3 major challenges in acquiring new customers!

First of all, why do we need to do customer acquisition analysis? It’s actually very simple.

In the Internet age, traffic is money, traffic is hard currency, and all major products hope to control more traffic, but traffic does not come out of thin air. Whether it is offline invitations, word-of-mouth communication, or promotion, it is a process of obtaining traffic.

The reason why we need to do customer acquisition analysis is that the key point is that it involves money and budget. This is often the first thing we need to analyze, so we need to have a clear understanding of traffic, from quantity, quality to subsequent conversions, so that we can get better results by spending the same money.

Let's first sort out the complete paid delivery scenario. There are mainly two types of third-party delivery content:

Advertising placement: For example, the splash screen ads of some apps are ads embedded in the information flow of such apps, or advertising banners.

Sem keyword placement: mainly refers to the paid placement of Baidu keywords.

The first one is to check whether the app is installed on the user’s phone. The user will be redirected to the app store to install the app and then open it.

Second, if the user has the app installed on their phone when clicking on the landing page, the landing page will be directly woken up and the app will be opened;

Third, the web-based product will directly enter the web-based product for use.

The first is to verify and monitor the data of partners. Enterprises have to spend money directly on advertising, but there is no way to objectively monitor their data from the enterprise's perspective. In this case, they can only rely on the data provided to us by a third party for verification.

The second is how to identify a unique user, which is actually a very fundamental pain point. The user goes from the landing promotion page to the web page, and then jumps to the app store to download the app. The app store is a complete third party, and the company cannot collect data. At this time, it is even impossible to measure whether this is the behavior of one user on two platforms, or whether there are just two users at all. If the problem of user identification cannot be solved, it will be impossible to confirm whether the user is natural traffic or traffic from promotional activities. There is no question of doing analysis.

The third problem is that the data from various platforms cannot be aggregated and analyzed in a unified manner. For example, a user on Android is constantly browsing products; the final order is placed on iOS. If behaviors cannot be aggregated and analyzed, we can only assume that there is a very valuable user on Android who is lost before paying. In this case, if we use other strategies or send SMS reminders to such users, it will be useless and may cause user disgust.

Ad monitoring technology has been evolving continuously since the beginning of advertising. The difficulty here is that the advertisements are on third-party platforms, and general companies cannot directly add data monitoring codes. Therefore, mainstream advertising monitoring technology will record the exposure and clicks of advertisements through the number of requests for images, thereby realizing the monitoring of third-party advertising-related data.

On the other hand, for advertisement detection we usually use the method of adding parameters. This was a solution developed by Google at the time. Now it has basically become a common method in the industry and is recognized by the industry. Through UTM parameters, different parameters are configured for the promotion pages of different channels. When the user enters the landing page, the browser will record the address of the URL, and directly parse the specific settings of the URL parameters to clarify the channel and keyword from which the user enters our product, and finally summarize it.

Therefore, when it comes to partner data issues, it mainly involves the time before the user enters the landing promotion page and the previous advertising exposure. Advertising monitoring technology is a complete solution. There are actually many products of this kind on the market. At the same time, Zhuge also has analysis on advertising, monitoring and evaluation of effects. If you are interested, you can visit Zhuge to learn more.

There are five independent dimensions: source, media, event name, event content and keywords. When these five dimensions of parameters are added to a URL, it is basically enough to cover all news scenarios.

Of course, the name of each parameter is essentially a code. The reason for defining the name is just to give you a preliminary overview of the application scenarios. You can also set parameters completely different from what the names mean. In essence, all you need to do is know which parameter corresponds to which dimension of your design. After configuration, the data of the transmitted content will be automatically recorded, allowing us to understand the number and trend of traffic brought by different keywords and different media, and ultimately help us conduct such an analysis.

There are several common scenarios in terms of user identification, and each scenario involves a process in which the user changes from a non-logged-in state to a logged-in state. Behind this is also the issue of uniquely identifying a user after using multiple devices and multiple technical platforms.

First of all, on the web page, cookie information can be stored in the browser for the convenience of users. It is a common practice in the industry to identify users through cookies. But there is a possibility that a user uses multiple browsers on multiple devices.

Therefore, after a user logs in, the user account is generally used as the only identification to connect the behaviors on various platforms and browsers;

If the user is not logged in, we mainly use the browser cookie to identify the user;

When a user changes from a logged in state to a logged in state, we use the user account information as the user's unique identification. And at the association level, through the uniqueness of cookies, no matter whether the user is before or after logging in, the cookie is unique, so the user's behavior before logging in can be connected.

The mobile side is actually the same as the js side, but the most difficult part is how to identify users between the js mobile side and the app mobile side.

For example, you promoted an event and then promoted it on Zhihu, Douban, and Tieba at the same time. When the user opens the promotion page and clicks download, he will be redirected to the Apple Store or App Store, Wandoujia and other app stores. When you launch the app, you don’t know whether the user was brought from Zhihu or Douban. You can only see those from the Apple Store or App Store. But this obviously does not help our promotional activities at all, and in the end they may all be placed on the same app platform or app store.

Let me explain to you the principles of Deeplink technology.

The third pain point mentioned earlier is that the data from various platforms cannot be aggregated for unified analysis. I just mentioned the problem of user unique identification. Once it is solved, it will be easier to conduct summary analysis. In Zhuge's final solution, we provide the ability to analyze data from all platforms as a whole. As long as the user's behavioral data is on our own platform, we can integrate it together for analysis to realize the complete user life cycle behavior.

For example, we can accurately see which users come from Zhihu's promotion channels, and how many of these users from Zhihu have registered, and what are the behavioral characteristics of these users in js, Android and iOS. By monitoring and monitoring user behavior across the entire platform, we can actually clearly understand the quality of users in each channel. It can help us understand, for example, whether the js side has the function of attracting traffic, whether the mobile app side has the function of retaining users, etc.

Today we briefly talked about the three key points in customer acquisition analysis. In fact, the value of these three dimensions is that they can accurately measure the user's acquisition source and new active status, and can see on the platform more macroscopically and realistically whether the new active users on multiple platforms are one person or multiple people; the second is to analyze the user's flow status between platforms. I believe that most companies have a positioning for the application end of each platform. When we see the user's flow status between platforms, we can know whether the user is as we designed.

In short, now that most companies have all-platform clients, it is very necessary to integrate data from various platforms for analysis.

Author: Zhuge io

Source: Zhugeio