There has been a lot of fuss lately about the list of top companies. Of course it is too late for my comments as to whether a company deserves to be in the top or not to have any impact on the outcome of the list. In any case it’s not up to me to decide on the criteria for inclusion in the list and in my humble opinion it’s only the privilege of the SDN administrators to do so. I would rather focus on the way this list is composed.
Mark Wipro passed 100K SDN Points. Congratulations. in his web log that it “is derived from the primary email address that you have posted in your Business Card (profile). We take everything from the @ until the next point and summarize it.” At first sight this would appear to be a rather objective and neutral approach, but I wonder if, in this case, it really is the correct method. One appears to be putting incomparable things in the same basket. Before everyone gets up on their hind legs, let me explain.
The problem is that there are different types of companies based on:
- the nature of the company. It is clear that a pure computer/development related company is different from a non-profit or manufacturing company.
- the size of the company. An SME is clearly not comparable with a multi-national.
- there is also a difference between SAP itself (as a company), SAP partners and SAP customers
All of this definitely has an effect on contributions, meaning respectively:
- a company consisting only of developers has more resources and expertise than a manufacturing company where the people are only there to keep the SAP system running (or for feeding the system with data) and the actual development is outsourced. It’s not so obvious that such a company can expend valuable time and people in order to contribute to SDN. The effort for such types of customers is thus bigger than for ‘dedicated’ companies.
- More or Less the same can be said for SMEs as compared to multi-nationals. But it can be difficult even for bigger companies. Take our university – which is not exactly a company in the commercial sense of the word – as an example. There are about 8500 employees at the university, but not even 0.5% of them are actually involved in SAP development. And even then I didn’t take into account the university hospital emplyees (about 8300), which brings the ratio down to 0.25%. If you ask me, there are, relatively speaking, not that many people that could contribute.
- It would appear rather obvious that SAP people and SAP partners can put more effort in to contributing. On top of that, they have more (re)sources at their disposal than SAP customers. The top companies list doesn’t make any distinction on that count.
In all fairness I have to say that SAP people don’t qualify for the annual prize-giving.
Where am I heading with all this? Well, the list doesn’t take the proportion of contributors per company in to consideration. Shouldn’t a company with only one contributor that ends up in the top 10 get more respect than a company with multiple contributors? Btw, I’m basing this on the assumption that there is a one to one relationship between an account and a physical person. I sometimes see an account contributing 24/7. It is clear to me that multiple people are using the same account on a multinational base. If not, there must be a robot with artificial intelligence behind the account. Whatever it may be, it is –as with anonymous users – against the philosophy of things.
In conclusion I would like to go back to Mark’s explanation. If we take it literally, then there are some (serious?) gaps in this algorithm. It considers only the second level domain name, excluding the top level domain (TLD). What happens with multinationals that use the TLD to distinguish between countries? It works fine if one doesn’t mind putting all the countries in one pile, but does one always want this?
It could also occur the other way around. Some companies put the country in the second level domain name, e.g. company-country.com. They would wrongfully be considered to be different companies. What if they use a third level domain name to do the same? At my previous place of work we used be.company.tld to make the difference. If we put it against the algorithm, we would end up with ‘be’ as company name. At the university we assign the departments to the third level domain name. Applying this algorithm, we would thus be known as ‘aiv’. Even if the FQDN was used, contributors from other departments would be assigned to other companies, despite being from (the same) university.
Some countries, like the UK, have a working standard with three domain levels. It could be even more complicated. We used to have four domain levels here: department.kuleuven.ac.be (the ac stands for academic). The phasing out is still going on, so there is still a mixed use.
Finally, the TLD can also be used to mark the difference between companies: sap.com, sap.be and sap.org don’t have anything in common.
In this forum , Craig suggested not working with subdivisions in order to be toted up as one company. I wonder thoud whether many companies allow their employees to work without the third level domain name. Craig pointed out in this thread that everything after the @ will be considered for the top company list. If that’s the case, some of my remarks would be obsolete whereas some would still remain valid. A clear statement in a FAQ would prevent confusion.
Nevertheless, I would suggest abandoning the current mechanism and instead simply ask the contributors to update their profile and fill in company and country, and whether they want to be grouped or not.
The World According to SDN. Put yourself on the SND world map and earn 25 points. Spread the wor(l)d!