Hello everyone, I am the strategy product manager of Xia bluff.
Today, let me talk about strategy productive thinking?
Why do you want to talk about this topic?
causes it because I recently discovered a phenomenon that when my team’s friends output strategy solutions, they are easily trapped in a vicious circle: the optimization strategy directly sees whether there are loopholes in the logic, and when optimizing the logic, it is complicated, that is correct, and awesome, that is, that is, the effect is good.
is strange.
This idea has deviated from the way of thinking of being a product manager.
If you ask everyone during the interview, how to make a product?
I believe that a relatively complete process will emerge in every product manager’s mind: planning, research, analysis, plan, implementation, and finally iteration online. When making products, you need to look at the business, users, scenarios, etc.
However, when they were asked to do strategies, they no longer had such an idea and could not do it.
In fact, the strategy has no fixed form in terms of form, but the essence of the strategy is actually a solution to the problem. However, compared with C and B-side products, the strategy is more often based on a large amount of data to output the solution, and in the end they are all to solve business problems and bring business indicator benefits .
Therefore, also requires productive thinking . The implementation of a strategy will not be short of product processes before you make it. The following is a strategy flowchart I drew before. It seems that it is basically applicable:
What is strategy product thinking? Simply put, making strategies also requires productive ideas to be implemented. A strategy should be treated as a complete product requirement, even if it is a small optimization.
How to shape this product thinking? In my opinion, you can deliberately practice the following points in work and study:
1. Data pre-analysis must be done for strategies
Usually, we mainly focus on the three directions of user, industry and business, but what is more important for strategy needs is that data research is required. Focus on the following aspects:
1. Data status and expected improvement value. Understand the current status of indicators, clarify the ROI after the strategy is launched, and formulate core indicators for subsequent iteration basis. At the same time, the expected increase in this strategy demand is determined by affecting the traffic size and past project experience. At the same time, I have also mentioned the relationship between expected improvement values and AB tests many times in the training camp.
2. Judging the coverage and impact of strategy requirements. Compared with other C and B requirements, it is difficult to directly locate specific cases for certain specific scenarios. For example, when optimizing a sorting strategy, it is difficult for you to judge what the sorting effect will be after the strategy is launched, but you need to predict the scope of the impact of the strategy and conduct early testing. For example, GSB.
3. Data can support the implementation of the strategy. This is a very easy-to-neglect link. There are many strategic products that students can recommend the process when they think of solutions: write a plan, do review, and find that data is missing here when they are developing, which causes the delay of demand.
In the implementation of strategy products, we generally focus on whether the user, item, and event-related data are available, and can be analyzed and processed. In the specific implementation process, the strategy is actually an implementation of data engineering logic, and the most important means to ensure data quality is high-quality data collection capabilities.
2. Strategy needs to pay attention to business background
Strategy is a solution to solve a certain problem in a certain business scenario. Therefore, the output strategy should be based on solving business problems. This is the most direct and important link.
However, the reality is that when outputting strategy solutions, many strategy product managers pay more attention to what algorithm is and what logic is added this time, rather than what problems are solved and how much can be solved.
In many people's subconscious mind, they feel that strategies without algorithms will not have good results. Therefore, the output of the strategy plan depends on whether complex logic is used and awesome models as necessary conditions for whether the effect can be produced.
However, a good strategy does not necessarily mean a high-end algorithm. The same sentence: Putting aside business and talking about strategies is all about no water. A good strategy must be based on the output after clearly defining the current business problems and user expectations.
Before making a strategy, focus on the following aspects:
1. What are the business problems solved by the strategy? Including business indicators, business needs, etc.
2. What are the user expectations in the current scenario? That is, what kind of results the user wants to see in this scenario
3. The results produced by the current solution can meet most of the situations.
3. Strategy needs to be focused on the impact
. In many companies, especially large companies, the strategy department usually exists in the form of a middle platform, which means that strategies usually support multiple businesses, especially similar to search recommendations, which must be a scenario of comprehensive traffic distribution.
So, at this time, whenever a strategy is launched, it usually needs to consider its impact. Generally, it needs to pay attention to the following:
1. Definition of business scope. Before making a strategy, we need to clarify the scope of the business effectiveness of this strategy. Many times, our strategies are effective against a certain or several strategies. If there is an error in the scope definition, it will cause abnormal situations to occur after the policy is launched.
2. The impact of the association link. Similar to search, recommendation system involves many links, and they are all interlocking. Therefore, when you optimize the strategy of one of the links, you need to consider the impact on other links at the same time. For example, if you need to add some synonyms in the intention recognition process, you need to consider whether the added synonyms will have an impact on other links that are using the intention interface; at the same time, whether the newly added sorting features are only suitable for the current scenario, but will not be used in other scenarios.
3. Strategic business objective estimate. How to estimate how much benefit a strategy can bring to current business indicators. Can you expand it according to the following ideas?
· The traffic range affected by the strategy is determined by
· The current core indicators of the strategy are now determined by
· The possible benefits of
or above are points that need to be considered in the process of making a strategy. The point I call it product-based thinking is: when you do strategies, don’t narrowly understand it as logical optimization, but just like you are making a complete product, so your strategy may be a relatively "good" strategy.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-