by Doly » Fri 07 Jul 2006, 05:01:02
Just posted how tech support in my company is done. Hope they post it. It's simply WTF.
In case they don't, and for your entertainment, here it is:
I work in first line tech support of a large multinational corporation. For reasons that only High Management could comprehend, first line support has been outsourced to one company, while second line support is outsourced to a different one, and third line support is shared between the two in ways too arcane to explain in a short paragraph.
The final result of this arrangement is that if a user has a problem with his/her computer, what will happen is this:
1. The user calls us (the helpdesk in the UK, ran by company A), often fully aware that the guy that will actually solve the problem is in the same office, a couple of floors away.
2. We log a case in our system, and send it to second line in Canada, ran by company B.
3. Somebody in Canada calls the user, figures out that the problem can only be resolved by somebody physically there, and sends it to the local guys. For this, the second line tech has to copy and paste manually the case from the logging system used by company A to the logging system used by company B.
4. If everything goes well, the user will get support by the guy a couple of floors away. This "if" is very important. Several things could easily go wrong in this process, and often do:
a. The user, being a practical person, calls the local tech first. The local tech actually solves the problem and tells the user to log a case so that it will appear on the system and he gets his proper bonuses. The user calls the helpdesk, thinking that it's all straightforward, even giving us the name of the local tech. We log the case. But the guys in Canada have been instructed by company B (for reasons I cannot imagine, since the local guys also work for company B) to close all such cases if the case has already been resolved, and it never reaches the queue of the guy that actually did the work, and the local tech is represented as a lazy b***. As you can see, the system is perversely arranged to maximize the time it takes to support anything.
b. The computer that has the issue, through some oversight, has not been entered in the inventory. The guys in Canada will flat-out refuse to deal with a case where the computer information hasn't been filled in, even if that happens to be impossible because the computer is not on the database. I had one such cases recently, where the computer in case was a file server, and the user needed to recover something from the backup tapes. I had to explain to him that his request was impossible to perform, not because it was technically impossible, but because it was burocratically impossible.
c. The user happens to be in a different office from his/her usual one, travelling for business reasons. Even if you put clearly on the call notes that the user is somewhere else, it will get routinely ignored and sent to his/her usual office. There seems to be no way in the system of company B to change the case to a different location, so a new case needs to be opened once the mistake is detected. And start again the same process.
d. The user doesn't get any support. Naively, the user believes that calling the helpdesk he/she will be able to chase the case. This is an illusion that we are supposed to maintain, though in practice this is simply impossible. For a while, we were able to access the logging system of company B through the Internet. The information we got there was only call notes, and lacked critical details such as who was dealing with the case, but at least was something. But since a few weeks ago, we can't see cases any more. I have flagged this to company B, and nothing has happened. I have flagged it to my teamleader, and he asked me with wide open eyes: "But what do you want to see those cases for?"
e. In the worst case scenario, the user won't be in a big office, but one of the small ones that doesn't have any permanent tech guys. Then, company B is likely to subcontract some or other local company C to do the job for them. In that case, the chances of figuring out what is actually happening if anything goes wrong are exactly equal to zero.
I still haven't seen Dilbert wandering round my office, but I expect to see him any day.