closer than where they are now.
The first thing that comes to mind is better business intelligence practices at each point of the supply chain. The use of BI within retail has yet to reach its peak point, but if retailers where able to clean and organize their own data first, and weed out irrelevant information, it might make for better sharing externally.
Second, retailers and their trading partners must learn to play nicer with each other and trust each other. Lack of trust is one of the reasons CPFR has never really taken off (GREAT concept–but still makes IT and business staff jitter at the prospect of sharing data). Trading partners need a greater sense of teamwork before they coordinate their data is an organized fashion. Given the complexities of today’s trading environment referenced in the article, this is clearly a top-line problem that must be reconciled.
I wouldn’t go as far as to say that data synchronization involves the establishment of a universal truth, but it is highly possible for a community of businesses to choose to work together in a form of a network and to seek a high level of synchronization in data description, best practices and standards.
Kanban, a Japanese word for Community Practice, is a well accepted concept universally. Standards established by ISO, or specific industry standards are commonly embraced by businesses big and small. Large organisation like Coca Cola, Wal-Mart can specify standards which they required their vendors to comply with. Such standards, if they make sense–not punitive or anti-competition–can quickly become industry standards.
With globalisation, increasing international trade and relations, networks or communities of businesses will become larger.
I have no doubt that all these developments will stimulate greater incentives for synchronization and alignment. A universal truth is perhaps quite distant or perhaps a moving target given today’s dynamics.
The emergence of China as the “manufacturer of the world” has, however, impacted the dynamics. In order to exploit the low cost, major businesses–even those in the West–are beginning to adjust their rules to accommodate the generally lower standards of Chinese made products.
Ultimately, it is probably not about what is right or the truth but what is feasible, practical and pragmatic at a given point of time.
And if we can in our short term mindedness contemplate once a while on what’s universally good for the world in the longer term, it would certainly be a bonus.
Two factors appear to be hindering data synchronization. First, even in today’s digital world, few companies have the discipline for accurate, clean data. For most management, the cost is too high and the payback does not significantly increase from 95% accurate to 100% accurate. It is for this reason that modern computer systems simply don’t have the level of data editing required to achieve higher accuracy. Not letting any data into the computer until complete and passing all reasonable edits delays and increases the input time, thus cost.
Second factor has to do with data definition. Not the basics of case or pallet, but the meaning. For example, is it sales or set sales? What is included and what is excluded? Each data element in each company has an assumed definition. This variance among companies is greater than the data synchronization proponents would have you believe.
Great topic, Bill. If the biggest factors in the digital flow were data and the things the data refers to and triggers, I doubt we would be having this discussion at all. Instead, our studies show that among 1200 organizations, at least 40%, and an average of 53%, of the time usage in all data flows is organizational and human interruptions of the flow. And note this: in those flows, data directly triggered events, or was directly generated by events, less than 10% of the time — humans interacted in the flow and made decisions regarding the data 90% of the time. This is amazing to me — it suggests that data flows are so weak, or are perceived to be, that humans “have” to interact in the flow 90% of the time.
I think it is easy to see data systems as some kind of separate, isolated, stand-alone system in organizations. At the present, however, our information suggests they are so fused with the organizational error processes that they sometimes create more work and more problems than they save. We have a long way to go, in my view, to making data systems function the way we assume they do, want them to, and hope they do. Only after that will we be able to make them live up to the promises used to sell them.
As more and more companies use the same small group of supply chain and ERP software packages, it will be easier and easier to achieve data synchronization. If Coca-Cola and the majority of its highest-volume customers all used off-the-shelf systems, especially if those systems were hosted by their creators, the challenges would be much easier to solve at a lower price. As a comparison: the airline industry largely shares SABRE, a hosted software package for reservations and ticketing. SABRE was formed by American Airlines using IBM’s infrastructure in the 1950s, but was rolled out to include airlines and travel agents worldwide. SABRE is a successful supply chain data synchronization tool. Wal-Mart or P&G or a similar large company with excellent supply chain technology could leverage that service across the CPG and retailing industry, with the support of a major technology partner (SAP, IBM, etc.)