Real customer service
A few days ago, a new Progressive commercial caught my eye. It is the one where “the other guys” insurance company claims they invented what Progressive does before anyone else. You see their pants catch on fire; and I’m reminded of an old childhood memory of Liar, Liar, Pants on Fire.
Then between Thanksgiving day and just the other day, I’m hearing about how NTT/Verio has thousands of accounts completely down for days with phones not being answered, back end status reports not being updated, etc.
That brought back memories of other companies — in one way, shape, or form — stating they provide customer service or fanatical support or some other catchy “marketing” phrase. Yet, they turn out to be like career politicians who say what they think voters want to hear.
The last time I checked there was approximately 30,000 hosting providers in the United States alone; a number of them stating they are “managed.” I’m sure if you asked each one if they provide good customer service, they will answer yes. But what does that mean?
What is “real customer service?”
We try to share with our daughter, whom we adopted as a young teenager, actions speak louder than words. To share what we at Dynamic Net, Inc. believe are examples of real customer service in action, I would like to present three recent cases.
Fulfilling the silent request – there has to be a better way
Our Global Security Service has been in production for some time now. This world wide security patrol is extremely accurate rarely having false positives.
Yet, false positives do occur; and we would have customers call us asking why such and such an IP address was blocked. The more technical savvy of our customers would take it as far as saying they checked their firewall, and count see the IP being blocked, but could not determine why. Or when told of how to know it is our global security service, since the steps for removal was manual, would run into trouble in that area. So they would call us.
We would go in, and then spend time verifying the situation; and if a true false positive, remove the IP making for a happy customer.
Yet, real customer service is going the extra mile — going above and beyond the request — there has to be a better way.
Most of our customers use the Advanced Policy Firewall by R-fx Networks.
Rather than using the firewall directly (the old — well current at the time — way), what if we used APF itself to add and delete entries? That way APF would record the block in its log files (in addition to our own); and provide the customer with an easier way to identify a block. If the block is a false positive, then they could use APF tools to remove it (if they don’t want to wait out the timer)… wouldn’t that be a better way?
Yes, going the extra mile with additional programming would make it easier for our customers as well as our own staff.
Programming complete, testing and quality assurance checked, and now our global security service customers using APF hopefully have their silent request — there has to be a better way — fulfilled.
Holiday season, slow site, their customers getting disconnects – identify the problem, solve it, but then what?
On Cyber Monday, we get a call from one of our managed VPS customers asking us what operating system updates did we perform on Black Friday. They shared their site has been slower than normal, and some of their customers were getting disconnects.
Unknown to us at the time, the DoS vulnerability patch for BIND/DNS we applied the Monday before Thanksgiving deleted a critical network configuration file on one of our name servers. Something that didn’t raise its nasty head until this past Tuesday evening when we rebooted our servers for an operating system update.
Digging into the problem revealed nothing on the VPS or physical server itself was the source of the problem, but the reboot of one of the name servers with the missing configuration file meant one of our name servers was down.
(As an editorial aside, DNS — domain name services which is like the postal service of the Internet in so far as determining what specific address / destination is to be used based on what the Internet consumer / user is trying to accomplish — is round robin.
DNS being round robin means that even though you might see words like primary, secondary, tertiary, those words do not hold their true definition. Each name server listed with a domain name at the registrar gets used in turn (round robin) without any regard to the words, primary, secondary, tertiary.
So if you have a down secondary or down tertiary even with the primary one up, there will be problems for any person / client / end user whose turn it is to use the one that is down).
While we had monitoring of the physical machines used for the name services, since we never had a case of a name server IP address being lost or missing, we didn’t have quality assurance checks in place.
While real customer service is not about excuses, it is about telling the truth, being real, identifying the problem, and fix it. Shouldn’t real customer service try to prevent future problems if it could?
We restore the network configuration file, and fix the problem. While we don’t have control over what a patch or update may or may not do, we did take control over being aware of the situation.
We added to our monitoring system checks for the name server IP addresses so we would be alerted, in real time, when a name server IP address was lost or missing.
We alert our customer with a full report of what happened, when we believe it happened, how we fixed the problem, and the steps we put into place for the future — handling the then what?
In return, we get a happy customer who states:
“We very much appreciate your prompt resolution of this issue and for putting the proper checks in place to assure that it won’t occur again.“
Avoiding telling customers “that’s the way it is”, rather dig in, provide meaningful details, and hold the customers hand through it
We’ve been using Parallels H-Sphere — a complete hosting automation system — since the year 2002 when it was owned by Positive Software founded by Igor Seletskiy (who is now the CEO of CloudLinux).
A hosting automaton system with billing automation that doesn’t work gets noticed quickly; it would be like having a payroll service that could not calculate paychecks correctly. H-Sphere has a relatively excellent track record for being accurate in its billing (I use the word, relative, because in the past nine years there has been two bugs that were fixed promptly related to billing).
So when a customer calls asking why their email bandwidth is now around 40 GB per month when it used to be 2.5 GB per month, it would be easy to say “that’s the way it is” or to otherwise put the burden on their shoulders (i.e. you prove it’s wrong).
Is that real customer service?
First off, while earning revenue fairly matters, money is never the end all (people matter more). While we do have terms of service, acceptable use policies, and other legal words in place to allow us to bill for bandwidth overages, we do look at any question of overages on a case by case basis.
Since the customer’s regular usage prior to November 2011 was 2.5 GB per month, clearly something is going on for the bandwidth now to be so high in comparison. While the billing calculations might end up being correct, let’s put the customer’s best interest at heart and credit their account for what we believe might be the overall overage rounded up.
That way, the customer doesn’t have to worry about any financial issues while we hold their hand and walk with them through this event.
Next, we work with the customer to alleviate any fright over whether or not hacking is involved. We determine the only IP addresses logging into email is that of the customer.
In the process, we did find out the customer was using extremely weak passwords; and we alerted them to that along with offering to help change them (we did tell them what makes a password more secure — at least 12 wide, mixed case, alphanumeric, special characters when able with zero use of words or phrases).
From there we looked daily reports to identify the specific day of the week in November 2011 the bandwidth spiked up above normal day to day use.
Once we identified the start day, we were able to see from that day forward, while the bandwidth was much higher (i.e. 3 GB per day vs. approximately 500 MB per day), that from the day the issue started, bandwidth usage was consistent (not a spike or otherwise trending upward).
Checking logs we determined the bandwidth utilization was not from sending email. Narrowing down the date the issue started, along with the issue now being email receipt, we looked into our DNI Mail Complete (an enterprise level anti-spam service) system logs.
What we found were a number of emails, all appearing to be from valid senders, who were sending emails that appeared to have large attachments (it needs to be noted we did not see or otherwise review the contents of the emails); and that was communicated with the customer along with the from and to email addresses involved.
For now it is in our customers hands, but through this experience we held our customers hands, showed by our actions its not all about the money, and gave them the details they need to make a decision of what to do next.
Real customer service needs to be patient, kind, gentle, and loving words that are put into positive actions for the best interest of the customer.
Please contact us for more information.