E-commerce search traffic allocation strategy

For major e-commerce platforms, search is an indispensable function. At present, full free competition has basically been achieved in the distribution of search traffic. The author of this article has given some thoughts on future traffic distribution and summarized five aspects that can be continuously improved.

1. Introduction

Looking at search from the perspective of system construction provides system builders with a holistic picture. Product managers are not limited to a specific functional module, but look at search from the perspective of the industry, company, and system, including the value of the search platform (both qualitative and quantitative), the interests of stakeholders (users, merchants, and companies), interest coordination mechanisms, and rules, regulations, and standards.

The entire e-commerce search platform consists of five major systems (product system, strategy system, operation system, data system and IT system). The inherent logical relationship between them and the division of responsibilities among departments within the company enable each participant to not only see their own local work clearly, but also clearly see the overall picture of e-commerce search, the industry and the economic principles behind it.

As product managers, we encounter many problems in our work. Many times, we are not clear about the working mechanism of the platform, the stakeholders behind it are unclear, the product plan goals are out of focus, and there is no value judgment standard. If the product manager has a comprehensive understanding and a value judgment standard in mind, he will have a guiding star to do things.

The biggest responsibility of e-commerce search is to achieve traffic distribution to achieve a win-win situation for users, merchants and platforms. Based on the principles of fairness and transparency, it activates market themes (merchants), provides users with a good user experience, and achieves sustainable development of the platform itself.

For e-commerce platforms, in the distribution of search traffic, there is currently full free competition, certain market macro-control, and full commercial monetization. In the future, I believe that traffic distribution can be continuously improved in the following five aspects. This article serves as a few thoughts on e-commerce search traffic distribution, and I hope it can serve as a starting point for further discussion.

2. Five core problems and solutions for current e-commerce search

• Problem 1: There is a lack of top-level macro-control mechanisms. The main method is to use micro-level single product competition to divide traffic, which results in the market being solidified in the top single products of the top brands, which is not conducive to the growth of mid-level and tail merchants and is not conducive to activating market vitality.

Strategy: Change the traffic distribution mechanism that is based solely on the competitiveness of individual products, and build a traffic distribution system for an open ecosystem. Mid- and tail-end merchants will also have a certain share of the market, ensuring that mid- and tail-end merchants also have the opportunity to be exposed in the top positions of searches, and through competition between mid- and tail-end merchants, stimulate the top-end merchants to increase their investment in the platform.

In terms of specific operating strategies, a certain share of the top slots will be allocated to mid-level and tail-end merchants, and at the same time, the assessment of the output of these slots will be strengthened.

The former is to give opportunities to mid- and low-end merchants, allowing them to see hope for growth. The latter is to force this type of merchants to strengthen operations and management through assessments of pit output. After strengthening operations and giving up part of the profits (such as price cuts), it can make up for the lack of competitiveness of some products, increase customer acquisition and conversion capabilities, and at the same time, user experience can be made up, pit output can be guaranteed, and at the same time, a certain amount of pressure is put on the top merchants/top brands. They feel the competitive pressure from latecomers, forcing them not to relax. Once the output of the pit is guaranteed, the interests of the platform are also partially protected.

In terms of specific operations, the traffic on the APP is generally focused on the top 20 slots (the specific situation is based on the actual data of the platform). In this way, one slot can be allocated from the top 20 slots and given to mid- and low-end merchants for free. At the same time, product and operation standard assessments are set for this slot, and the assessment rules are communicated to the merchants. The conditions involve supply preparation, price level, preferential strength, basic operations, customer service and other aspects. This slot selects the best products from the mid- and low-end merchants, and after the traffic arrives, it is adjusted according to the actual traffic situation.

• Problem 2: At the micro level, it is difficult to motivate merchants to operate in addition to sales by allocating traffic based on sales.

Strategy: Increase the weight of traffic efficiency and reduce the weight of sales at the micro level. Through efficiency and operation quality tiered access, merchants can optimize operations and gain the opportunity to get a front-row position in search results.

Our current traffic allocation strategy places too much weight on sales volume and sales revenue, which is caused by the platform's pressure to monetize. However, this leads to frequent hot-selling products and a serious solidification phenomenon. To alleviate this phenomenon, in addition to personalized matching, we also need to change the strategy of placing too much weight on sales in traffic allocation and increase the assessment of traffic utilization efficiency. That is, some products, as long as they are given traffic, have high exposure click-through rate and order conversion rate, but the sales accumulation in the past period of time is insufficient. Increasing the weight in this aspect will help new forces stand out, give the old products in the front row a certain impact, and break the monopoly of the old products in the front row.

In terms of specific operations, after observing that new and cutting-edge products have arrived in the front row every day, we must keep up with the assessment of the monetization capabilities of the slots, and reach the merchants through the merchant-side products to guide them on how to proceed next, adjust prices, adjust discounts, follow up on services, optimize basic operations, optimize evaluations, and take over new traffic.

• Problem 3: Insufficient human-computer interaction in the search scenario, and search results do not change in real time based on positive and negative feedback from users and the system

Strategy: By drawing on the human-computer interaction feedback mechanism similar to Toutiao’s information flow, search results can be adjusted in real time based on user feedback.

In today's e-commerce search systems, once each search is completed, the result is finalized. The user's actions on the search results page cannot affect the ranking of the result page for this search. However, the headlines' information flow recommendations and the current user's behavior will affect the content of subsequent information flows. An ideal search should be able to reflect the user's real-time decisions.

For example, if a user has viewed several products but has not clicked on them, it means that the user may not be interested in such products, and similar products should be reduced below the current search results page. If a user stays at a certain product for a long time, it means that the user is interested in this product or similar products, but the user may need to compare. At this time, a comparison button will appear under such products to provide comparison services. In short, the ideal search is dynamic, and the system can interact with users in real time, becoming a real-time intelligent shopping guide for users.

• Problem 4: No collaboration between the site and the external site

Strategy: Reward a certain percentage of on-site traffic based on the merchant's off-site traffic

The biggest problem facing the current platform growth is user growth. From a social perspective, the scale of Internet users has basically peaked, and the cost for e-commerce platforms to attract new users from the existing Internet user pool is high. As of Q1 2021, Pinduoduo, the best at attracting new users, has an average cost of 190 yuan per new user, not to mention JD.com's 290 yuan and Alibaba's 458 yuan. It can be seen that attracting new users is a thankless task.

At this stage, attracting new customers requires gathering more forces and more sources. Among the three main roles of consumers, merchants and platforms on the e-commerce platform, consumers need to be acquired. In the past, the platform itself was working hard, but now the strength and wisdom of merchants must be incorporated into the overall goal. Merchants have more diverse means and scenarios for attracting new customers, and there are still many low-cost measures to attract new customers, especially omni-channel new customer acquisition. We should make full use of merchants' offline scene resources, implant them into users' minds, and achieve omni-channel collaboration.

In general, merchants can attract new customers outside the site through advertising, social networking, live streaming, membership marketing, offline scenarios and many other means, and cash in on the e-commerce platform. In order to attract merchants to acquire new users on their own platforms, e-commerce platforms need to formulate an on-site traffic reward or compensation mechanism for off-site new users, so that the merchants' cost of acquiring new users can be amortized over the free on-site traffic. It is also one of the main measures to attract off-site users.

• Problem 5: No synergy between paid and free

Strategy: Currently, the distribution of free traffic and paid traffic on many e-commerce platforms is still independent. The distribution of free traffic must be able to leverage the monetization of paid traffic. The free traffic here is not just the free traffic in search, but the free traffic of merchants on the platform. The payment here is not just the payment in search, but the overall payment amount of merchants on the platform.

The acquisition of free traffic is dynamically adjusted based on the proportion of merchant paid traffic to achieve a balance between costs and benefits for the platform and merchants. This issue is actually an economic issue. The level of mechanism design behind it tests the platform designer's depth of insight into the problem and his ability to regulate the system. Currently, there are not many platforms that can implement this mechanism, mainly because there are not many platform designers who look at the problem from this perspective. Secondly, there are higher requirements for the platform's own traffic regulation capabilities.

The former requires platform designers to have a rich and profound understanding of platform economics, an economic understanding of the "bottom-level" operating mechanisms of society, and a grasp of commercial practices; the latter requires the platform to have the ability to regulate the flow of traffic across society and to be able to achieve the ability to reduce peaks and fill gaps in a single game.

At this level, a virtuous interaction is formed between the platform and merchants, and the platform distributes incremental value on the basis of sustainable development for both parties. The platform can take the initiative in market competition and decide the division of incremental value distribution, otherwise it will become a passive follower.

Conclusion

E-commerce in a broad sense includes both traditional e-commerce, such as Taobao, JD.com, and Pinduoduo, as well as other online transaction businesses, such as Ctrip's travel business, Meituan's various online local life services, various online vegetable, medicine, and car sales, etc. E-commerce in a broad sense is the most mature and most competitive digital business.

The distance between these businesses and money is second only to financial businesses. At present, Internet finance is subject to state supervision, and a large amount of funds continue to flow into the e-commerce business. Catalyzed by this, a large number of talents are also pouring into this industry. Currently, many strategies are being implemented on e-commerce platforms. In the early stage, we sorted out the traffic distribution of e-commerce searches, from the free competition stage, the market macro-control stage and the commercialization and monetization stage. Of course, these stages are not separated in time, and are now running in parallel.

On this basis, the platform has not reached perfection. This article sorts out the evolution of traffic distribution from five aspects, involving the optimization of platform ecology (questions 1 and 2), the positive interaction between consumers and platforms (question 3), the interaction between merchants and platforms in user growth (question 4) and the free and paid collaboration of the platform (question 5). The goal is to optimize the sustainable development ecology (merchants, users and consumers), continuously create incremental value and distribute the incremental value in a free and collaborative manner. These are the ultimate goals of traffic distribution on e-commerce platforms.

IV. Postscript

The current society has entered the second half of the demographic dividend. Every company is improving its digital capabilities and working to tap into the value realization and value enhancement of existing users (Kuaishou, Douyin, Baidu, and Sina are all engaged in e-commerce, and offline companies are busy transforming). Matching supply and demand transactions between goods/services and consumers is the main way to commercialize and realize profits, and search recommendations are the main mechanism for achieving supply and demand matching.

At present, only a few leading Internet companies in China have a mature and complete talent team. Other companies either have scattered professional talents and insufficient knowledge structure, or have no such talents at all. The contradiction between the huge talent demand in society and the serious shortage of high-quality talent in society has restricted the improvement of many companies' business in this area, and also restricted the improvement of young people's professional income and career space.

Given the professional nature of search recommendations, there is relatively little publicly available systematic information, and the difficulty of the information varies, making it difficult to meet the urgent needs of businesses and professionals.

I have focused on e-commerce for the past 10 years. I have been deeply involved in and led the construction of the search recommendation traffic distribution and product management systems of two major e-commerce platforms. I am familiar with the strategies, products, operations, data and R&D aspects of e-commerce platforms. I have summarized and refined them into building a practical e-commerce search system, building a practical e-commerce recommendation system, building a practical e-commerce product management system, and a series of articles on observations on new retail in e-commerce.

Author: Mao Xinnian

Source: Senior e-commerce expert Mao Xinnian