When we do data analysis, we have not false. Data analysis is difficult to get closed loop, this time, how can we do to achieve data analysis results? The author shared his own actual floor experience, let’s take a look.
How does data analysis results?
I analyzed the results I think of valuable data analysis, but I don’t know how to land, especially in large enterprises, I want to implement my results, it is very difficult, I do a valuable data analysis, Over the end, do you have any means?
As a data analyst with dead bumbles from all year round, I have just retired that I will have a problem that the owner said: Why can’t you always have a data analysis project? The following is the land experience after I experience multiple projects.
First, what is a valuable data analysis result?
1. Analyze logic closer to business scene
First asked, this demand is the business proposed, or actively issued?
Because the title does not introduce the relationship between the department and the business, in my personal experience, there is indeed a lot of data analysis needs is not the case of the business side, that is, the analysis is likely to be away from the actual business.
If the business is proposed, it is also necessary to avoid a point: don’t first, it is the main reason for the business.
When analyys analyzes demand, analyzes to critically, use the validity reliability think to determine the needs, and reduce the scene in the bottom of the planing. This process should be dominated by a professional data analyst.
This part is actually ensuring that the direction is correct, and the data analysis logic of the support conclusions is recognized.
2. Analysis conclusions and recommendations are combined with specific business
Unless there is a person who has better data thinking and has been used to and this, the data analysis may be a little result for most business, but it is still difficult to imagine how to apply landing in the specific business.
To analyze the conclusions, you need to reach two points:
(1) In-depth drilling analysis capabilities
This is actually the data analyst itself, facing business problems, can continue to drop the analysis to find the finest granular solution and business growth point.
For example, the example in my actual business: Business Company should force the K products next year, sales increase by 100%, how to do?
First, because the docking is the user’s business colleague, the KPI indicator of its assessment is the proportion of reused populations and the proportion of the amount, that is, to operate through the finest population, enhance the replay rate.
Second, the RFM model is used to analyze the existing population, and it is found that the proportion of important people will fall year-on-year, and the price of guests has a certain degree.
Third, for the important people discover, DuPont analysis – indicator dismantling analysis, found: If the decline in the user’s customer unit price is mainly caused by the average tag price of the store, it may be due to the change of sales of store products Therefore, it is further analyzed that the population with the product is discovered, and there is no significant change in sales structure, and the problem is that the problem is in a high price A and B type transaction price decline.
Fourth, in summary, for the growth of K products next year, users operate this piece of operations to operate the RFM population, for the importance of important value users, pay attention to the price strategy of A, B high-cost price strategy
After meeting the analysis of the business company, the product, operations can be treated in response to this specific situation, and actively propose their views, action plans and further analytical needs.
2. Thinking Ability of “Point Line Face”
The “point line facody” said that “Liang Ning’s Product Thinking Course” said that in terms of the same business problem, you need to know more, consider multiple modules such as product, operation, planning, supply chain. , Construct a thinking ability of three-dimensional business models, know abstract composition facing a specific business problem.
Another example in my actual business, before the traceability analysis of Jingdong Festival holiday reused crowd: such as Christmas has a total of 100 old customers back to buy, traceability is what is the last consumption of these vans, the purpose is For the presence of a preceding marketing of the day, the recruption is selected.
When meeting with the business company, the business boss can have a specific policy direction based on this analysis:
1 Combined with Jingdong channel characteristics: Male digital product enthusiasts, knowing that emotional festivals such as Valentine’s Day, women’s festival and 520 festival buy our first-jewelery products member needs are probably gift
2 If it is assumed to be established, these customers may be in these three festivals, which may be purchased in a row, so the operational colleagues can consider selling packages in three festivals.
3 Planning colleagues can be carried out in the packaging design and new development development of the gift scene; user operation colleagues can focus on the month and recent active population.
We all know that leadership is to control the overall strategic direction, that is, our specific business problem is actually the composition of a three-dimensional multivariate. So only the business model of “wire facures” is built, not to analyze a point, in order to match the boss, the analyst said that he can hear it.
Second, how do you apply landing?
1. Hook with business KPI, operation plan
Not every department recognizes refined management, digital decision, at least from action, not well implemented. I docked with the business company’s user departments, although the leaders have already done, but they will also work in operations, products, and planning departments.
Specifically, even if the leadership meet the business units for the results of the data analysis, it is basically the state of “all walks, each to find a mother” after the meeting. An analyst is needed to actively communicate.
The business is still in this way, our data-related staff is to clear how the data analysis project conclusions can help them do performance improvement. And this is to understand the process of KPI and operational plans, as long as it combines these content, naturally they can make them.
2. In the leadership, where is the interests of this matter?
Yes, the world of adults is interesting. Of course, directly promoting the business is the best interest point, but the real situation may be more complex:
Target Management – Smart Principles
If the business is inside the door, everyone is what they have. As long as the test scene is clear, based on the Samrt principle, the Gantt Chart and other tools, the test plan, estimated effect, ROI target, etc., the probability can Executed.
But if it is a cross-departmental project, such as analysts are independent data sectors, and the business is a sub-brand company, and “management”, flexibly use the direct superior “management tool”, build a group, email, etc., hurry At the analysis report interpretation meeting, the test plan is attached to the PPT, and will be put forward, and when the leaders of both parties are present, it is best to have clear future conclusions.
It is necessary to continuously promote the landing and business value of the analysis conclusion in the form of project management.