|
|
STRATEGIES: XML
XML's Return on Investment
XML removes the barriers to entry for small and mid-sized enterprises that want to participate in electronic trading networks. Why does that matter to large enterprises?
By Rik Drummond, Contributing Editor
Why is XML's ability to support small and medium enterprises (SMEs) in the value chain (that is, supply chain, distribution channel, and Electronic Funds Transfer) so important to the large enterprises using EDI? Why should large enterprises (LEs) care if their SME trading partners are participating electronically?
Qualitatively, this isn't hard to answer, but quantitatively, it's harder to evaluate. Ultimately, it's important for SMEs and LEs alike to trade electronically.
Expanding the universe of trading partners
I've strongly believed since the advent of XML that XML can save large amounts of money for any enterprise. But what's the value proposition for smaller and mid-sized businesses to participate in electronic marketplaces with large enterprises?
The reality is that in today's value chain, half of all transactions come from SMEs -- not large -- organizations. If the SMEs aren't able to handle electronic transactions, then three things happen:
- There are a lot of transcription errors in the value chain.
- The value chain is vastly sub-optimized from all participants' point-of-view.
- Paper cost vs. electronic transaction processing cost is much higher, especially for LEs.
Those three points indicate that no one gets anywhere near a sufficient ROI. So how does XML figure in? Many of us believe XML reduces the cost for SMEs to participate in e-business, compared to what EDI offered. Unlike XML, EDI was cost-prohibitive and difficult for SMEs to integrate into their systems. Let's look at the three problems outlined above in more detail.
Transcription errors
This is one area that EDI has helped reduce significantly over the years, but there are still too many transcription errors that occur, which makes the cost of managing a value chain very high. Most studies show that fixing an error can cost 15 times the actual amount of the transaction. That's independent of EDI or the type of electronic communications used -- since it requires human intervention to fix errors.
Even though SMEs are increasingly online, many haven't integrated their front-end e-commerce software into their back-end systems. They often transfer the data between the systems by hand or by a cut-and-paste method, which introduces transcription errors in often the same manner as if those companies were using paper transactions.
Let's assume that an electronic value chain with no errors has a cost to manage of 1, which is the best one could achieve in a perfect world. Here's the formula for calculating this:
| (The cost of no errors = 1) * (the percentage of transactions without errors) + (the cost of errors = 15) * (the percentage of transactions with errors) = the relative cost of the value chain to a perfectly run value chain. |
If 1 percent of the transactions are in error, then it means that it costs (1*0.99) + (15*.01) = 0.99 +.15 = 1.14 times the cost to run the value chain over the lowest cost value chain possible.
If the errors are in the range of 5 percent, which isn't unusual, then it costs (1 * 0.95) + (15 * 0.05) = 0.95 + 0.75 = 1.7 times to run the value chain over the perfect value chain.
That means any transcription errors greatly increase the cost of running the value chain. So, if a value chain costs US$1 million per year to run without errors, then it would cost US$1,140,000 to run with 1 percent of errors, and US$1,700,000 to run with 5 percent of errors.
In some value chains, 50 percent of all costs are from fixing errors, which represents an error rate of about 7 percent. One of the reasons errors occur is human data entry. The more times a human interacts with copying the data, the more chance there is of transcription errors. Many SMEs hand copy or cut and paste data from/to their back-office systems into EDI systems, which introduces errors. If those companies integrated their e-commerce systems with their back-office systems, the cost would drop, subsequently decreasing the cost of managing the value chain.
The value chain is sub-optimized
Following Metcalfe's Law, the usefulness of a network is the square of the number of participants , which is n2. This means that non-participation of even a few of the community reduces the value of the network to all. For example, Metcalfe's Law argues that a network that becomes twice as large actually increases its value four times.
Now, many e-commerce value chains aren't networks, but are instead trees, so n2 may not be appropriate in all cases. In some cases, there may be just n. Many e-commerce value chains are not yet (they will become more so in the future as e-commerce matures) complete fully meshed networks where everyone communicates with everyone else, so in some cases at this time their usefulness computation may be closer to n and not n2.
Let's assume that if everyone in a community participates in the e-commerce value chain, that it has a value of 1. The 1 indicates the best possible use of the value chain. In some of the most aggressive value chain implementations, 80 percent to 90 percent -- not all -- SMEs get online. The movement then comes to a standstill as the remaining SMEs resist because they don't see a solid return on their investment.
The formula for computing the usefulness of the network is:
| (Number of possible participants in the trading community)2 = usefulness of the network. |
Following Metcalfe's Law, if only 90 percent of the possible participants actually participate, that means the network is only 81 percent as useful as it could be. Likewise, if only 80 percent of the possible participants actually participate, then the network is only 64 percent as useful as it could be.
Hence, without the participation of even a small number of trading partners -- say one or two percent -- it still impacts the network's usefulness. However, all but a very few companies at this time have all of their trading partners online. It isn't unusual for only 20 percent of the possible trading partners to be online in any trading community.
Paper vs. electronic transactions
Paper can cost four-to-10 times the amount to process over an error-free electronic transaction. That means that at 80 percent participation and 20 percent non-participation, the value chain is 2.2 times as costly to operate from the LE's point-of-view than if all trading partners were online. If 40 percent participated and 60 percent didn't, then 1 * 0.4 + 7 * 0.6 = 4.6 times the cost to run the value chain from the LE's point-of-view than one with everyone online electronically.
Here's the formula to compute this:
| (The cost of no paper = 1) * (the percentage of error-free electronic transactions) + (the cost of errors = average of (4+10)=7) * (the percentage of transactions with paper) = the relative cost of the value chain to a perfectly run value chain |
Conclusions
Following the computations above, what does it mean to a large enterprise if all trading partners are combined online?
We can gauge the cost of handling paper and errors to run a normal value chain with only some of the SMEs online and without integration between their front- and back-office systems.
Let's assume a fairly normal value chain implementation with 20 percent of partners offline, and 20 percent are those SMEs who have integrated e-commerce front-ends with the back-office operations. The error rate is 3 percent from those 80 percent that haven't integrated into their back-office systems. The computation for doing this is a combination of the paper and the error analyses above.
Thus that means (0.2 * 7) + [0.8 * {0.2 + 0.8 * 1 * (0.97 + 0.03 * 15) } ]=2.47.
So if it costs US$1 million to support a perfect-world value chain, then it would cost more than US$2.47 million to support 80 percent of partners online and 3 percent of errors for those online but not completely integrated. Now, if we make a leap and assume there are other types of overhead for the LE not associated with the above computations -- then network usefulness could be over 1/0.64 percent worse than the above numbers.
It's evident from the above that even though a large enterprise may have 80 percent supply chain participation, the transcription errors being produced by the non-integrated SME and the 20 percent of non-participants who still communicate with paper are costing the LE millions of dollars per year. Those costs can be reduced with the appropriate application of XML technology to the SME to help them integrate into their back-office systems, thus reducing their transcription errors. As discussed, the value of the network is only as good as the sum of all participants. So, it pays to look closely at XML-based products that enable any enterprise -- whether small, medium, or large -- to participate in an electronic trading network. That's when the word "value" starts to truly add meaning to the term "value chain."
Rik Drummond, the CEO of the Drummond Group, Inc., works with companies and associations to construct B2B strategies, build architectures, and drive interoperability in electronic commerce. Rik is on the Steering Committee for ebXML, is the chairperson of the IETF's EDI over the Internet workgroup, and leads the EDIINT Interoperability testing. Drummond Group offers a vendor-neutral integrated approach. This involves a concentration on the intersection of technology, business processes, and culture of an organization to build strategy that reduces risk, lowers cost, and fosters e-business. Their client list includes Fortune 100 companies to start-ups. http://www.drummondgroup.com, drummond@onramp.net. ARTICLE INFO
 FREE ACCESS
Keyword Tags: development, e-business, E-Business Management, EDI, ibm websphere, it networking, microsoft .net framework, web development, xml, XML
|
|