Introduction: product value = user scale user value *
Introduction: the scale has been an important factor in user measure product value, from 0-1 in the product, the entire cycle from 1-100 subscriber growth has always occupied a large proportion of the cost of inputs, like electricity supplier, O2O, social class, etc. Almost all product development can not avoid the "growth" word.
A common user growth model:
A new, retained reflux three parts, in fact, at the moment of the first user to log product, retained the work began , the user experience every step of the product flow, see each ad operations, received each push message, etc., each node determines the conversion of the next .
Referred to herein, "retained" in AARRR 4 comprising letters, i.e. Activation (active), Retention (retained), Referral (conversion), Revenue (spread).
Dividend flow disappears under Internet winter, we are talking about saving, but I think the systemic problem of user retention is every operator should think about patience.
Here we talk about growth hacking, talk about how well the user retained, then it should be how to do? This reference to the methodology of operational chiefs knife sister doris, combined with his experience, will retain the user's work is divided into three parts, new user retention, retain old customers, to explore the boundaries of retention strategies.
New user retention, focusing on the optimization experience, allowing users to quickly feel the product's core values.
We often experience will give users a set route, but the fact that users do not have to go, if not able to match the value of the product for the first time on the user's needs, it is likely that users will be lost, so do the new user's behavior to ensure that the user guide is retained premise, the premise is a product to carry out large-scale investment in the market.
How do the new user guide? We three steps, as shown below:
1. The first step: to draw a new map user experience , that is based on user roles to the timeline axis, listed in the product user experience of key nodes and feelings (Note: If the product is more complex function can be set to experience multiple paths)
2. Step two: return to reality , the reality of the path the user to use the product and user experience to compare the map
3. The third step: comparison of differences in expectations and reality , the reality of conversion product meets expectations or operational goals? We will result differences that may arise are divided into two categories:
3.1: user experience the core functionality intact
3.2: user experience is not complete core functionality, there are usually two cases:
a. After the user enters the home, did not experience any function directly out of the
b. The core function of the user experience process interruption
For a case, I will first analyze the source of the user, the new user quality judgment, some users are not our target audience, no matter what kind of operations are difficult to take to generate retained, these users we do not have to spend energy, but do pull new channels the strategy optimization;
Then analyze whether the product is the value of education, or are a new user guide does not do a good job. For the educational value of the product launch of the new time in pulling begins, content delivery whether the user precise hit pain points? Whether the expression of the value of the product? Meets user expectations?
For the new user experience to guide the path, there are multiple forms. Can be carried out step by step guide through the mask, the system can also create tasks set new tasks to complete to give some incentives, it can also be issued by novice gift packs, of course, use the best gift is to return to the core values of the product.
In selecting the form guide at the same time, we have to stand on the user's perspective, taking into account the completion rate of the user guide path, it is recommended to set the progress bar to the user expectations.
For b the situation, I will do according to user behavior path conversion funnel analysis model, step by doing research focuses on the behavior of interruption. Here the user's guide new users through education and product value of psychological expectations have been matched with the product, if you do not complete the core out of the path, it may be a link with the expectations do not match.
The reason is that new users are typically the first time to experience the core functionality of the interference of other factors , factor here may be a phased operational activities, possibly some conventional means, such as operating the old with the new, red envelopes and other divisions.
Solve this problem, we need to do a basic user stratified, hierarchical new users and returning users, the new platform for user touch of the highest priority should be the product do the core values of the path of experience, operational activities should be based on user points gradation layer serving.
Comparative analysis of differences in expectations and reality, we will have user behavior, demands have a better understanding, on the optimization of product and operational strategies based on business scenarios to be carried out. In addition, even some of the key actions carried out AB test, trial and continuous optimization.
Old user retention, focusing on refined operation, namely the refinement of operational strategies and refinement of data monitoring.
Each user of the content and depth of product requirements are different, how to distinguish between user demands for products? And how to determine the value of mining the user of the product? How to improve the proportion of high-value users and reduce the proportion of potential users?
A user clustering is that we do fine operating base, from top to bottom can be divided into nine steps, as follows:
1. clear operational goals, we hope to operate hierarchical what kind of effect, enhance MAU? Enhance GMV? Operational goals directly determine the dimensions of the grouping.
The target list partition factor, the operator has the goal of diversity, to cover a variety of goals, we summarize broadly divided into four categories:
2.1 basis of property factor - sex, young, region, occupation, income, equipment type, etc.
2.2 Properties demand factor - what the user wants to get from the product
2.3 Behavior property factor - the user's activity, product breadth and depth of experience to complete the critical behavior of the number of times
2.4 Properties Value Factor - user needs are met to schedule i.e. part which records the progress of the conversion funnel
- >> 2.1 and 2.2 can be combined to make user portrait
- >> 2.3 and 2.4 can be combined to make the user life cycle
3. Locate the corresponding index attribute factors and indicators dismantling. The following figure shows an activity index for reference: it is the company's business involving a code of secondary indicators
4. A combination of factors determine the user grouping, to RFM stratified example, we need to find the value of the property factor of three indicators:
R (Recency) a latest time of consumption, the F (Frequency) frequency and period of consumption, the amount of consumption in the M (Monetary) period;
User-type layered, for example, if it is to travel products, we can be divided according to the number of users browsing content and the user just need high-frequency type, quality low frequency etc.
The extracted data generation population distribution, determined according to the threshold value index distribution of the data.
To user activity metrics, for example, we may live into high users (MAU density ≥60%), the live (monthly active density ≥20%), low living (monthly active density> 0), silence (active loss 30-90 days), loss (active loss of> 90 days)
6. grouping of users for different operating strategy, where the logic core functionality and new user experience is somewhat similar path,
Namely: I look for blind -> find a cause -> Import Strategy
First, comparing different periods of the number of users clustering distribution, to find blind operation inconsistent with expectations. Taking a cargo logistics vehicle matching platform as an example, I recently discovered the loss of the driver of the number of users is greater than expected, so we were broken down by region, the main source of loss is found in southern China user, we again found by analyzing the behavior of the loss before churn , using the platform's main purpose is to find the user short supply, but the number of recent short supply due to reduced platform can not meet user needs, thus leading to the loss.
After analyzing the causes and take countermeasures. To address this situation, we took three measures:
1. For reducing the number of short supply, we have created a short supply experience Month, extract a list of shippers often publish short supply, the owner released more matching short supply through incentives to stimulate, to take some effective sources of supply in order to ensure the quality of conditions determination; while guiding line at the newly opened efforts to strengthen the team owner in this short period;
2. For already lost and about the loss of the driver, was to increase the activity for the purpose of login prize activities, because users only have to log in to find short supply changed much;
3. In order to return the number of drivers feel short supply upgrade, we had a push (push + SMS) message-based automation short supply of LBS, the user every city who will tell the driver: "somewhere - somewhere there is Article XX supply, how many meters the car needs, large price advantages, speed to inquire. "
In the implementation of the above three strategies, in addition to action taken respectively from the owner and driver two dimensions, but also the development of priority activities of the timing and intensity of contact, in order to achieve the most cost-effective costs.
7. operating strategy as scheduled landing. In the implementation of the policy, it is recommended to do according to the top-level logic indicators dismantling, clear priorities, establishing execution rhythm, clear time, each responsible for a breakdown of the matter, refer to the following chart:
8. Monitor the data grouping. Grouping user data for monitoring lateral point of view of each of the total amount of user grouping number / proportion, e.g. March and April, the proportion of high value to the whole amount of the user; longitudinal viewed Circulation of grouping different users, for example, March how many users of high-value high-value flows to non-users in April.
9. tracking results, continue iteration. After the new operating strategy on the line, we have to continuously monitor changes in the core index, conversion and retention user base continues to verify the validity of the adjustment strategy.
Develop more operating strategy, the basic work is completed the user retained the same time, we have to think if the current business scenario to develop other operational strategies.
As long as deep into the business scene, we can step by step, in-depth understanding of user needs change in order to understand the deep-seated reasons behind user behavior.
About develop other retention strategies, which can be divided into two types:
1. Meet the needs of users in-depth, namely the core business -> expand their business -> epitaxial business.
Taking a driver to find the goods platform, for example, first driver to find goods to meet the needs of the core business, that is to achieve a breakthrough in user retention, gradually carry out the driver refueling platform flow to a certain size, ETC cards, loans to expand their business, and then gradually to carry out the investigation violation, driver's license upgrades, inspection and other services.
2. Increase the cost of the loss of users, an increase in precipitation user platform (chain relationships, money, honor, etc.)
To grow (Member) system, for example, in terms of the platform can enhance user stickiness, self drive user growth; for the user to be able to get along with or enhance the level of the corresponding rights and interests, increase the cost of the loss.
Summary: a thousand eyes Garrett Ram had ten thousand, one hundred products have one hundred operating strategy, for their own is the best. And we are seeing things from the methodology of constantly identify, test optimization constantly find the most suitable method.
Author: summer bean soup