I have two core projects to share as a driver. One is to pull new and the other is to work part time.

Editor’s note: This article is from WeChat public account “Liuyanfeiyu” (ID: liufeinotes ), the author Liu Yanfei.

I have two core projects to share as a driver. One is to pull new and the other is to work part time.

First talk about the first one, pull new.

The driver’s new goal is very clear, that is, the conversion funnel for each step, and even the last driver will remain in the next month.

What’s interesting is that every step of the new funnel is not the same. You can simply divide the new one into: Channel->Registration->Activate->First Order->Second Month Retention.


All kinds of launching activities, as well as referral rewards, social fission, etc. In this funnel, the most important job of the product manager is not the exploration strategy, because there are a large number of market and operational students who are already studying.

Product managers are more flexible and rich in operational tools, so that market and operating students have enough room to try and adapt to local conditions. Different cities and different distribution channels can have different touches. Way.

This step of the funnel, the source of the demand, is basically internal.

Some experiences of driver growth

(Different channels are different. The incentive activities like drivers recommend drivers have always been the most stable and effective conversion channel for old users. It may not be the highest in terms of registration, but the retention is the best. The main reason is that These drivers are recommended by relatives and friends and have a higher willingness.)

Sign up:

Unlike the user registration of common C-side products, the driver’s registration complexity is very high. It is necessary to submit information on the three certificates (ID card, driving license, driver’s license) and vehicle information. At the same time, due to the high threshold, there are still some means to improve conversion.

The first thing to do in registration is to simplify the steps and “prevent” as much as possible.

The simplification step means that as long as you can reduce the number of steps, the conversion rate may increase every time you reduce it. For example, if the ID number of the ID card and the driver’s license are the same, then you can fill in less; or, to require face recognition, for travel safety,Instead of verifying your identity, you can ask for it before the first order, not when you register.

“Protection” means to make it as easy as possible for the driver to understand what to do. Most drivers are people over the age of 30. The level of education is average, and the simple content is difficult to understand. Therefore, it is necessary to use words and phrases instead of words.

Do some experience with driver growth

(Progress tracking is also a “defense” program that lets drivers know exactly what they should do)

These tools are aimed at users who have already decided to become a drip driver. This page, in fact, there are a large number of users who come in with a try-and-try attitude, so the registration aspect has to do the second thing: expected management.

What are the benefits to the platform, what conditions are needed for the driver, etc. There is no such thing in the past. (When the subsidy is high enough, the driver will flood into the platform without explanation.) In the past, it was obviously very inefficient by calling customer service or inquiring with friends around me.

Some experiences of driver growth

So on the registration page, Q&A supplemented with FAQs, and some car owners’ stories, that is, the real-life case adaptation of the drivers’ work on the platform, to give registered users clear expectations.


Activation refers to the conversions that are approved after the user submits the information. It seems to be a technical problem, but it is not. There are many situations in the middle, which require product managers to participate in the solution.

In the case of activation failure, data fraud, or non-compliance, of course, should be excluded, not allowed to enter the platform. However, there are two situations in which the user should enter the platform, that is, accidental injury, and the format is not standard.

Injury refers to the problem of the platform’s review and the accidental injury to upload the correct material. The audit includes two types, technical review (OCR) and manual audit. The former is accidentally injured, usually technically modified. The bottom plan is to transfer into the artificial channel; while the latter’s accidental injury is to be rectified, and more importantly, the bottom plan is Whether there is a complaint channel can be paid special attention. A user who is accidentally injured may be unsuccessful in repeated uploads, and the damage to the experience is also very large.

The format is not standard, which means that due to the complicated format requirements of the upload, a large number of users do not have problems, but the upload format does not match.Meet the standard. For example, there was a time when the photo format of the car and the car (confirming the ownership of the car) was very difficult to shoot. The requirements were: the phone was self-portrait, the face was clear, the vehicle was all drawn, the license plate was clear, and the door was opened. Most of the older drivers are difficult to take the correct photos according to the standard. The pass rate of this step has dropped below 30%, which is extremely serious. Later it was changed to a simpler way to achieve the same test objectives.

These two issues, product managers have to analyze a large number of badcase, in order to explore the problems behind.

First order:

When I was in the drip, the conversion rate of the long-term activation to the first order (received the first order) was only 60%-80%, which seemed to be very high, but the driver was successfully activated after various troubles. However, even one is not willing to pick up, there must be problems.

A study of activating a missed driver found that the driver’s anxiety was much more severe than we thought. This is the conclusion of one of the surveys:

Some experience of driver growth

Different solutions for different anxiety issues.

The experiment has a good effect, it is aimed at not knowing how to take orders and worry about unfamiliar routes. We provide the function of regional order, that is, you can set the destination in a certain area, so you can set a familiar area, and you can control the short order. The process of trying an order in a familiar route.

In the first single funnel, a big problem is that the conversion from “getting out” to “ordering” is not high. Getting out of the car means starting to accept the order, and we have counted that most of the new drivers have low patience with the order, usually turned off in a few minutes after the car is out, and there is no order.

This problem should also be managed as expected, so that the driver realizes that most people receive orders after n minutes; or let the driver see the situation of nearby orders, exaggerate, and even make a fight. A lot like that, “The nearby Master LiReceived an order, “The nearby Master Liu received an order.” Of course, I was worried that this exaggerated plan would create anxiety, but it would not be embarrassed.

Second month retention:

The new driver who lost a few orders is also anxious. The survey found that the cause of anxiety that “income does not meet expectations” topped the list.

If the income is reasonable but lower than the driver’s expectations, then the loss is normal. But usually the income of the new driver is not reasonable. There are two reasons: First, the new driver’s order handling skills are slightly worse, many operations are unfamiliar, which may result in fewer orders; second, the new driver is not familiar with the order. The resulting divisions have more contradictions, which will seriously affect the service points, while the service points directly affect the weight of the singles, which will lead to fewer orders.

So, new drivers usually have an increase in revenue after they have been on the platform for a while. This is another issue of expected management.

Solving problems can be done in combination with three options. The first is to give the new driver a clear income expectation. You can promote and improve the skills of the new driver through the promotion and promotion of the skills. Second, the protection of the new driver will be protected. The new driver made the mistakes to be excusable, and the deduction of the previous n orders was suspended; thirdly, the new drivers were sent to the position, because even if the new drivers did not deduct points, the default service points were much lower than the old drivers. Therefore, it is possible to slightly tilt the new driver who has not received the single ticket for a long time.

In the previous survey, we also found that the new drivers are struggling with the rules of the platform, and there are many steps in the order process. It is very difficult to learn and will be particularly anxious. So we made a relatively simple task system, I hope that the driver will come to the platform, what needs to be done, at a glance, and step by step, the acceptance will be much higher.

Some experiences of driver growth

In combination with these product solutions, the conversion rate of each funnel link has a certain degree of improvement from a few points to a dozen points. It is effective to put the number of users on the platform tens of thousands of levels per day.

The above is the first time I have done a relatively complete growth project, and I can benefit from a lot of experience and strategy. And you can also appreciate that this growth is the growth of products and technologies that I mentioned, rather than the growth of purely operational strategies (Talk about growth and growth methodology.

Second, is a part-time job of a driver, or a larger part, or a part-time project.

The problem of capacity resilience has always been a major problem that plagues the platform. The demand of passengers is peaked in time. The supply of full-time drivers is relatively stable. This contradiction must be supplemented by a more flexible transportation ecosystem, and the best elastic capacity is of course a part-time driver. (The same problem exists in the field of takeaway distribution, and many platforms are exploring more ways of income for free riders.)

Some experiences of driver growth

(Green is supply, blue is demand. If the capacity is mostly full-time, there will be a mismatch between supply and demand.)

The platform was not friendly enough for part-time capacity in the past. There are three points. First, it is the same as the problem faced by the new driver. It seriously affects the service points of the dispatched orders. The full-time drivers will obviously be higher, so the difficulty of part-time orders will be greatly improved. Second, the part-time drivers have various orders. Special requirements (time, distance, area, etc.) result in a higher probability of canceling the order, which makes it easier to deduct service points. Third, in the operating strategy of the platform, a considerable part of the full-time driver’s income is subsidies and rewards, while part-time drivers are not. Possible conditions for rewards.

Analysis, the third problem is the operational strategy of platform revenue distribution, which is not dominated by product managers; the first and second issues are similar, all of which are to provide a good part-time driver order experience, let the driver Willing to stay. In most cases, it is not the driver who feels that the income is low or the experience is poor, but the driver feels that the order is not received and will be lost.

This way, it is the most important thing to provide the order function for different workplaces.

The whole workplace is relatively consistent, for example, they work 8-10 hours a day, all have similar user portraits; and the workplace is very diverse:

  • Small business owners, 4 hours a day to pick up orders

  • Ordinary blue-collar workers can take orders for a long time on weekends

  • Programmer, commuting to work every dayCan receive orders (similar to the ride)

There are different requirements for the time and space attributes of the order. 4 hours of order time, usually do not want to send too far, and more importantly, as close to the end as possible, it is near the end; on weekends, it is more like a full workplace; The requirements for the start and end time are much stricter.

Different time and space attributes, and a hypothesis, is the relationship with income expectations.

Under the road scene, the driver is the psychology of fuel-saving, can receive the best, can not get the road, but the income is expected to be low; while doing business, occasionally can come out to pick up the order, then compare Other blue-collar jobs, income requirements will be higher, but still lower than full-time drivers.

Overall, there is an inverse relationship between degrees of freedom and income expectations:

Some experiences of driver growth

The hypothesis of this hypothesis is that it is possible to achieve a more complete ecology through the regulation of income + part-time functions, without having to distinguish between full-time drivers, and this is unlikely in practice.

What do you mean? For example, any driver user can use the function of qualifying the destination area at any time, but at the expense of a portion of the revenue (through pricing and subsidy strategies). In this way, part-time drivers will feel no problem, and full-time drivers will not use this feature to “pick”. At the same time, it will not restrict the use of full-time drivers. For example, the scene he may encounter is “I have to do things in the vicinity after an hour, I can’t go far”, then when he uses the function, the role has changed, that is, the transfer is over. Income, in exchange for a higher degree of freedom.

In an ideal situation, the regulation of income + part-time function can make the transportation ecology more healthy and the supply and demand matching more effective. This is why I am very concerned about this project.

In turn, the low-cost capacity provided by part-time capacity can also drive passenger demand in some cities, that is, “special cars” that have been on the line for some time in some cities last year. I won’t talk anymore here.

This project has been in progress for three or four months and has had good results in some experimental cities. It is a pity that later, due to the strict requirements of compliance, there is no room for part-time capacity, and it is regrettable to stop.

In addition to this project, there are many projects and scattered needs, like the driver’s personal homepage + honor system, like driver management and organization, and so on. I don’t talk about it either.

When I was in the direction of the driver, I really understood this sentence: the product manager should have a comprehensive and in-depth understanding of the user.

On the one hand, it comes from the fact that I have been a driver product for more than a year and have experienced hundreds of large and small demands.On the other hand, it also benefited from the excellent lineup of the Express product team. Under the guidance of Teacher Yu, the team worked closely with the brother product team.

About the understanding of the driver’s direction, I have written an article, I have not read it, or strongly recommend reading it: four key issues driver products.

I hope I can help you.