Data-driven decisions means that all actions are based on collected data that are interpreted and gives insights to what had happened and what could potentially happen. Decisions made based on data is the current strategy of modern business world.
He said “should be” and “I think”. The groundless, optimistic observation made me angry. – Dr Akiu played by Naohito Fujiki in the Japanese drama entitled Innocence, Fight Against False Charges
The excerpt above is from the Japanese TV show Innocence. Of course, the show is in the Japanese language and the line is translated shown in the subtitle. Reading it made me think about our predispositions which are most of the time based on what is available to us and consider it to be true, even if data does not support it.
And data is a powerful tool in making decisions.
Data-driven insights
In a BPO company, numbers are all over the place. Almost everything is based on numbers. Some examples of these are key performance indicators like service levels, average handle time, quality scores, and attendance. These indicators tell how the company, or just an individual team or person, is performing. Decisions are made based on these numbers. It can result in financial rewards such as incentives or salary increases, or at worst, an individual being terminated due to failing numbers.
Data is very useful in determining the state of the business and coming up with a strategy to make it better. For example, in a BPO setting, the most common metric is average handle time or AHT for short. This is basically telling the management team how fast each calls are handled from the time the customer and an agent engage in a conversation up to the time either party hangs up. Usually, a customer service call lasts for five minutes while a technical call lasts for about 20 minutes. If the average goes beyond this, it is worth taking a look and understanding the reason behind it. New process? Technical glitches? New people? From here, a plan is devised and implemented to bring it back to its level before.
The example I provided may seem simple but there could also be some real-world events that are not so easy to decipher. Financials can be grueling data when analyzed. It is not just about a quick understanding of where the money comes from or going but it will branch out to other metrics like sales, retention efforts, customer service, IT infrastructures, and so on. With these in mind, data-driven insights are not just about the numbers. Experience is also needed to explain the numbers.
Background knowledge
For you who have not worked in the BPO industry, the concept of AHT may be something that is new to you. Maybe that is something that you just encountered now and does not really fully comprehend it. But for those like us who have been in the industry our entire careers, AHT is an everyday metric for us. It is something we look at daily. It is in our blood.
And since we are so exposed to this metric, it is easy for us to start looking for the causes of the change. Just like what was mentioned above, new processes that were implemented can affect it making calls longer. New agents can also make it spike since they are still familiarizing themselves with the system and the type of conversations and issues they encounter. Even the implementation of new technological tools can hamper the speed in handling calls as the tool may not be fast enough or familiarization may be required.
Knowledge of what the numbers are showing is then important and a key requirement to interpret the data. Without such, it would be impossible, or at least very difficult, to understand, get insights, and create actions from what the data is telling us. Thus, making decisions that are data-driven is not just about math and data visualization. It requires exposure and experience to the world where the data was from or happening, at least a a general knowledge of it.
Dangers of Data-driven Decisions
Using data to make decisions and actionable insights can also backfire if the data is not carefully understood, if data is made to show something that is directed to an action already assumed before seeing the numbers, and if data is presented in a way that conceals something for the purpose of showing good or bad numbers.
Data not carefully understood. Think about investments especially stocks. Most people buy stocks of companies that are earning at the moment. When we see a stock gained 20-25% from yesterday, we think and consider that as a good stock to invest. And for those that are showing prices gone down in the current year compared to the previous one, we consider it as not a good investment. This is where we go blind.
If we only consider the data that is immediately in front of us, it is easy to make mistakes. Data-driven decisions require us to scrutinize the dat and look at other datasets for comparison. The example of stocks investments above may mean that the company is not doing well this year because of economic conditions (happened to most companies during pandemic), but has a track record of more than 50 years of an average growth year on year by 10-20%. And comparing that company to others in the same industry it is in would give us a better view if the company is worth investing.
Assumed action before data. People fall prey on this especially in business. We have a solution in mind ready to implement and we make the data say it to justify our preconceived solutions. Doing this is a poor decision-making strategy as we manipulate the data to implement an approach that may or may not work because it is based on real data.
Taking AHT for example, when we see it going up and assumed that agents are having a hard time, the data can be manipulated to show that the cause of it would be high talk time. This would trigger coaching and training of the agents to make them more efficient in handling the calls. But what if it is not about handling the call but a new system implemented that is slower than what was used before? This could cause agents to wait for customer information to pull up delaying the service provided to the customers who calls.
Concealing the real story. When data is manipulated to show good numbers for the purpose of having a good trend or story, it could have a very detrimental effect to the company.
Let’s stick to the AHT scenario as an example. If it is shown as month-on-month trend, or even week-on-week, it may show static changes of maybe about a few seconds. This would mean that AHT is stable and there is nothing to do about it. However, if it shown on a daily basis, aberrant data can be spotted. In a typical scenario, Mondays are always high. This is because most customers call on a Monday and high number of calls lead to high AHT. Missing this out would cause solutions not to be implemented. Maybe more people are needed every Monday. Or maybe, opening operations every weekend could help alleviate the volume on a Monday. Maybe there is something else that can be done. But essentially missing it out could hurt the decision and the business in the long run.
Data-driven decisions is not easy
Using data to make decisions is not that easy as it seems. Other articles and training around it may look and may sound simple but it is not. The work of getting, processing, and manipulating the data to create a story takes time and effort. Additionally, the knowledge is important to ensure the correct data is used to see the current state and create analytics around it.
Being aware of the dangers of using data to make decisions helps to better analyze it and design approaches that would solve any challenges that are presented. It could also help anticipate future issues and implement a plan to prevent them. Having a good understanding of basic math can help while knowledge in statistics can go a long way.
In a nutshell, data-driven decisions remove the “should-be” and “I think” way of thinking. Facts are laid out. Performance is measured. Reality is shown. Applying data before making decisions and creating a base from it is a more realistic approach. It removes the anecdotal comments that are affected by bias and aberrant events. By using data, there is a stronger foundation for making decisions compared to simply thinking about what it “should-be” or “I think”.
Leave a Reply