The entire process of product construction from 0 to 1

This article reviews the entire process of building a data analysis product from 0 to 1 in which the author participated. On the one hand, it is to better understand the overall picture of product development, and on the other hand, it is to summarize the construction process and learn from experience.

1. Project Review

This section is divided into four aspects: project background, project implementation, project results, and future planning.

1. Project Background

The project is mainly based on a self-developed new technology for product construction. The capabilities of this technology can be simply described as: collecting user behavior data on web pages, and then mapping the data into video to achieve a "screen recorder" effect. The capabilities and advantages of this technology enable it to have a wide range of application scenarios, such as behavioral quality inspection and user research.

Our goal in launching this project is to find specific business scenarios for the technology, verify the technology and product value, and ultimately achieve commercial success.

2. Project implementation

During the entire project, my work can be divided into two parts: product commercialization exploration and product design . In actual work, these two parts are carried out in parallel, but in order to facilitate sorting and summarizing, each part is explained separately.

(1) Commercialization Exploration

How to explore product commercialization? In fact, the core point of the problem is to determine user value and commercial value . To further break down the problem, determining user value requires determining the three factors of "people, goods and place", that is, what kind of service the product provides and what problem it solves for users; determining commercial value requires determining the three factors of the current market size and future prospects, and the situation of competitors.

Only by fully understanding user value and business value can we know ourselves and our opponents.

Commercial exploration work is carried out around these six factors.

• Understand the technology capabilities. Through technical exchanges, industry sharing and document learning, you can gain a general understanding of the principles and capabilities of the technology, as well as the scenarios in which each capability can be applied.
• Conduct user research to understand user needs. Conduct research and gain understanding through interviews with internal and external users, product presentations, and data collection. After analyzing and summarizing the needs surveyed, we found that the needs can be mainly divided into two categories: 1. User behavior analysis, which is used for business empowerment; 2. Behavior retrospection, which is used for evidence retention, quality inspection, etc.
• Competitive product analysis. Find competing products with corresponding technical capabilities, conduct interviews, trials, functional analysis, etc. Based on competitive product analysis, we can know the existing commercial scenarios of the technology, which can prove that there is market demand. On the other hand, we can understand the solutions of competitive products in various scenarios.

The above three tasks are a general survey of the possible commercialization directions of the product, which need to be further focused and refined. After multiple user surveys, we broke down user needs in each scenario to make them clearer and independent of each other.

After sorting out all the survey content, we obtained the table below.

Finally, the commercialization direction is determined to be behavioral analysis, mainly based on the following considerations:

• The company's future development strategy is in the direction of big data
• Existing potential customers are more interested in user behavior analysis
• Developing products that are not behavioral analysis products requires business involvement for promotion, and it is unknown whether the customer will be successful or not

Therefore, in order to verify the technology and product value more quickly, after comprehensively considering the above three points, we decided to use behavioral analysis as the product positioning. Specifically, by visualizing and digitizing user behavior, we help customers fully understand and analyze user behavior and achieve data-driven business growth. When customers use the product, there is no need to collect data. They can perform user behavior analysis by simply defining data indicators based on business needs.

Regarding the exploration process of product positioning, you can read an article I wrote before: "Review of the Exploration of Product Positioning"

But when reviewing, I found:

• When analyzing the six factors, the analysis of market size and future prospects was omitted.
• When conducting user surveys, the survey subjects were not fully covered, for example, technicians were not surveyed (debugging ability).

In addition, the current product positioning is still relatively coarse. Whether it is subsequent product design or real commercial output, it is necessary to further refine the product positioning and find specific product entry points. There are so many behavioral analysis products on the market, and you need to give customers a reason to choose you instead of other products. This reason is the product entry point. It is a concentrated reflection of the product's capabilities and a moat, so it must be "sharp" enough.

Based on the above ideas, a preliminary segmentation plan has been developed, but it needs to be verified through in-depth research (SWOT analysis, etc.). This will be the focus of work in the coming period.

(2) Product design

In this part, the main work is to design the product according to the product positioning. The design ideas are:

1. Conduct business research , sort out business status, and summarize business problems;
2. Establish product portraits , summarize and deduce demand analysis, transform user needs into product needs, and clarify product context;
3. Sort out the overall product plan , abstract and refine product functional modules based on business understanding, and determine priorities;
4. Product function design to promote product launch and iteration.

We first used the company’s post-loan collection department as a pilot scenario and designed the product around this business scenario.

1) Business research

Conduct research on the post-loan department through user interviews, field observations, etc., and sort out the current status and problems of the collection business.

The main business goal of the post-loan department is to improve the collection rate. To do this, we can find the core factors that affect the collection rate by sorting out the case collection process and breaking down the process. The business processes sorted out according to the survey results are as follows:

When a loan is overdue, the case is collected (an overdue loan is called a case). First, the collection system automatically runs a batch deduction to collect the user's bank card balance. When the deduction fails, the case is transferred to the case distribution engine system, which assigns the case to the collector according to certain rules. The collector works on the electronic collection workbench system. Cases that have not been collected after multiple collection attempts are outsourced or taken to court.

From the above, we can see that the collection of overdue loans is mainly completed through two methods: automatic deduction by the system and collection calls by collectors. The operability of automatic deduction by the system is not strong, and whether the deduction is successful depends entirely on whether there is a balance in the user's bank card. Therefore, the core factor affecting the collection rate lies in the collection call by collectors .

However, in real business scenarios, the actual work behavior of debt collectors is unknowable and uncontrollable :

1. Currently, the assessment is only based on performance completion, but it is unclear what causes the performance differences among expediters;
2. All the work of the collector is done on the electric collector workbench, but it is unclear how the collector uses the workbench, the usage of each function, and whether additional functions or case information support are required;
3. Debt collectors often report system operation problems encountered during the collection process. Product or technical personnel need to understand the specific scenarios when the problems occur in order to locate and solve them.
4. The debt collection industry is an industry with strict policy supervision. Debt collectors need to strictly abide by compliance and quality inspection regulations. Currently, this is done through manual inspections and audio recording quality inspections, but it is not clear whether the debt collection behavior violates regulations.

There are many similar situations as above. Obviously, the management is unable to fully understand the actual work behavior of the collectors , and there are management blind spots.

2) Create a product profile

By analyzing the needs of typical users, we can understand what services the product provides to users and what problems it solves, so as to clarify the product context and build a product portrait.

Customer problem diagnosis : Extract three typical user roles from the survey results, analyze the needs of these three types of users, convert user needs into product needs, and derive product solution design.

According to the above analysis of typical user portraits, their demand is to understand the work behavior of the debt collectors. Digging deeper, we can find out why the management wants to understand the work behavior of the debt collectors. This is related to the department's goal, which is to improve the collection rate. In other words, the management wants a tool product that can measure the collection effect. This product must be able to comprehensively portray the work behavior of the debt collectors, optimize the workflow and improve efficiency, and comprehensively manage the case collection process in combination with actual business data.

Product portrait creation:

Combining the above analysis conclusions and product positioning, we decided to develop a product for analyzing and tracing employee work behaviors. By digitizing employee work behaviors, we can help them fully understand and analyze their work situations, measure collection results, and ultimately help customers drive business growth with data.

The key value point of the product is the analysis and application of employee behavior data. The business planning path is: first digitize and visualize employee behavior, build a portrait of employee work scenarios, and achieve a comprehensive examination of employee work conditions. Finally, combine business data to examine the performance of each link and influencing factors in the collection process.

3) Overall product solution

Based on the implementation path, further abstract and refine the product functional modules .

Generally speaking, when building the information architecture of data analysis products, you first need to complete the construction of the data indicator system based on business understanding, find the core components, determine the analysis methods for each component, determine the corresponding product function modules based on the analysis methods, and finally connect the product modules through work processes and core pain points. However, during the construction process, standardization issues must be considered. After all, post-loan is only one of the customers of our product.

First, we break down the core indicator of collection rate, as shown in the following figure:

According to the above indicator system, the factors affecting the collection rate can be divided into three categories: cases, collectors, and behavior.

• The analysis methods for users include: user portrait, user grouping, etc.
• The methods of behavior analysis include: behavior path, etc.
• The analysis of cases can be done by using distribution analysis methods, etc.: ① Analysis of digital intervals (different overdue stages, overdue amounts); ② Analysis of fields by type: different types of cases (first reminder, inventory, personal, third party)
• The funnel analysis method can be used for the entire collection path

Based on the above three factors and analysis methods, the product can be divided into four major modules. Each module analyzes the impact of one or several factors on the collection effect: display the changes in core indicators and the basic situation of various factors through the dashboard , understand the changing patterns and characteristics of various factors through behavioral analysis , examine the use of collection tools through website analysis , and understand the basic characteristics of each collector on each factor through user analysis. The details are as follows:

• Dashboard : Displays the changes in core indicators and the basic situation of various factors
• Behavior analysis : Visualize the work behavior of the debt collector through behavioral retrospection to intuitively understand the debt collector's collection behavior; understand the basic work situation of the debt collector through work analysis; understand the behavioral rules and behavior patterns of the debt collector through behavioral paths; find the points with low "conversion rate" in the case flow process through funnel analysis; understand the distribution of cases, calls, behaviors, and debt collectors through distribution analysis; find out which behaviors of the debt collectors contribute to the successful conversion of debt collection through attribution analysis
• Website analysis : Through page analysis, you can understand the basic access situation of each page on the workbench; through heat map analysis, you can observe the approximate usage of specific functions in the function page; through page click analysis, you can accurately know the click ratio of each function module, the number of clickers and other data information
• User analysis : Classify users with specific common characteristics through user grouping as screening conditions for various analysis models; understand the basic situation of each agent through user portraits

After determining the main functions of the product, you need to determine the priority of each function . According to the survey, the first demand of the management is to fully understand the work behavior of the debt collectors, and the actual business analysis mainly lacks the behavioral indicators of the debt collectors. In combination with our technical core, which is to map the data into "video", in terms of implementation ideas, we first use behavioral backtracking as the entry point to digitize and visualize the behavior of employees, and then analyze and refine the employee behavior patterns, work conditions, etc. based on business and methodology. Finally, we digitize the entire collection process, combine process indicators with business indicators, and analyze the performance of each link.

After the product design is completed, some analysis functions can be used alone or in combination to feed back data to business development. A typical usage scenario is to establish a user portrait to conduct personalized matching of collection cases, thereby improving the collection rate.

We can further imagine that in the more distant future, when data is enriched, automation and intelligence will be achieved. For example, one of the current pain points is that the call connection rate is very low, and it is impossible to manually judge whether a case is connected. We can use data to predict the connection probability of a case, and the collection staff can perform collection work based on the case connection probability, which will reduce the workload of collection staff in making invalid calls.

Of course, thinking from a standardization perspective :

• Clearly, post-loan collection is only one of the scenarios. By analyzing this scenario, we can find out what functional modules the data analysis product may have, which ones are general and which ones are customized.
• At the data processing level, how to standardize and configure data indicator definitions and metadata management also needs to be considered.
• Improve the technical solution. Because the current data collection method is all embedded points and supports the web side, for some data that cannot be collected, whether to add other technical solutions to supplement the lack of source data

4) Functional design and implementation

Based on priorities, we design specific functional modules and coordinate resources to promote product launch and continuous iteration.

When designing a module, factors that need to be considered include: who is the user, what is the value of the functional module, whether the value is consistent with the user's goal of using the module, what is the operational path to achieve this goal, whether there are any abnormal situations that have not been considered, etc. Finally, the product function design is completed in combination with relevant product design rules.

Align the designed prototype with the business side first, call together technical staff to determine the implementation details, and after the meeting, determine the personnel arrangements and schedule to control the project progress.

3. Project Results

When the project is launched, the goal is to commercialize the product within six months to one year. How to convert goals into measurable indicators requires breaking down the goals into two aspects: ① the completeness of product functions, and ② the user value volume, which can be measured by the number of connected customers.

The current project results are as follows:

• Product functional integrity: The product has completed 4 major version iterations, 60% of the functions are online, and the online operation is normal
• Customer access status: 4 customers have been connected, and several departments also have access needs (willingness to access was expressed after the product roadshow)

In general, after more than half a year, the product commercialization progress bar has reached about 70%.

4. Future plans

Still divided into two parts: commercial exploration and product design

(1) Commercialization Exploration

• Vertically, further refine the product entry points, as well as the corresponding market size and prospects, and evaluate the value of each entry point through SWOT analysis
• Horizontally, evaluate the possibilities and opportunities of other commercialization directions, etc.

(2) Product design

• Based on business scenarios, further product detail solution design
• Continue to track the performance of existing online functions, collect user feedback, and perform iterative optimization

2. Thoughts and Conclusions

I joined the project team by chance. Looking back on the past six months since I started participating in the project, I have taken many detours and moved forward by trial and error. This process has taught me a lot, not only in terms of supplementing and improving the product manager's unique work skills, but also in terms of improving some potential soft skills. I have gained a lot.

First, summarize the difficulties encountered in the project

• When exploring product commercialization, due to the lack of relevant experience and knowledge reserves, many pitfalls occurred, wasting time and energy. Later, through practical exploration (more communication, more review) + some methodological learning (books, courses), I knew how to think about this problem and what the solution was.
• Lack of ability to control project progress. In the process of pushing products online, sometimes they cannot be delivered on schedule. In addition to objective factors (personnel changes), the main subjective factor is that I am not a technical person, which leads to inaccurate assessment of the development workload. Later, we increased our control over the project by ① asking technical experts to help with the evaluation and ② refining the development work, setting goals and regular checks, which allowed us to better identify the risk of project delays.

Secondly, summarize the gains in the project

• Deepen the understanding of product work. Participating in this project from scratch has given me a more comprehensive understanding of the entire workflow of a product manager and a deeper appreciation and understanding of each link.
• Improve product work skills. For example, communication skills. B-side products must interact more with business parties to understand their needs. In addition, when doing business promotion, the ability to speak different words to different people is exercised; such as learning ability, logical thinking ability, problem-finding ability, etc.

Finally, the road is long and arduous, but I will continue to explore. Keep moving forward!

If there are any shortcomings in the article, please point them out.

Author: Smell the Rose

Source: Smell the Rose