You want your systems work perfectly smooth and provide the best service to your customers. But real life can be a bit different. Systems can fail, electricity cut off can happen, humans can make errors. Although you continuously work hard to prevent any “problem”, it’s important to invest in infrastructure and mechanisms to identify instantly if any issue is raised.
Let’s go over an example I’ve recently faced with: After the latest patch, an online mobile game started to get disconnected very often.
The life-cycle of this problem is:
- A problem pops up. E.g. A coding bug causes disconnection.
- The problem is identified.
- The problem is solved.
- The solution is communicated to the customers (optional).
In real life, you cannot avoid step 1. Problems will occur. Once the problem is identified, usually it will not take too much time to solve it. Therefore, I believe the most critical step is the 2nd step: problem identification. It is where companies should focus and invest more.
Now, let’s deep dive Step 2 to understand what sub-cases can happen during problem identification.
A state-of-art monitoring system will be costly to implement
2a. Ultimate Real-time Monitoring. A company can have a perfect monitoring system that can instantly alert that an unwanted disconnection case is happening. Since control systems cannot monitor everything, this case is very unlikely to happen.
2b. Standard Monitoring. You do not need to monitor disconnections individually. If you are monitoring standard KPIs like, # of logins, # of active players, minutes of game sessions on a daily basis, a smart analyst will be suspicious when he/she sees a sharp increase in # of logins per user. Analyzing “why/how it happens” will let the analyst find out “disconnection problem”.
2c. Bug reports. Most of the online applications have a “bug report” link or forum topic for this purpose. If enough users report “disconnection problem”, then the problem is identified.
2d. Customer support. When the user faces a problem, customer support is one of the first contact points to resolve his/her problem. Here, the important aspect is that the customer support agent should take the feedback seriously and escalate the topic for immediate investigation.
In my case, the game support failed at this stage. When I contacted customer support saying: “Disconnections started after the latest patch. I tried both 3G and wireless connection, but the result was same. I often disconnect although I don’t experience any disconnection issue in other games I play…”, the response was: “There is no technical problem, please check your internet connection and restart your device”.
The result? My problem is not solved and since the support did not seem to care about my feedback I’m discouraged to contact them next time I face a problem.
Listen to your customers!
Many companies wait for a significant portion of their clients complain before they accept that there is a problem.
It is a big mistake.
Every feedback from the customer is valuable and should be investigated, especially when your monitoring systems are not state-of-art.
It’s also important that customers are encouraged to give feedback.
Provide channels like easy “bug reporting”, “customer support chat” for feedback. Ensure customers feel that feedback is taken seriously. Otherwise, they will not give feedback next time, and it will take longer to identify the problem, resulting potential losses.
Identify the problem fast!
Does not matter how many problems you have or how big they are, if you can manage to identify them before your customers recognize it’s still a good result. You will keep your brand reputation and customers.
So, the key is to identify the problem ideally before the customer does, briefly as quick as possible. Between 2a and 2d, as you move down, you’ll be affected more on profit level.
Investment on “problem identification” is not necessarily a costly one. 2a may require a significant capital investment but all others are just proper implementations of simple analytics and customer service flows.