User operation: new funnel model for conversion analysis!

In the paid conversion scenario, we often use funnel charts to analyze problems. However, this funnel chart analysis can only tell us the current business status, but not the specific reasons. We cannot find the iteration direction based on the funnel chart data.

A new funnel: the user decision process funnel

1. Structuring the decision-making process

What exactly do users want to do? The buying process is generally very fixed.

Generally, it goes through the following stages: attract attention → generate interest → collect information → evaluate value → purchase decision → post-purchase behavior

2. Determine model indicators

Once you have a model, you know what information to collect.

Attract attention: For example, users who are exposed to the advertisement can be considered as the number of users covered in the step of "attracting attention".

Generate interest: Clicks on ads can be considered as the number of users who have "generated interest".

Information collection stage: This stage is more complicated. Depending on the actual situation, the number of users in this step can be “the number of users who read to the bottom of the page”, “the number of users who browse the product brochure”, etc. If this is an early experiment, you may want to intentionally design some buttons in order to collect data. For example, you can hide some key information and click "Expand" to see the complete information. That way you can determine which information users prefer to see.

Value assessment stage: Users who have completed the "information collection" stage can be considered to have entered the value assessment stage, or the price information and other assessment-related information can be placed directly on a separate page, with the number of users entering the page as the basis.

For some non-standard products, I suggest putting the price information at the end, because price is an important part of evaluating value. If the price is announced too early, we cannot be sure whether there is a problem in the "information collection" stage, or whether the price is announced too early, causing users to "evaluate the value" very early.

I have done experiments:

Original page: Product introduction page (with price) → Purchase page → Payment page

Change to: Product introduction page (without price) → Purchase page (with price) → Payment page

After the revision, the proportion of people entering the purchase page increased, while the conversion from the purchase page to the payment page decreased, and the overall conversion rate remained almost unchanged.

Although this change had no impact on the business at the time, it greatly facilitated my evaluation of the effectiveness of the revised product introduction page because there was no interference from price factors.

Of course, as time goes by, prices will become more and more transparent, and it is inevitable to use price as a selling point. However, the above point is still relatively effective as a preliminary experiment.

Purchase decision: Users who click Buy are considered to have completed the purchase decision.

The above indicators can determine how many users there are in this step. We also need some auxiliary indicators, such as: the average length of time users stay in this step, the average number of button clicks, etc.

3. Decision-making process funnel

Create a funnel chart based on the user decision process and develop a user decision funnel from the indicators in the previous step. For example, in the previous step, the number of users entering the evaluation stage/the number of users entering the information collection stage can be used to judge the pros and cons of the information collection step.

Comparison with traditional funnel chart:

Traditional funnel: exposure → product landing page → purchase page → click to purchase → successful payment

Decision-making process funnel: attract attention → generate interest → collect information → evaluate value → purchase decision → post-purchase behavior

The new funnel can directly identify the root cause of the problem and guide the direction of revision. For example: the user conversion rate between "collecting information → evaluating value" is low, the average stay time per person at this stage is relatively short, and the click-through rate of each selling point in the information collection page is relatively high. It can be basically assumed that the user has glanced at the page but was not impressed by the selling points.

In this way, the direction of the revision will be much clearer than the traditional funnel.

The difference between the two funnels

1. If the process itself is not set up properly, the traditional funnel chart is meaningless.

Because the funnel chart is built according to the steps of our page settings, if our page process itself is not set up reasonably, then the basis of the analysis will be wrong.

Case: A purchase conversion on a fund sales APP.

The original purchase process is:

Go to the recent hot sectors → Hot sector information → Purchase hot sectors → List of funds in this sector → Fund information → Purchase funds → Fund purchase page → Purchase successful.

Later, I analyzed and believed that: during the stage of purchasing popular sections, users had already completed the purchase decision process for popular sections and were already in the purchase decision stage. The existing process requires users to re-enter the process of selecting specific funds and re-enter the stage of collecting information.

So, I suggest the following modification:

Modified process: enter hot sectors, hot sector information, purchase hot sectors, fund purchase page, purchase successful.

After modification: The overall sales conversion rate increased by 240%.

2. The traditional funnel tells us about user behavior, but does not clearly define user needs.

Even if the process is correct, the traditional funnel chart can only show the current status of the business. After finding the problem, we still need to continue to analyze the root cause and countermeasures. The user decision process funnel can directly see the problem points and user needs, and can quickly find the solution direction.

When we analyze a problem, there are generally several levels:

1. User Conduct
2. User needs
3. Our Goal

Generally, we proceed step by step, inferring user needs through behavior and finding our goals through needs.

The traditional funnel chart can only show the user's behavior, while the decision-making process funnel can reflect the user's needs, which allows us to find out more quickly: what do we need to do?

The traditional funnel chart is a way to look at problems from the company's perspective. This will naturally lead to improvement plans that are out of touch with the user. For example, the proportion of users entering the purchase page from the product page is not high. Based on this information, we may prompt various stimulating information, such as "limited to 2 hours", "limited to the first 1,000 users", etc.

This will increase the number of users entering the purchase page, but these users will still be lost during the payment process. (This is not to say that such measures are ineffective, but the effectiveness of such measures will be greatly reduced if the user's basic processes are not met.)

But if we use the user decision process funnel, we can clearly see that there is a problem in the stage where users collect information. It may be that our selling points are not chosen correctly, or the expression is too complicated, etc. The correct approach should be to adjust the information or display method of the product page. Moreover, the funnel chart is prone to fall into the KPI trap, which results in a forced pull and no major change in the overall monetization.

3. The original funnel easily deviates from user thinking.

For users, a purchase process is nothing more than collecting information, evaluating value and other stages. The product page and purchase page in the traditional funnel diagram are just specific manifestations of these user needs. If we use the traditional funnel chart, we are likely to develop rigid thinking and make minor modifications to the existing purchasing process.

for example:

• Does the purchase button on the product page have to jump to the purchase page?
• For the user, at this point he has already made a purchase decision, and he doesn’t really care whether he enters a separate purchase page or other forms. Since there is usually a 1-2 second delay in jumping to the page, would it be better if a floating window popped up directly after clicking to purchase to make a selection?

From the user's perspective, there can be more ideas for revisions rather than just minor repairs.

4. The new funnel helps cultivate data thinking

When we do business analysis, the first thing we need to solve is: "What data should I collect?"

Data analysis does not start after the business goes online. We have already started data analysis when designing the page.

If the page design is not conducive to data collection, it will be difficult to conduct in-depth data analysis later.

If we build a new funnel model and consider the indicator system for each step, it will be easy to know what data to collect. Some of the originally designed pages are not easy to distinguish the specific steps, so we designed them in a way that is suitable for data collection. For example, if there is too much information on the original page and it is impossible to tell which step the user is in, then the evaluation information can be placed on the second screen or made into a jump.

Related reading:

1. User operation: How to use B-side operation thinking to increase user growth?

2. APP user growth: How to use data analysis to increase user growth?

3. Product operation: How to use data analysis to drive product user growth?

4. APP user growth: One model solves 90% of growth problems!

5.How to increase users? Take Pinduoduo and Xiaohongshu as examples

6. Triggering user growth: Is user operation just about attracting new users?

Author: Three-dimensional variance

Source: ternary variance (sanyuanfangcha)