Online education product promotion practice

In this article, the author introduces how to implement data product promotion . Let’s take a look at it together.

1. Introduction

Not only data products, but all B-side products will face the problem of implementation effectiveness. The B-side products mentioned here include not only commercial B-side products of ToB enterprise services, but also B-side products built within the enterprise.

The implementation difficulty of self-built B-side products is relatively small because your users are your colleagues around you, but if you build a larger platform yourself, you also need an implementation team.

For companies providing enterprise services, maintaining a large implementation team is inevitable. Huawei may be the company that has done the best in implementation, having already deployed 5G base stations on Mount Everest. However, not all companies have Huawei's financial strength and organizational management efficiency.

A fellow operations guy once told me a joke: the boss of his previous company decided to buy a competitor of Sensors (I won’t mention the name), but a year later, none of the operations staff knew what the system login address was. Maybe only the boss himself was using it. From another perspective, in order to maximize the implementation effect, both parties A and B need to work together, and internal factors may still be the main factor.

The advantage of Sensors product functions is their flexibility. It is more difficult to get started than similar products such as growingio and Umeng. Dialectically speaking, this is both its advantage and disadvantage. In order to make up for the shortcomings, the Sensors implementation team has worked very hard. What I have seen is that the front-line analyst team is almost teaching the people who have reported the problems how to operate. However, there are too many operations staff, and you can't just go through all of them like reviewing a text.

After the data product is deployed, it is equivalent to giving you a fishing rod and a method of fishing, but it is hard to say whether you can catch fish with the fishing rod, because each person's cognition and practical ability are different.

So how can we enable users to integrate methods and tools and truly teach them how to fish? This article attempts to break the deadlock based on a summary of my own practical experience.

This article also continues the ideas of the previous article: analyze the problem - provide a solution model - practice iteration. This article will explain the problem analysis faced in the early stage of product implementation, implementation and promotion ideas, how to promote to core users, and establish a standardized usage mechanism.

2. Analysis of early use problems

1. Low enthusiasm for use

Although everyone talks about data-driven operations when reporting on their work, when it comes to implementation, most people are unwilling to make changes due to work inertia. Giving you a system and asking you to use it yourself actually increases the cost of work in the entry stage. It is not surprising to hear the phrase "from getting started to giving up".

In the early stage, those who used it voluntarily focused on teams such as advertising and website operations, and the number of people was less than 10. These teams only focus on the core indicators of their KPIs: number of business cards, number of registered users, PV, UV, etc. The use of analytical models focuses on event analysis, funnel analysis, and user attribute analysis.

2. Not understanding the basic functions

We conducted a survey on those who were very enthusiastic in the first full-staff training but did not actually use the product much later. We found that they did not have a good understanding of the basic functions, let alone a deep understanding. Some people have read the tutorials but found it difficult to get started; some people found some problems they did not understand during use but were too embarrassed to ask; of course, some people had too much transactional work to use it and had no time to do so.

In the end the result is the same, from getting started to giving up.

3. Difficult to integrate with business scenarios

The above analysis also shows that the operations personnel who consciously used it in the early stage also focused on several common analysis model functions. In addition, the analysis indicators were also focused on their own KPIs. There was insufficient analysis of the indicators of each link that affected the final KPI results, and there was a lack of ability to sort out and quantify business processes.

How to skillfully use system functions and apply them in business scenarios is indeed a challenging problem.

4. Different statistical calibers

The inconsistency in statistical caliber is mainly reflected in two aspects. The first is that the data viewed by business colleagues on Sensors Analysis does not match the internal business system, and there is a point tracking error (see the point tracking calibration section in the previous article).

However, the situation I want to talk about in this article is man-made, because some operations partners do not have a thorough understanding of atomic indicators and dimensions, which leads to inconsistent selection of atomic indicators or dimensions during queries in different systems.

Another situation is that different people have different definitions of the same indicator (with the same name) on the same system (all queried on Sensors Analysis).

3. Implementation and promotion ideas

After the above analysis, we have a clear idea of ​​the general problem, so how do we solve it? Let’s go directly to the solution, please see the figure below.

1. Operation Alpha

From the above solutions, we can see that Operation Alpha is at the core.

There is nothing mysterious about this action. It just has a cool name. Its essence is to cultivate core users. I will explain it in a chapter later. The training of core users is conducive to improving the usage atmosphere, exploring the company's training and assessment mechanism for the use of data systems, and incorporating it into the training system for new employees.

2. Promotion and iteration

After several rounds of core user training, if nothing unexpected happens, the company's data-driven operation atmosphere will improve, the number of core users will increase, and demand will naturally increase, which will promote the system to enter a positive iteration stage.

3. Promotion of governance

As the number of core users expands, the number of people with analyst role permissions will increase, and irregular usage will increase. It is necessary to govern the usage process to avoid disorder, so it is necessary to standardize key usage links.

4. Operation Alpha

1. Action Overview

The purpose of organizing the Alpha Action is to cultivate and expand the core user group, publicize and enhance the data-driven operation atmosphere, and establish a regularized training and assessment mechanism during multiple rounds of implementation and iteration of the action.

The specific implementation steps of Operation Alpha are shown in the figure below:

1) Form an action team

The initiation of an action requires the support of senior management. The empowerment of senior management not only clarifies its legitimacy, but also determines its influence and level of attention within the company.

The members include:

• Sponsor, product manager or operations manager;
• Business consultants are key personnel who are familiar with the company's operational business and provide business guidance during the operation;
• Technical support: Technical support here refers to people who are familiar with the system functions. They can be internal product managers. Of course, it would be better if you can invite analysts.
• Participants: The trainees are recommended to select young people with growth potential in the first few sessions, who have relatively strong ability to learn and accept new things.

2) Develop practical training courses

The formulation of the course plan needs to be combined with the company's business, and the theme should be found from the actual work content of the trainees. This can not only achieve a practical effect, but also not add extra workload to the trainees.

It is recommended to set two themes, from easy to difficult. Hold regular weekly meetings to review activities, summarize patterns and reflect on shortcomings. The recommended time is 3 to 5 weeks.

3) Review sharing & evaluation

This session is the final climax of the event. All employees will publicly review and summarize the action and invite all operators in the company to share cases, gains and experiences.

After the review and sharing, outstanding students were selected and awarded by company leaders, which was used for internal news publicity to further emphasize the concept of data-driven operations.

2. Even small actions need PDCA

Although Operation Alpha does not involve software development and is only a training activity, we should also look at it from a project management perspective and adopt the PDCA project management approach to reduce the risk of failure. After all, it is not easy to organize an event.

1) P (Plan)

It includes the determination of policies and objectives, as well as the formulation of activity plans.

2) D (Do) Execution

Based on the known information, design specific methods, plans and plan layouts; then based on the design and layout, carry out specific operations to realize the contents of the plan.

3) C (Check)

Not only should the final result be checked, but also checkpoints should be set during the process to detect problems in a timely manner. For example, during the implementation of Operation Alpha, each topic may last for about 2 weeks. After about a week, I will ask the trainees about their progress in the group and ask them to send mind maps to check whether their analysis ideas are correct.

4) A (Action) processing

Process the interim results, affirm the successful experiences and standardize them; the lessons of failure should also be summarized and paid attention to. Unresolved issues should be submitted to the next PDCA cycle for resolution.

The above four processes are not completed after one execution, but must be repeated over and over again in multiple actions. After one action is completed, some problems are solved and the unresolved problems are used in the next action, in this way, we can move forward in a step-by-step manner.

V. Standardized Usage Mechanism

As the number of core users expands, the number of people with analyst role permissions will increase, and irregular usage will increase. It is necessary to govern the usage process to avoid disorder, so it is necessary to standardize key usage links.

1. User stratification and clear division of responsibilities and rights

Personnel participating in Operation Alpha training will be assigned to analyst roles with the following responsibilities:

• The analyst role needs to be designated by a dedicated person in the department;
• Responsible for the use of the Sensors system within the department (report preparation, specific scenario analysis);
• Training on the use of internal departmental systems;
• Establishment of internal data indicator system of the department;
• Summarize feedback requirements.

2. Demand feedback process

• As shown in the flowchart, each department needs to designate an analyst role;
• The data needs generated within the department are first solved by analysts relying on the Sensors system;
• Requirements that cannot be resolved by analysts need to be fed back to the data product manager in the format of requirement cards.

3. Account management mechanism

1) Account application mechanism. Since it involves specific roles and personnel within the company, it is not convenient to describe in detail, so I will give a brief explanation here. Analyst accounts are allocated and the number of vacancies depends on the headcount of the department. Those who apply for special data analyst accounts (authorities) need to be approved by the department leader (or VP).

2) Mechanism for changing authority and shutting down authority for employees who leave the company.

3) Account regular screening and cleaning mechanism. Inactive accounts are easy to become targets of vulnerability attacks. There is a large amount of business data in the Sensors system. Once leaked, the impact will be unimaginable. Based on data security considerations, a corresponding account regular screening and cleaning mechanism is established.

• Ordinary permission accounts with a recent login time of more than 30 days will be cancelled;
• Analyst privilege accounts with a recent login time of more than 15 days will be downgraded to general privileges;
• Analyst accounts that only use the dashboard for common functions will be downgraded to normal permissions.

4. Standardization of statistical caliber

Based on the survey results of each business department and the AARRR model, we compiled the "Statistical Caliber Standards for Commonly Used Atomic Indicators" and released it to operations colleagues. It is a must-have tool, especially for novices.

The screenshots are as follows:

5. Regularization of training and assessment

A regular training and assessment mechanism has been established. It is worth mentioning that those who are responsible for the training and assessment are outstanding students of the first phase of Operation Alpha.

1) Training schedule

• Week 1: Training + Q&A + problem communication;
• Week 2: Assessment + first round of scoring;
• Week 3: Second round of make-up exams + final elimination list confirmation.

2) Assessment method

• Assessment method: completion rate (20%) + online test of objective questions (30%) + data form preparation report (50%).
• Assessment criteria: 100 points in total, with a passing score of 70 points. If the data form is not submitted, it will be deemed as giving up this round of training assessment and the account will be directly closed.
• Reward and punishment method: If you fail the first round of assessment, you can take a make-up exam within three days (the conditions for passing the make-up exam: the score of the second round of objective test is above 80 points, and the submitted data form has a large iteration after evaluation); after the make-up exam, the elimination list of this round will be confirmed, and the corresponding business line manager will be synchronized on xx day and the analyst account will be closed.

6. Final Thoughts

The article is coming to an end here. This article summarizes and sorts out the practical plans for the implementation and promotion of data products that I have promoted, and systematically explains each link. In each article, I try to start from my personal practice process, combine it with the online education business background and the objective constraints at the time, and explain it to you as detailed as possible.

I have used three articles to explain the path to enterprise digitization, from underlying construction to MVP solutions, and then to promotion and implementation. This article serves as a node.

In the next series of content, I would like to summarize and sort out how data empowers the marketing business line from the perspective of another marketing CRM product line that I am responsible for, combined with the marketing business scenarios of online education.

Author: Tigerhu

Source: A data person’s private land