The basic data knowledge that designers need to master has been summarized for you by experts from big companies!

Preface

In vivo's professional course system for campus designers, there is an introductory course called "Basic Knowledge and Application of Data", which received unanimous praise from college students last year. In today's world where data-based design and verification are popular, every designer needs to master some basic knowledge of data. So I specially found the courseware and combined it with my own understanding to write this short article, hoping that it will be helpful to you who want to learn data analysis.

The full text is divided into 3 major modules:

• Clarify the value of data to designers. (First establish the value recognition, then there will be motivation to learn)
• Understand the data models and data indicators commonly used by designers. (Sort out the learning map of data-related knowledge, from macro, meso to micro, and be targeted)
• Learn how to do simple calculations and data analysis. (A few Excel formulas + pivot tables + data comparisons will do it. I believe you can do it too.)

Identify the value of data to designers

In "U Yidianliao", the author compares data to the rational light of design, which is a very appropriate metaphor. In real life, designers are more or less emotional. Making good use of data can help designers add rational thinking, so that the design plan takes into account both emotional and rational light.

In short, data has the following benefits for designers:

1. Before designing - finding problems

By comparing the data horizontally (with competitors), we can understand the current situation and gaps and find the focus of design.

For example:

After the launch of vivo Wallet V1.0, by extracting conversion data and comparing it with competing products, we can find the bottleneck of the conversion rate funnel and the gap with competing products, so as to carry out targeted optimization design of the entry, guidance, card selection and card activation links.

Through longitudinal comparison of data (and historical comparison), we can understand the patterns and changes and iterate the design solutions.

For example:

After the browser history record was revised, the conversion rate of the history record was significantly reduced, especially the first history record, which dropped by about 10%. Considering the click frequency and operation convenience of the first history record, we made an iterative design to strengthen the first history record.

By comparing the data among different groups (user stratification), we can find out the functional attention/involvement of different groups and present different design interfaces to different users.

For example:

In view of the low involvement of new vivo video users, insufficient motivation for interaction, and inaccurate recommendations, different design solutions can be presented to meet different user demands.

2. Design time - assist decision making

When designing a plan, if you are undecided about multiple directions, you can use rapid surveys to obtain data to assist in design decisions.

For example:

When we were designing the video product presentation mode, we had different opinions on light mode and dark mode: the operations staff preferred light mode, worried that dark mode would be depressing and that the children and parents on the preschool channel would not like it. The designers, on the other hand, believed that the dark mode had a small contrast between the content and the background, making it more visually comfortable and more immersive to watch videos. It was difficult for both sides to convince each other, so we obtained user feedback through a quick questionnaire: most users preferred dark mode and believed that dark colors were more eye-friendly, thus reaching a consensus on the design direction.

3. Post-design - Verify the design

By comparing the grayscale or formal launch data with the expected data, we can judge the degree of achievement of the design goals and summarize and accumulate the corresponding design experience.

For example:

In the optimization design of the cash register page, the project team had different opinions on whether to add a premium membership activation module. Finally, they decided to launch two plans and observe the data results. From the grayscale data results, the payment click rate and success rate of plan 1, which did not display premium membership, were higher than those of the original plan, indicating that the design optimization effect was significant; the payment rate of displaying premium membership was reduced, and finally it was decided to launch plan 1. After this grayscale, everyone is more convinced: on the payment page, do not easily add design elements and functions (to avoid visual and cognitive loads, resulting in lower conversion rates).

Since data can bring so many benefits to designers, from which dimensions should designers understand data?

In "30 Questions for Product Managers to Cultivate Data", the author puts forward such an idea, which I deeply agree with and would like to share with you: You can build a global view of product data from three dimensions: macro, meso and micro:

Macro industry insights

Studying macro industry analysis reports helps us understand the size of target users, the industry structure and ecology of the product, and thus helps us better understand the advantages and disadvantages of the product relative to competing products.

Zhongguan Product Overview

Sorting out the product's data system helps us understand the product's core indicators, the current status and fluctuations of the indicators, the significance of the indicators and their impact on the business, as well as the relationship between the indicators and the design, and identify the focus of the design.

Micro data insights

Grasping the usage links of each core function and understanding the conversion rate funnel on the core links will help us restore the user's usage scenarios, think about how to better serve users and user goals, and make better designs.

From macro to micro, it is the process for designers to reshape the product in their minds and establish a holistic view of product design. With such a data perspective, designers will be more rational and objective about the current status of the product.

Understand the data models and data indicators commonly used by designers

Common data models in the industry are Google's HEART model (on this basis, the Alibaba design team iterated the 5-degree model which is more widely used in China, so this article mainly introduces the 5-degree model) and the AARRR model.

1. 5-degree model

The user experience cycle is divided into five stages: reach, action, perception, return visit, and communication. The core goals corresponding to these five stages are attraction, completion, satisfaction, loyalty, and recommendation.

Attractiveness

Attractiveness refers to whether a product/function can be noticed by users when they first come into contact with the product/function before operation, and whether it can attract the users' attention and interest, thereby generating corresponding behaviors.

Relevant user experience data indicators include (but are not limited to) awareness rate, reach rate, click rate, exit rate, etc. At the level of attractiveness, what designers need to analyze most is the exposure rate and click rate of the page/function/content to see whether the design entrance meets the design expectations.

Completeness

Completion refers to whether the user can complete the operation process corresponding to the product goal during the operation, as well as the operation efficiency in the process of completing the goal. The main user experience data indicators include (but are not limited to) first click time, operation completion time, number of operation completion clicks, operation completion rate, operation failure rate, operation error rate, etc.

At the completion level, what designers need to analyze most is the conversion funnel of the core path and the length of user stay, from which the usability of the core link design can be judged.

Satisfaction

Satisfaction refers to the subjective feelings and satisfaction of users after the operation is completed. The main user experience data indicators include (but are not limited to) subjective evaluations of layout rationality, content comprehensibility, ease of operation, and interface aesthetics.

Satisfaction data mainly comes from qualitative and quantitative user surveys. The user research team generally conducts a satisfaction survey every six months or a year. Designers should pay close attention to the peaks and troughs of the satisfaction survey, and try to fill the troughs and raise the peaks in subsequent designs.

Loyalty

Loyalty refers to whether a user will use a product again after completing a single use. The main user experience data indicators include (but are not limited to) 30-day/7-day return visit rate, usage overlap rate on different platforms, etc.

Loyalty is closely related to user retention rate. Designers can design appropriate triggers based on return visit rate and user usage scenarios to remind users to use the product. At the same time, they can learn from the experience advantages of competing products and convert competing product users.

Recommended

Recommendation degree refers to whether users will recommend this product to others. The main user experience data indicator is: Net Promoter Score (NPS).

Recommendation is similar to satisfaction, and is also obtained through qualitative and quantitative surveys of users. The difference is that satisfaction emphasizes more on the user's own experience, while recommendation is related to the user's external word-of-mouth communication.

Designers should grasp the advantages and disadvantages of products in word-of-mouth communication, quickly improve disadvantages, block the spread of public opinion, and at the same time continuously strengthen advantages, guide users to spread, and form a virtuous word-of-mouth cycle.

2. AARRR Model

The AARRR growth model comes from growth hackers and is also known as the Pirate Model.

AARRR is the abbreviation of Acquisition, Activation, Retention, Revenue, and Refer, which correspond to the five important links in the user life cycle.

Customer Acquisition

Through certain promotional methods, the product can be displayed on some channels, and the users who see the display can be converted into product users.

During the customer acquisition stage, it is necessary to calculate the product's promotion methods (CAC, CPC, CPT, CPM, CPS, CPA), promotion costs, and promotion effects to evaluate the return on investment (ROI) of the promotion channels and help make product promotion decisions.

activation

Convert users introduced through channels into active users of products, and improve product stickiness and depth of usage.

During the activation phase, you need to calculate the conversion rate of the product’s core links, understand the conversion bottlenecks, and make corresponding design optimizations.

Retention

How to keep users using our products, reduce user churn, increase user stickiness, and prevent users from leaving the product.

During the retention stage, we need to combine the user's usage scenarios, design appropriate triggers, and improve user retention rate.

income

Obtain revenue from users or advertisers through certain means and channels.

In the revenue stage, it is necessary to analyze the product's revenue methods and revenue proportions to stimulate user consumption frequency and amount.

recommend

By improving product competitiveness and providing benefits, users can recommend our products to their friends.

The recommendation stage should be combined with scenarios to encourage and guide users to share and bring about positive product reputation.

Regarding the AARRR model, I found a picture like this on the Internet, which marks the data indicators that designers can pay attention to at each stage. If you use the AARRR model in your design, you might as well use the data indicators in the picture to analyze your design results.

If you are new to data and are confused about the English abbreviations and data indicators in the figure, you can click to view the "Common Data Indicator Definitions List" compiled by our team, which I believe will be helpful to you. The link is as follows: https://docs.qq.com/sheet/DS1pHcWl0SFJQdHZp .

Learn how to analyze data simply

In a project team, a designer mainly uses data for design insights and analysis, rather than collecting, processing, presenting, deeply analyzing, and accumulating data. Therefore, if a designer can complete the following basic data analysis, he or she will be able to perform well in daily design analysis.

1. Look at the data

Know what data platforms the company has, what data can be viewed on each platform, and be able to speak out the core data indicators of the product.

Know and understand the product's tracking data table. When you have questions about the interpretation of the data, you can consult the tracking data table to understand the true meaning and statistical caliber of the data;

Know how to view or download data that is not included in the data report, reduce design dependence on data analysts, and improve design efficiency;

2. Calculate data

Be able to use the company's data platform. If the data platform supports it, you can directly select simple data points in the data platform to complete the data calculation of click-through rate and conversion funnel (faster and more intuitive).

If the company's data platform is not perfect, you can use Excel to complete the following simple data calculations.

Find the average of the data (=AVERAGE(cell range)).

When performing data analysis, in order to avoid errors caused by daily data fluctuations, the average value of data over a period of time is usually taken as the basic calculation benchmark value.

(When I take data myself, I usually take a week's worth of data to calculate the average, and try to avoid holidays. This is to avoid the impact of holiday traffic and special demand fluctuations, and to avoid differences in user behavior on weekdays and weekends.)

Find the click-through rate (= click PV/exposure PV)

This is the most basic and most commonly used data operation, which is a simple division.

(Note that the exposure PV should be taken from the corresponding click element, not the entire page. Select the PV click rate or UV click rate according to the specific situation)

Be able to use pivot tables to batch calculate averages/click rates/conversion rates.

It is quite troublesome to calculate data one by one. Pivot tables can help us batch calculate and present the data we want to view, which is very convenient and fast. To be honest, although I have been paying attention to data for many years, my data calculations are limited to Excel pivot tables, which can basically solve my daily data analysis needs. If you encounter data that you really don’t know how to calculate, you can directly ask a data analyst for help. After all, everyone has their own expertise, and we can focus more on data analysis and thinking about design solutions.

3. Analyze the data

Data is ultimately just a bunch of numbers. Knowing these numbers is meaningless. What is meaningful is to extract your own insights from these numbers and assist your own thinking. This is the real value of data.

For specific data analysis, please refer to the following analysis context:

• Confirm the analysis objectives;
• Identify data indicators relevant to the analysis objectives;
• View and extract relevant data indicators and make corresponding calculations based on the analysis objectives;
• Compare the calculated values ​​horizontally/vertically/layered, find the gap between the data and competitors/history/crowd, and think about how to make up for or even surpass these data differences through design. (For specific cases, please refer to the first section of this article)

summary

Now, let's review the key points of this article:

• Clarify the value of data to design (discovering problems, assisting decision-making, and verifying design) and establish a global view of data analysis.
• Understand the data models (5-degree model and AARRR model) and data indicators commonly used by designers, know the framework of data analysis, and know which indicators to start the analysis.
• Know how to perform simple data calculations and analysis to eliminate the fear of the unknown in data learning.

Finally, I would like to add that for designers, it is not important whether they can do data calculation or not (indeed, many designers feel overwhelmed when it comes to data calculation, but they can still cultivate data design awareness through cooperation with data analysts/data product managers). What is important is to have data thinking, be able to discover problems from existing data, assist in decision-making, and verify solutions, thereby helping us make more rational decisions.