Data, data, data. Every organization today, is collecting data in one way or another. Due to the declining cost of data storage, data collection has become an obsession for most organizations. This data can come from employees, customers, suppliers, governments and many other sources. The basic premise behind collecting all this data is that it can be used to make informed decisions. But is it?
Informed decision-making should be based on sound data which requires data to be collected in a way that does not portray a false landscape. In other words, if the way data collection methodology is incorrect then the decisions made on that data will also be incorrect. Whenever we do data collection, there are a couple of things we should consider:
- Is the data coming from a primary source or a secondary source?
- Is the data coming from an individual or an organization?
- Is the data coming from conducting a survey?
- Is the data coming from conducting a study?
- Is the data coming from ongoing business activities?
Sometimes comprehensive data collection is time-consuming, costly, cumbersome and impractical. Considering these restrictions, we have to collect sample data and have to be cognizant that if this data can or can’t be generalized for decision-making. The wrong generalization of data from a small data sample can result in errors that might not be evident to the people who are making decisions on this data.
Let’s assume that you have been given the task of collecting data that can help the organization in Business-IT Alignment. For this, you conduct a survey in your organization to get a feel of what is going on. Your goal is to collect all this survey data, make sense of it and present it to the executives so they can make decisions. Here are the steps you take:
Step 1: Create a survey to collect data
Step 2: Reach out to relevant respondents
Step 3: Understand what the data is saying
On the surface, the above steps sound good. But here are the problems with each of them.
In step 1, when you are creating the survey, you can run into issues when:
- You ask leading questions that direct the survey into a certain direction
- The tone and mannerism of the survey/surveyor can make respondents uncomfortable
- A standard question across various respondents can be easily compared but based on the context the answer may differ drastically
In step 2, who you think the respondents are can affect the survey when:
- You only ask a subset of the respondents but you are unaware it was the wrong subset
- Respondents provide no answers
- Respondents don’t have access to the survey
- Respondents provide false information
In step 3, your tallying and interpreting the data can have issues when:
- Your personal biases (we all have them) influence your interpretation
- Others’ personal biases influence your interpretation
As you can see from the above errors, you have to be careful in data collection so it reveals the truth rather than a skewed version of a hypothetical scenario. The basics start by asking the following questions even before you start creating the survey for data collection:
| Today | Tomorrow |
Who is going to respond? IT (Help Desk, Software Developers, Management, Database Developers, Network Support, Cybersecurity, etc.) vs. Business (Accounting, Sales & Marketing, Finance, HR, Operations, Management, Customer Service, etc.) | Who should respond? IT (Help Desk, Software Developers, Management, Database Developers, Network Support, Cybersecurity, etc.) vs. Business (Accounting, Sales & Marketing, Finance, HR, Operations, Management, Customer Service, etc.) |
What areas are covered? Person-to-person interaction vs. Organization-wide capabilities | What areas should be covered? Person-to-person interaction vs. Organization-wide capabilities |
Where do you think is the organizational misalignment? Offline vs. Online | Where should be the organizational misalignment? Offline vs. Online |
| When did organizational-misalignment appear and reported? | When should organizational-misalignment be identified and reported? |
| Why is the data being collected? | Why should the data be collected? |
Welcome to Khan's Blog 