When analyzing data to reveal trends, the more data, the better. For any type of statistical analysis, the larger the sample size, the more accurate the results. It is difficult to see future trends by just tracking a company’s sales data for one week. Three months is better, and six months is even better. Even if you are not sure what you are looking for, make sure the information contained in the data you collect is as detailed and accurate as possible. Try to figure out how to get the best data you need, and then start collecting. If you don’t have data, you can’t analyze it.
Collecting data is the link of how to record data. There are two principles that need to be emphasized in this link, namely, full volume rather than sampling, and multi-dimensional rather than single-dimensional. Today’s technological revolution and data analysis 2.0 are mainly reflected in these two levels.
1) Full volume rather than sampling
Due to the constraints of system analysis speed and data export speed, data analysts in companies that are not supported by big data systems are rarely able to collect and analyze data in full volume. But this will no longer be a problem in the future.
2) Multi-dimensional rather than single-dimensional
On the other hand, it lies in the dimension of data, that is, to achieve comprehensive refinement of SWIH for customer behavior, comprehensively record the time, place, person, reason, what was done, and how it was done during the interaction process, and refine each section. Time can be subdivided by start time, end time, interruption time, cycle interval time, etc.:
Location can be subdivided by geographical features such as cities, communities, climate, channels, etc.: People can be subdivided by multi-channel registration accounts, members, salaries, personal growth stages, etc.: Reasons can be subdivided by hobbies, major life events, demand levels, etc.: Things can be subdivided by themes, steps, quality, efficiency, etc. Through these subdivided dimensions, the diversity of analysis is increased, so as to discover patterns.
Purposeful data collection is the basis for ensuring the effectiveness of the data analysis process. It is necessary to plan the content, channels and methods of data collection, with the following main considerations:
①Convert the identified data analysis needs into more specific requirements. For example, when evaluating suppliers, the data that needs to be collected may include their collection capabilities, measurement system uncertainty and other related data:
②Clarify who collects data, when and where, and through what channels and methods:
③The record sheet should be easy to use:
④Take effective measures to prevent data loss and false data from interfering with the system.