the job from hell minus thirty degrees
Lately things have been getting hectic and each day "the job" is converging closer to a nightmare. Today's progression in four acts is a stereotype not just for the parody of customer service, but for any kind of dispute management you might conjure.
Scenario (last month): Our bleeding edge e-business integration product X came out way before anyone in support gets a chance to have any exposure to it, leave alone try installing it or deal with any kind of issues. To make the comedy more dangerous, the product is expensive enough and the customers few enough to justify ignoring the whole product line for all practical purposes.
Cadence (last week): Sales actually manages to lie their way to a few strategic pockets who acquire it, rapidly test their futuristic applications with X and off they go. Experiencing crashes and hangs they realize that they can't fulfill their mission and they call in desperation expecting not only to be walked through complicated setups that none of us have ever performed before, plus expecting to find out immediately what causes their specific problems and how to fix them right away.
Counter-theme (yesterday): Support makes some feeble attempts at contacting the product's developers through the QA liaisons, knowing that they're busy to their ears in the next version already, since any product that has passed QA is already obsolete in their perspective and they are totally bored with any problem reports.
Drama (today):Customer sends an ultimatum with a deadline and many threats, not just to end the contract, but to get everyone involved in all sorts of uncomfortable situations. I end up calling the customer and this is roughly how it gets played out between the wily Techie and the irate Analyst:
Techie (me): I'm sorry about the problem you are having with the install. You need to uninstall the previous version and also have super-user rights.
Analyst: I have used your software for over a year and never had to do that before. Is that something new, if so, why isn't it in the docs?
Techie: (sheep led to slaughter) I'm not sure, that's what QA recommended. I was root when I installed it and it worked...
Analyst: (axe swinging) Then you didn't try hard enough. As we told you numerous times, we are always running into problems with the new line of products. We are experiencing mysterious crashes when there is no server activity late at night. We need to know of other customers who are experiencing the same problems and what their resolution was.
Techie: (more sheep live here) hmmm... no activity whatsoever? I've never heard of that.
Analyst: (ironic) What, never heard of an idle server with no activity??
Techie: (yes we play ping pong) No, I meant I never heard of a server crashing because there was no activity.
Analyst: (bile and vomit are free too) Are you saying I'm lying?
Techie: Not at all, but I think my suggestions for server crashes are still A... B... C...
(at this point I am likely to get a little creative). Your first report was that the server was crashing under a load TEST, not under production. (I am starting my offense). Now you are saying your production server is crashing. Why wasn't the issue reported from your testing first before going to production?
(here I could hear the phone line humming withdeep thought at the other end)
Analyst: (It's not exactly production, it's more like a limited release. What I'm telling you is that it's not purely a testing issue, it is now becoming a production issue and my boss expects a resolution today. Additionally we are still seeing performance issues with your new product X, because we require a transaction time of less than 2 seconds, the previous result was 5 seconds, we now see a 15% improvement with the patched jar, our transactions should be less than 1 second, (more bullshit statistics following in rapid incoherent succession).
Techie: (losing track of the point being made). So you are going to test with the new patch?
Analyst: (sarcastically) I just said two seconds ago that we saw a 15% improvement. If you do the math, that's now like 4.25 seconds where it was 5 seconds before.
Techie: (echoing the sarcasm) So you need a 1000% improvement.
(by this point in time I realize that I am physically shaking with repulsion. Several people have called in the meanwhile and I was seeing their caller ID on the other line: the account manager called, the director called, the people from QA emailed me, my wife called me. It seems my whole little universe has suddenly become interested in me and is aware of my every twitch and convulsion attempting to play along with this customer. Everyone expects a certain outcome from this call and odds are stacked on both sides.)
Techie (re-starting the offensive disguised as helpful) QA said the best way to address your issues is to run a newer VM which we don't actually officially support. They saw similar problems being resolved this way.
Analyst: I'm not going to run an unofficial VM which you won't support with your product. Given our tests, I'm not even sure how you can call your product scalable, even less so distributed computing. You're creating objects within loops of code and that is a no-no in object oriented frameworks, your product does not perform as promised by yor sales people and the documentation is the worst and most lacking I've ever seen. The fact that it crashes mysteriously only complicates and prevents any useful testing we might have done otherwise.
Techie: (at the end of my fucking rope, with a dry hurried tone) I'll pass on your documentation suggestion and to be honest with you if I just wanted to play by the official rules I would not make any suggestions myself, but I will pass on your comments (to whomever happens to be as masochistic as to listen to them). Additionally I will open a bug report about the product X installation problem. I'll get back to you as soon as I hear any suggestions from QA.
Analyst: I have your account manager on the other line. Awaiting your fixes soon.
(the call ends on a surprisingly even note.
Scenario (last month): Our bleeding edge e-business integration product X came out way before anyone in support gets a chance to have any exposure to it, leave alone try installing it or deal with any kind of issues. To make the comedy more dangerous, the product is expensive enough and the customers few enough to justify ignoring the whole product line for all practical purposes.
Cadence (last week): Sales actually manages to lie their way to a few strategic pockets who acquire it, rapidly test their futuristic applications with X and off they go. Experiencing crashes and hangs they realize that they can't fulfill their mission and they call in desperation expecting not only to be walked through complicated setups that none of us have ever performed before, plus expecting to find out immediately what causes their specific problems and how to fix them right away.
Counter-theme (yesterday): Support makes some feeble attempts at contacting the product's developers through the QA liaisons, knowing that they're busy to their ears in the next version already, since any product that has passed QA is already obsolete in their perspective and they are totally bored with any problem reports.
Drama (today):Customer sends an ultimatum with a deadline and many threats, not just to end the contract, but to get everyone involved in all sorts of uncomfortable situations. I end up calling the customer and this is roughly how it gets played out between the wily Techie and the irate Analyst:
Techie (me): I'm sorry about the problem you are having with the install. You need to uninstall the previous version and also have super-user rights.
Analyst: I have used your software for over a year and never had to do that before. Is that something new, if so, why isn't it in the docs?
Techie: (sheep led to slaughter) I'm not sure, that's what QA recommended. I was root when I installed it and it worked...
Analyst: (axe swinging) Then you didn't try hard enough. As we told you numerous times, we are always running into problems with the new line of products. We are experiencing mysterious crashes when there is no server activity late at night. We need to know of other customers who are experiencing the same problems and what their resolution was.
Techie: (more sheep live here) hmmm... no activity whatsoever? I've never heard of that.
Analyst: (ironic) What, never heard of an idle server with no activity??
Techie: (yes we play ping pong) No, I meant I never heard of a server crashing because there was no activity.
Analyst: (bile and vomit are free too) Are you saying I'm lying?
Techie: Not at all, but I think my suggestions for server crashes are still A... B... C...
(at this point I am likely to get a little creative). Your first report was that the server was crashing under a load TEST, not under production. (I am starting my offense). Now you are saying your production server is crashing. Why wasn't the issue reported from your testing first before going to production?
(here I could hear the phone line humming withdeep thought at the other end)
Analyst: (It's not exactly production, it's more like a limited release. What I'm telling you is that it's not purely a testing issue, it is now becoming a production issue and my boss expects a resolution today. Additionally we are still seeing performance issues with your new product X, because we require a transaction time of less than 2 seconds, the previous result was 5 seconds, we now see a 15% improvement with the patched jar, our transactions should be less than 1 second, (more bullshit statistics following in rapid incoherent succession).
Techie: (losing track of the point being made). So you are going to test with the new patch?
Analyst: (sarcastically) I just said two seconds ago that we saw a 15% improvement. If you do the math, that's now like 4.25 seconds where it was 5 seconds before.
Techie: (echoing the sarcasm) So you need a 1000% improvement.
(by this point in time I realize that I am physically shaking with repulsion. Several people have called in the meanwhile and I was seeing their caller ID on the other line: the account manager called, the director called, the people from QA emailed me, my wife called me. It seems my whole little universe has suddenly become interested in me and is aware of my every twitch and convulsion attempting to play along with this customer. Everyone expects a certain outcome from this call and odds are stacked on both sides.)
Techie (re-starting the offensive disguised as helpful) QA said the best way to address your issues is to run a newer VM which we don't actually officially support. They saw similar problems being resolved this way.
Analyst: I'm not going to run an unofficial VM which you won't support with your product. Given our tests, I'm not even sure how you can call your product scalable, even less so distributed computing. You're creating objects within loops of code and that is a no-no in object oriented frameworks, your product does not perform as promised by yor sales people and the documentation is the worst and most lacking I've ever seen. The fact that it crashes mysteriously only complicates and prevents any useful testing we might have done otherwise.
Techie: (at the end of my fucking rope, with a dry hurried tone) I'll pass on your documentation suggestion and to be honest with you if I just wanted to play by the official rules I would not make any suggestions myself, but I will pass on your comments (to whomever happens to be as masochistic as to listen to them). Additionally I will open a bug report about the product X installation problem. I'll get back to you as soon as I hear any suggestions from QA.
Analyst: I have your account manager on the other line. Awaiting your fixes soon.
(the call ends on a surprisingly even note.