Show me a corporation and I’ll show you where they need, they want metrics.
Once upon a time, when an issue occured with a security patch or service pack and you didn’t know immediately what was wrong and you were not aware of the nntp newsgroups you called Microsoft. They got the call. They’d set up a case. You got help. They had a metric.
Show me the current state of social media and I’ll show you that trying to gather metrics from social media is a study in estimates at best. Take for example my recent exercise to take an accurate head count of those people impacted by the “C34” error in Windows 7 sp1 after WSUS was used to deploy it. I could tell from the folks I personally knew that were impacted that few of them called into Microsoft. When you ask a consultant they will tell you that it’s inefficient to call into Microsoft. The client/customer needs the PC working now, not a several hour call back later. So they do what is now the norm for today’s world. We search. And in moving from a support venue that has exact counts of impacted customers and contact information via SRX case numbers to the world we live in now of Google, Bing, Twitter and Facebook I’d argue we’ve lost that metric value and not replaced it with anything that comes close.
Fast forward to today. Twitter is now seen as a support venue. Facebook and forums are the new normal. But yet in these venues I’d argue it’s near impossible to gain a true count of issues. Again using the Windows 7 sp1 c34 issue as an example…go to twitter and search for http://twitter.com/#search?q=0xc0000034 . Count the tweets. We have our metric right? Not so fast. Look at the tweets. Many are retweets of headlines. So how do we know how many of those in twitter are really impacted or just echoing what they see someone else tweeting?
Okay so let’s search on blogs… http://www.google.com/search?hl=en&ie=UTF-8&q=0xc0000034&tbm=blg Count the posts….. oh wait… hold on… again the posts seen there may be and truly appear to be echos and reblogs of other blogs. How many of those are actual impacted customers?
Now let’s look at google trends of the 0xc0000034 over the last 30 days. That has to find something right?
http://www.google.com/trends?q=0xc0000034&ctab=0&geo=all&date=mtd&sort=0
Wait. Nothing? No results in the last 30 days at all? Okay so what about 2011 as a whole?
http://www.google.com/trends?q=0xc0000034&ctab=0&geo=all&date=2011
Hmmm not much better, okay so we can’t see any pattern or blip.
Okay so now what? Okay so let’s hop on over to the forums and email all of those people in the forum and ask them. We see these threads where people have posted….
http://social.technet.microsoft.com/Forums/en-US/w7itproinstall/thread/1c9a7151-b48c-4a98-aae7-a4b82682ea8e
http://social.technet.microsoft.com/Forums/en-US/w7itprogeneral/thread/fb81fb04-2854-48f6-a627-7be66baf43e8
http://social.technet.microsoft.com/Forums/en-US/w7itprogeneral/thread/20d3f158-c198-4f4a-af0f-9c95e2ed8d3f
http://social.technet.microsoft.com/Forums/en-US/w7itprogeneral/thread/c06fb719-bc91-4c4e-8fe7-a385d139617b
http://social.technet.microsoft.com/Forums/en-US/w7itprogeneral/thread/608ecca8-b815-4ff6-8f3c-a828518434a7
http://answers.microsoft.com/en-us/windows/forum/windows_7-windows_install/fatal-error-c0000034-applying-update-operation-282/2750aef6-958e-43bf-a890-2dac95c5f43d
http://answers.microsoft.com/en-us/windows/forum/windows_7-system/fatal-error-c0000034-applying-update-operation-282/bd99ead9-ece9-48af-9ad8-3c0377d01301
Oh wait, we don’t have their email addresses as that’s private and the forums don’t support private messages.
And in the case of the new Microsoft answers site — see that “me too” button on there?
Can we use that as a metric? We can surely use that as our count of people with issues right?
Wait, not so fast. You see when the new Answers 2.0 forum opened up those of us that answer questions that need to know if the poster came back and needs more help found that the only way we could set up an alert to get an email when someone posts on the thread is to click on that “me too” button. So any “me too” count needs to be questioned as it may be hyperinflated by answerers of questions that really do not have the same problem, but merely clicked “me too” in order to set up an alert to get notified of a response to a thread they posted to. A “Me too” of 5 may not be 5 impacted customers. Instead it may be one guy with a problem and 4 other people that posted to the thread and then wanted to get alerted when someone else posted on the thread. Yeah, not a good metric to count on at all.
Okay so let’s hop on over to facebook and see what the chatter is over there.
http://www.facebook.com/#!/search.php?q=fatal%20error%20c0000034&init=quick&tas=0.47961585820526425
hmmm okay so that’s only coming back with what we already know. If people use facebook as a help venue, we can’t see the results.
Get the idea here? Searching may have revolutionized how we get our computer help these days, but I’d argue it’s not helping to get a better count and feel for true customer impact.
On the one hand we don’t want to build in a big brother service that people would object to, but at the same time, no social media venue I see is ensuring that it can track number of computers impacted, OS’s involved, or any sort of basic metric information that anyone would need to make decisions.
If social media will rise in the scale and use of a support venue, and I’d argue it will, then someone needs to take the bull by the horn and ensure that these venues don’t turn into echo chambers and instead include tools that get real metrics and customer counts and build in a means to get truer counts than there is now.
In my unscientific process of posting out on this blog, the www.patchmanagement.org listserve, the SBS2k, smallbizit and smbmanagedservices yahoogroups and asking people to contact me to give me numbers of computers impacted and trying to count the bodies in the posts of various blogs, I came out with a head count of 370 impacted computers. Of those 370, only two computers had support cases set up. 2. Out of 370. Including two customers that had over 70 pcs in one firm nailed by the c34 issue.
Microsoft only sees two phone calls.
I see 370 dead pcs that are now screwed up and need to be dealt with.
A challenge to those in charge of customer support these days. Think about this impact. Build into your data gathering processes better ways to track. Don’t just put up a twitter account and say you’ve done your job. Don’t just build a listening venue, but the online version of that support call system and do real tracking.
When you build your forum venues, consider ways to gather information about impacted pcs, operating sin an easy but not big brother way. Yes, I know that’s asking a lot but you are going to need this … in fact you need this right now in order to better respond to your customers.
“When you ask a consultant they will tell you that it’s inefficient to call into Microsoft.”
Inefficient or they are scared that it will make them look stupid and ruin the client relationship ? They are being paid to know everything right ?Then they rush into a solution and show the client how they saved the day. Two days later they find out from Microsoft that all of the clients workstations are now in a torn state and have to be reloaded. They really saved time not waiting for that call back from Microsoft didn’t they. Yes, sometimes the client is to blame in pushing for a solution but then you should manage the client or walk away.
Microsoft actually has a great system for tracking issues, for issues that don’t involve a machine not being able to boot anyway, called telemetry. If you opt to turn it on. They get all kinds of data that way. Other companies are starting to copy it.
In this case the box wouldn’t boot. No dr. watsons can work then.
And inefficient in that they have to charge the customer.
Again let’s get back to the root cause… didn’t Microsoft make the change in how this service pack was released? Which in turn caused the issue in the first place? Who’s the blame now?
The issue is not about blame and never should be when either one of your machines or another companies machines is down. The issue is finding out what happened and coming up with a solution that doesn’t cause more problems.
Isn’t a call to Microsoft about a servive pack issue free like it is with security patches ? And even if it isn’t how is it going to cost the client more ? You have to charge for Google time don’t you ?
At the time they did it, it was clear it would get the boxes operational, it was only later on that the ‘fix’ was more understood. Sorry Dean I don’t think it’s fair to let Microsoft wiggle out of this one and say sorry, you are screwed, you didn’t call us.
When I’ve called and stated that I’ve had a security patch problem, I can get in without a credit card. When I’ve had a service pack problem, it’s a different call venue and it’s hit and miss regarding no charge.
I don’t think they should say you are screwed either but that is after the crisis. During the crisis blame of any kind should never be an issue. It’s counter productive. At that point the only issue is solving the problem. In the best way. Not necessarily the first way.
Look at the nuke crisis in Japan. Is it productive for them to go around pointing fingers of blame or is it better that they work on getting the reactors shut down ? The talking heads on all of the news shows can do all of the blaming later.
I didn’t come across the problem so I don’t know what you mean by “At the time they did it, it was clear it would get the boxes operational”.
“When you build your forum venues, consider ways to gather information about impacted pcs, operating sin an easy but not big brother way. Yes, I know that’s asking a lot but you are going to need this … in fact you need this right now in order to better respond to your customers.” Listen up, Microsoft! I agree 100%!!