Fixing Customer Service Costs Money, But Not As Much As Angry Customers

Our apps and mobile website allow people to leave feedback. I see the comments every day, and they can be quite interesting, occasionally useless and sometimes irritating. But the ones that make me cringe the most are the ones that have no issue with the app or website, but with our customer service.

I can fix bugs and improve features, but I have no control over customer service and it makes me feel terrible the pain that people go through. Why do companies skimp on helping their customers when pissing them off usually makes them tell all of their friends how awful the business is and destroys whatever good will advertising might create.

Not pissing off customers seems a no brainer but for whatever reason customer service is viewed as an expense to be minimized. My best experience with this was a project I did for Cheaptickets back in 1998-1999. Note that this story has nothing to do with the current brand owned by Orbitz.

At the end of the 90's CT was based in Honolulu and made their money selling air reservations, generally involving Hawaii travel. The had a big call center but little or no web presence. Their phone agents were paid by the reservation which meant they had no incentive to spend any time with a caller if they didn't seem to want to make a reservation really quick.

If a customer got a ticket (most of the tickets were printed and shipped in those days) and couldn't use it or had to change it they would want a refund. However the company's refund process was a nightmare; on average it took 3 months to get your money back and during the process no one had any clue what the status was. People would call the call center and as soon as they mentioned refund the agents would hang up. These customers would be livid and likely told everyone how crappy CT was.

Another issue was that people would take advantage of the broken system and claim refunds even when they had used the ticket, sometimes also doing a chargeback on their credit card. Then they would fly and get two refunds. Often CT had to pay just because they didn't research the chargeback fast enough.

CT didn't seem to care for the longest time, as there were plenty of new customers to be had. Eventually the sheer mass of angry people started to put a crimp in their income. Very reluctantly the leadership decided to hire someone to do something about the horrible process. But they would only pay for one person and a limited amount of time. They hired my employer and I was to be that person.

To this day I can safely say I will never again see a more awful mess of a business process. Four woman slaved away at a system they had built themselves out of whatever technology they could grab and understand. They used paper, the SABRE reservations system, Access, spreadsheets, boxes of unsorted records, various printouts and worked 100 hour weeks trying to handle the refunds. Management didn't want to waste any money on something that brought in no revenue, so these four did the best they could with nothing. I spent several days shadowing them and couldn't believe what pain they went through. It was obvious why the process took so long, everything was manually done. Data about any particular refund appeared in many different places and there was no index of any kind and things got lost. Airline records and credit card information were in paper form and took hours to search. The worst problem was the limited time that credit card companies allowed for chargeback research. The woman who did these did the best she could but every week thousands of dollars were lost because she couldn't work fast enough. Talk about a stressful job.

Once I understood what all the data was I did some research and found downloadable data sources for the paper records, such as ticket usage and the credit card data. I built a database and services to get the data regularly. Then I put a system together for any refund requests to be logged as soon as they came in that started a process. Every piece of incoming data was now tied to that initial request. I made a simple web application that anyone in the company could access and find status based on any bit of data a requester had such as ticket number, name, dates, whatever. This way anyone could quickly answer a query.

I added to the system the ability to automatically construct and print requests for the refundee, or send as email, that filled in all the details. These were requests for more information, please send in the ticket, etc. Since everything was now being logged on arrival once enough information was there it would appear in a work queue where a decision could be made on the refund request. If it was granted it would be forwarded to the manager for approval and then routed to another system for automated crediting. All data would be searched for and attached to the request so there was little manual searching needed. I even came up with a filing cabinet process to organize the paper items.

In testing it looked like it would around a week or so to grant or deny a refund. I was a hero to the four woman; it was probably one of my favorite projects as it solved both a business and personal problem. No more angry customers and stressed out employees!

Plus I got to work in Honolulu for four weeks (and four more at home).

The bad part of this was that halfway during the project management wanted a major change that I estimated would take an extra week to complete and they refused to pay, so they had to finish that last week themselves (I left instructions) and even then they didn't pay the invoices for weeks.

Sadly I never heard what happened after that. Hopefully the system went into production as planned, but I have no idea. A few years later CT was bought and I expect whatever was left of my process probably vanished.

Clearly the lesson was that if you piss off enough customers it will eventually affect your business, and trying to be cheap in this area is pretty shortsighted. Customers generally bring more money than they cost in support and it's easier to keep selling to happy customers than acquiring new ones. Even a programmer like me knows this simple equation.

If I get crappy service I surely (and surly) go elsewhere and tell my friends about it; today it's so easy with social media to spread the word I would think companies would get this point better today than in that mostly pre-internet era. Yet bad customer service is still more common than the good.

Keeping customers happy costs money, but happy customers bring you money, and angry customers bring torches and pitchforks. Your choice.