Reading: Data Driven, by J. Dearborn

The author aims to draw out some insights from her past experience in implementing data analytics to improve sales performance in her organisation. I particularly appreciate the style of using a fictional story, of a company struggling to improve their sales and their eventual adoption of data analytics, allowing most readers to empathise with the protagonist, as well as relate the situations to the readers’ workplace. Here are some of the key learning points that I have picked up:


Sales as the Guinea Pig – we should all consider this when pushing for changes in our own organisation. Dearborn stated that about 80% of any companies is related to sales in some ways, and improvements in sales are easy to measure, and therefore more potential to gather support for the project following any initial successes.

Company’s Responsibility – I like the author’s thinking, that a company is responsible for preparing their staff with sufficient skills to manage technology changes in the workplace. Without any focus on proper training, layoffs are inevitable should the company continue to grow.

Chapter 1: Playing the Blame Game

The game that we are all too familiar with. When confronted with finding root cause of declining sales, all team leaders, from sales, per-sales, marketing, training, recruitment, product development, and operations, start pointing fingers at each other and making excuses for their situation.

Without making use of data, these leaders will end up setting unrealistic goals, solve the wrong problems (as root causes are typically identified based on “gut feel”), and end up measuring the wrong performance indicator.

I appreciate Dearborn pointing out the common bad habit of teams measuring efficiency instead of effectiveness as a justification of their performance. For example, an IT team is extremely focused on ensuring and reporting system up time, however they may be missing the fact that this system is not critical and does not contribute to the sales of other business units at all. When confronted with declining sales, the IT team never fails to pull out the system up time to justify their value and contribution, and starts shifting the blame to other teams. This is one of the factors that perpetuates the “silo” mentality that plagues many organisations.

Chapter 2: Pulling Back the Curtain

An important warning for every company venturing into data analytics. The landscape evolves constantly and there is no standard analytics taxonomy, so leaders who are embracing data analytics in their company must be comfortable with this mess. If you require every step of the journey to be properly structured and defined by some kind of “industry best practice”, then this is probably not for you.

Also, there are a lot of hype in the market, and some sellers pushing products and solutions to your face may be making unsubstantiated claims on their analytics capability. They may not be analytics at all, just some re-packaging and re-marketing of traditional solutions. So do yourself a favour, gather more knowledge before making any investments.

Chapter 3: Changing Mindsets

Start Small – If you start big and fail, everyone will lose their trust in data analytics, and be rest assured future projects will never take off, ever.

Company Leadership – The project must be headed by the management, preferably someone interested in fact-based decision making. Successful transformation of the company depends on changes in processes, skills, culture, not just the implementation of a solution.

Internal Capabilities – I feel that this is a crucial point that many top management failed to appreciate. No consultants or solution providers will be able to transition your company, without you first establishing internal analytics capabilities. Your internal team will understand the nuances particular to your business and are driven to use analytics to make changes with more motivation than any outsource parties. Of course, assemble the team with staff of the right skills and mindset is the key to success.

Other pointers from Dearborn are to not rush to outsource your analytics capabilities, and do not expect analytics solution to work like a silver bullet for your problems.

Chapter 4: Finding the Keys

Always start off by considering all the possible variables that affects the performance output that you are trying to improve. This is a brainstorming exercise, and a lot of the factors maybe eliminated eventually as they are not as significant as others.

Next, try to locate the data. It may be in multiple systems within the organisation, or it may even come from external sources, like firmographics. Not all data that corresponds to the variables may be available, and some data cleansing will need to be performed.

Chapter 5: Descriptive Analytics

Descriptive analytics simply use software tools to present the data in a meaningful manner. I like that the author warned the readers not to be too particular with the terminology, and to accept that all the analysis in the subsequent chapters may make use of the exact same tools, the difference is merely the way we use data and the objective we are trying to achieve.

Chapter 6: Diagnostic Analytics

The story described the use of a machine learning model trained to identify high performers and poor performers through a large set of input data. Once the model is able to predict performance with high confidence, the configuration within the model will point out which input variables has the greatest impact on the outcome, and those are likely the areas that the company wants to focus on.

Once key variables had been identified, the performance of each staff in the respective variables can also be identified, which highlights the areas for improvement tailored to each staff.

Chapter 7: Predictive Analytics

With limited resources, the company will have to prioritise which staff and customers deserves more attention. Using the same machine learning model in the previous chapter, the company will be able to identify deals which are more likely to close than others, , and staff who are more likely to under perform. This helps the company to decide on resource allocation, just in time to bring up overall performance.

Chapter 8: Prescriptive Analytics

The author admits that prescriptive analytics is an extension of all the above analytics, and it really isn’t a tool or a method. In my opinion, it is a systematic way of using the results derived from the above analytics and take action.

Descriptive analytics had helped to provide the mechanism to look at performance objectively in greater details. Diagnostic analytics had identified the key areas of improvement that matter more than others. Predictive analytics then highlights customers and staff that the company can work on for quick wins with limited resources. Finally, all it takes is to communicate with the staff that are most likely to improve with the least amount of resources, work on key areas tailored for the staff, and focus on customers that are most likely to close the deals.

Chapter 9: Celebrating Success

[Spoiler Alert] Of course the story has a happy ending. The company saw encouraging sales improvements and adoption of data analytics.

In summary, this book teaches an action model that is applicable to most companies:

  1. Identify key factors that affects performance, and collect data for those factors.
  2. Diagnose which factors have the most impact on results.
  3. Predict the performance and identify quick wins.
  4. Act on the areas highlighted, and continuously measure and improve on the action plan.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s