Pinduoduo Marketing Promotion Methods in 2022

As Pinduoduo grows in size, there is a saying that goes "It was looked down upon before, but now it's really popular." Many brands have also settled in Pinduoduo, but how can they achieve sales growth on Pinduoduo? We recently invited the former Pinduoduo store manager of a beauty brand to share replicable operational experience.

This article introduces Pinduoduo's marketing methods in detail, helping you to systematically understand how to conduct marketing through Pinduoduo channels.

The sharing content is as follows:

1. Pinduoduo’s recommended traffic characteristics

2. Operational characteristics of Pinduoduo cosmetics industry

3. Pinduoduo Resource Position Operation

4. Pinduoduo’s on-site advertising: Lookalike logic and excessive fee ratio (may affect DSR)

1. Pinduoduo’s recommended traffic characteristics

1. How to understand Costco + Disneyland

This is a statement that Huang Zheng repeatedly mentioned in the prospectus. When I was an investor, I could understand Costco's ultimate value for money, but I was confused about the Disneyland part. After working in the business for a year, I have a deeper understanding of:

(1) Costco is really pursuing Costco's ultimate product selection model. Costco pursues fewer SKUs, customized super-large packaging, and ultimate cost-effectiveness. This is highly consistent with the fact that Pinduoduo is currently the only mainstream e-commerce platform that has achieved in-depth development of industrial belt merchants (i.e. white-label supply chain).

Pinduoduo has not only produced many white-label SKUs that dominate the entire network (such as paper towels and some agricultural products), but also continues to promote branding with the support of the platform.

Although there are not many Pinduoduo brands that have really become popular at present, I think there are quite a few that have sold well, if we don’t emphasize “brand sense”. Therefore, Pinduoduo will continue to promote the most cost-effective SKUs in terms of investment promotion. The same applies to brands. It will come up with cost-effective models that are differentiated from Tmall in exchange for various resource positions of Pinduoduo.

(2) Disneyland is worthy of its name. It is the company that has done the best job in gamification and entertainment among shelf e-commerce. Of course, it cannot currently compare with content e-commerce companies such as Douyin and Kuaishou that started out as Disneyland.

The role of entertainment is to increase activity at low cost. Duoduo Orchard is still one of the largest traffic entrances on the homepage. The waiter often sells the orchard’s resources to NPCs, and the current bidding is exaggerated to 200,000 yuan per day before and after the big promotion.

The problem is that the people who play the game of stealing vegetables have insufficient purchasing power, and the level of traffic is comparable to that of dou+ without selecting any options. Therefore, for beauty products, which is the second smallest category on Pinduoduo, the accuracy is very poor. At least the ROI did not exceed 0.3 in the few times I bought them. I heard that brands such as Yunifang once sold facial masks for 9.9 yuan and achieved an ROI of 3% of Orchard NPC. I can only say that for extremely broad populations, it is necessary to choose products that are universal enough and have sufficiently exaggerated prices.

2. Pinduoduo search/recommendation traffic share

Pinduoduo uses the Taobao ban as an excuse (partially true, partly to get out of trouble), and has extremely low data transparency. It is often impossible to see the data dimensions provided by Taobao (even if they are indexed).

Regarding the traffic share of the overall market, waiters from different backgrounds have different opinions.

From my experience, I agree with the statement that, when adding up all the traffic sources on the entire platform, recommended traffic accounts for 40%, resource positions account for 20%, search accounts for 30%, and social portals such as Pinxiaoquan account for 10%. Of course, this ratio is based on the UV caliber, and the UV values ​​of different traffic sources vary greatly. However, the recommended traffic accounts for 40%, which should be more than twice the proportion of Taobao. It can be considered as the source of all Pinduoduo operation logic that I will write later.

I think there are two basic understandings for operating Pinduoduo: one is to think about the logic Pinduoduo uses to make recommendations; the other is to think about the actions Pinduoduo uses to differentiate itself from the Taobao system. It just so happens that I am currently working on Douyin e-commerce, and I found that Douyin’s core operating logic is similar to these two points, but Douyin has taken a completely different recommendation logic from Pinduoduo and has taken a differentiated path from the Taobao system.

3. Strong single product and weak store logic, from the platform's traffic distribution efficiency + economies of scale

The most efficient recommendation is of course to promote a very universal single product into a best-seller on all platforms within a period of time, rather than truly achieving the so-called "one size fits all". The quantity of each product promoted is relatively small. Without inventory depth and scale effect, it is not enough to reversely require the brand/white-label supply chain to reorganize the supply chain of this single product, make customized specifications and cost-effectiveness, and differentiate itself from the Taobao system.

Therefore, Pinduoduo's traffic recommendations are more based on the weight of the product itself (even for products with multiple links, the weight of a certain link will be mentioned below), rather than on the weight of the store.

The various indicators of a store can be considered as the weighted total of all the products in the store, but they do not have a strong impact on the products themselves.

The reason for this judgment is that we have pulled up a very important piece of data to support it: we have two large stores AB and several other small stores. In order to increase linkage and thus increase the average order value, we pull out all the products with the highest sales share in store A (such as A1/A2...) and the products with the highest sales share in store B (such as B1/B2...), and then sort them by sales together, such as A1/A2/B1/A3/B2..., and we uniformly call them X1/X2... Then we pull the cross-repurchase rate of X1/X2... in the last 60 days (X1-X1, X1-X2, X1-X3...X2-X1, X2-X2, X2-X3...), which includes various dimensions such as the repurchase rate of this product, the repurchase rate of cross-products, the repurchase rate of this store, and the repurchase rate of cross-stores.

As a result, products with high cross-repurchase rates are basically products that have relatively high sales themselves. For example, the repurchase rate of X1-X2 will be much higher than that of X2-X5, even though X1-X2 are across stores and have weak category association (foundation-lip makeup), while X2-X5 are in the same store and have strong category association (foundation-foundation).

This set of data shows that the logic of strong single products and weak stores on Pinduoduo has been confirmed, and the best way to improve linkage is not to promote same-store combinations, but to promote single product links one after another.

4. The importance of the recommendation system and DSR

Under the recommendation system, the platform needs to provide credit endorsement for the products recommended to users. Only with better user conversion can the efficiency of the recommendation system and user satisfaction be higher.

The core of recommending products is to look at the price-performance ratio. "Price" is relatively simple, just look at the absolute price directly (this is also a feature of Pinduoduo. For example, 2 bottles of makeup remover are 60 yuan, but they are not as good as one bottle of makeup remover at 35 yuan, because many Pinduoduo users really do not have the money to buy two bottles. This is the reality of the world, and it is even more true on Pinduoduo). The absolute low price is strongly weighted; the judgment of "performance" needs to rely on DSR.

DSR is the weighted sum of more than ten after-sales indicators, including quality return rate, return dispute rate, etc. There are single product DSR and store DSR (the latter is the weighted sum of the former). Once the single product DSR declines, the natural traffic of the link (the calculation method is shown below) will drop significantly. Since DSR is composed of many indicators, the workload of analyzing them one by one is not low, and some DSR indicators are extremely strict (if there is a return dispute, the product will be directly downgraded for 7 days), so improving DSR after-sales service is an extremely important link for Pinduoduo.

As mentioned above, Pinduoduo needs to differentiate itself from Taobao. Therefore, Pinduoduo will be extremely biased towards users. As long as users file a complaint, there is basically no chance that the merchant will win, and the appeal will be basically useless.

There are many low-price users on Pinduoduo. The lower-price users are, the more and more outrageous the problems they encounter, and the more they need to be appeased by compensating customers with goods, which can be regarded as the cost of advertising. Therefore, the way to improve DSR is also very simple and crude. Just stuff a parcel card, give a red envelope of more than 2 yuan, and call to ask for a good review. You must give money.

2. The overall number of orders in South China increased by 10% compared with last year

1. Pinduoduo Cosmetics Scale vs. Tmall and JD.com

The estimated market share of cosmetics in 2020 is: Tmall has 40 billion, JD.com has 6 billion, and Pinduoduo has only 2 billion, and about half of them are estimated to be white-label. Nearly half of the 1 billion of branded products are contributed by international brands with billions of subsidies. Therefore, the scale that a brand can truly achieve is actually quite small.

To give a simple comparison, in August last year, a certain beauty brand had sales of 150 million on Tmall, JD.com and Vipshop each had sales of 30 million, and Pinduoduo only had sales of 10 million. Small is small, and fast-growing is also small. We must have reasonable expectations for the scale of Pinduoduo stores.

2. Pinduoduo makeup price range

The price range is mainly white-label, with unit prices ranging from 12 to 19 yuan. From the perspective of the brand price range of domestic products, our experience is that 32 yuan is an obvious sales turning point. For products priced above 32 yuan, even if they receive traffic resources with billions of subsidies, it is difficult for most of them to sell in large quantities, unless they were/are currently popular products on the entire network.

3. Characteristics of Pinduoduo’s three-level cosmetics categories

My logic is that offline dealers in lower-tier regions will be about 6 months slower than online dealers. Therefore, last year's online hot-selling model may become an offline hot-selling model in the lower-tier regions this year, and the online inventory of this model is shipped on Pinduoduo, forming a temporal resonance with offline consumer perceptions.

Therefore, we found that the models that were popular last year by using all the resources of the entire network (typically such as Li Jiaqi) can still sell well on Pinduoduo this year as long as the inventory price is given at around 60% off, while new products on Tmall that are not discounted are not selling well. Pinduoduo users’ style awareness seems to be about one year behind that of Tmall. From the perspective of the tertiary categories, the time lag seems to be longer, possibly 3-5 years.

The main sales volume of Pinduoduo cosmetics are lip makeup and foundation makeup, while eye makeup and various upgraded tertiary categories (highlighter/contouring/setting spray, etc.) are difficult to sell. Therefore, the eye makeup advantages of a certain brand cannot be reflected at all on Pinduoduo (it just so happens that the flagship product such as the 12-color eye shadow is unlikely to have a significant price reduction on Pinduoduo). Instead, the hot-selling models are out-of-season anti-gravity lip glaze, small gold diamond lipstick, old loose powder, etc.

4. Pinduoduo Cosmetics Investment Strategy

Pinduoduo’s beauty category has 1 brand group (directly dealing with brands), 2 industry belt groups (white-label), and 2 catfish groups (for authorized dealers, the purpose is to force brands to provide cheaper goods by setting random prices, otherwise they will not be able to hang up the 10 billion yuan subsidy).

The competition between the brand group and the catfish group is very fierce. The catfish group often takes out genuine inventory goods at prices that are completely unacceptable to the brand, and claims tens of billions of subsidies. Appeals are basically useless (a certain brand has already become the number one domestic cosmetics brand on Pinduoduo). If they really want to sell inventory goods, they will continue to lower the prices, and then use the name of a specialty store, declaring that the price has nothing to do with the brand, but that it is genuine.

The shop assistants who deal more with brand operations include category investment promotion, special event operation shop assistants (CPS, expert live broadcasts, etc.), review shop assistants (10 billion review, flash sale review, etc.), and risk control shop assistants (influencing DSR, live broadcast room permissions, etc.). However, the latter two types of shop assistants can only communicate indirectly through category shop assistants (firewall rules), and there are also certain conflicts between the KPIs of category shop assistants and these two types of shop assistants, which further amplifies the uncertainty of operation rules.

For example, at the end of 2020, due to Duoduo Maicai's black hole-like absorption of Pinduoduo's internal talents, the number of clerks in charge of the 10 billion yuan audit was significantly reduced, the audit became slower, but the audit standards for price comparisons became wider.

By the way, due to Tmall’s “Choose One of Two” policy, the name of the brand’s flagship store cannot be revealed, so authenticity verification becomes very important.

5. Low transparency of Pinduoduo’s backend data and its response

Pinduoduo’s shop assistants will use competition from Taobao as a shield for all topics, especially data transparency. Pinduoduo's backend is scattered and not interconnected. For example, for a single product link, I can only see its daily sales, promotion sales, CPS sales, and live broadcast sales from different backend entrances, but there is no full-link, full-traffic composition analysis from the perspective of this product. The most important of these is the sales brought by natural traffic = total daily sales of the link - promotion sales - CPS sales - live broadcast sales - other quantifiable sales. Natural sales are strongly correlated with the DSR of a single product and are the ultimate goal of all paid operations.

Therefore, we need to build an Excel function table ourselves, collect the background data every day using the most primitive copy and paste method (Pinduoduo's background anti-crawling ability is very strong), and then let the data automatically calculate after pasting, analyze the composition of the full link and full traffic port of each product, and provide a reliable judgment basis for our operation strategy. Let’s take a counterexample: the sales volume of a certain product has been growing rapidly recently, but the promotion costs have also been growing rapidly. It is not known which factor is driving the growth of this product.

Through the above semi-automated analysis in Excel, you may find that the natural sales of the product have dropped significantly recently. Further inspection reveals that the DSR is dropping. Looking at the negative reviews, you find that the product has quality problems due to overheating during transportation in summer. Therefore, you can quickly feedback to the logistics department for processing, avoiding the embarrassing situation where the natural sales decline rate exceeds the promotional sales growth rate.

3. Pinduoduo Resource Position Operation

1. 10 billion yuan subsidy

Since the price range characteristics of most categories are similar to those of cosmetics, that is, white-label industrial belt merchants account for the majority, and the price range is less than half of the price range of mainstream brands, in addition to the search traffic brought by the brand itself, the 10 billion subsidies have become a core outlet for brand merchants in terms of recommendation-side traffic.

The transparency of the rules for obtaining the resources for the 10 billion yuan subsidy is very low. From the perspective of the brand authorization chain, any brand product that claims to be "within the third level of authorization" can be included. In practice, many products from unauthorized dealers (but genuine) can also get the 10 billion yuan subsidy. Judging from the price comparison rules, manual review is very common. For example, Vipshop's multi-piece multi-discount products compete with Pinduoduo's single-piece prices, or non-brand Taobao C stores show lower prices, or even Tmall Juhuasuan deliberately subsidizes small Bs to help certain SKUs break the price. All of these may be captured by the price comparison system and Pinduoduo's store manager may be notified to allocate tens of billions of resources within 24 hours, requiring public relations.

Most brands regard the 10 billion yuan subsidy as their core traffic, which will cause the entire product portfolio to be heavily dependent on the 10 billion yuan subsidy.

In the short term, there will be two major risks: one is the targeted attack by Tmall, especially in the sensitive period of buying volume before and after the big promotion, and most company bosses will still abandon the car to protect the driver and comply with Tmall's interests; the second is the loss of active control over the goods. After reaching 10 billion, the link cannot increase the price but can only reduce the price, and the price must be slightly lower than the data of shelf e-commerce that can be captured on the entire network (including the brand’s own other links to the same products must also be higher than the 10 billion link price), and other types of activities such as flash sales and some resource positions cannot be carried out with 10 billion, which is equivalent to "giving your life to the traffic outlet of 10 billion subsidies" and also affecting the operation of other flash sale links.

Here is another piece of experience: if a new billion-dollar product has not seen any growth in volume within three weeks, it is basically considered that the system recommendation algorithm has sentenced this link to death. You can consider changing the link and re-submitting it (the review will be more difficult) or switching to other products that have reached billions.

2. Various flash sales

Compared with 10 billion, flash sales have a higher degree of operational freedom, require higher experience in operational operations, and have greater uncertainty in the volume of product links (for 10 billion, it is very clear whether the volume can be increased given 3 weeks).

On Pinduoduo, flash sale resources include ordinary flash sales, billion-dollar flash sales, brand flash sales, super flash sales and other types of resources. The rules are obviously different, and there is also a lot of gray area in the price comparison rules for cooperation with the store manager.

The randomness of the number of flash sale links is similar to the randomness of promotional advertising units. The price-performance ratio of a product itself is equivalent to an advertising plan, which basically limits the ceiling of the number of links under this product;

A link in this product is equivalent to a unit in a plan. The randomness of the running volume will converge to the final ceiling of the plan as the unit is continuously replicated and rerun.

Therefore, the more common link-cultivation strategy on Pinduoduo is during daily operations (not before the event resource position arrives! It is very important). In order to disperse the risk of over-reliance on the 10 billion subsidies, it is necessary to set up 2-3 non-10 billion links for the same products as the 10 billion, sign up for various flash sales, and continuously race out the links with the highest sales volume (the common link is to report 10441 first and then 584/585), so as to maintain the volume through long-term flash sales.

3. Resource slots (paid/free)

Resource positions are a source of traffic that requires very strong operations and experience. They are characterized by high uncertainty, strong explosiveness, and large ROI fluctuations. They are typically divided into big promotion resource positions and daily customized resource positions.

The resources for big promotions are usually available half a month before the promotion, and most brands can purchase these resources through fair bidding, with prices generally being in the region of 50,000 yuan. But in the field of cosmetics, the general situation is that there is only exposure, and the ROI is likely to be lower than 0.3. Daily customized resource positions are resource positions for which leading brands can receive targeted invitations from store clerks, and are mainly exchanged for free resource positions in exchange for limited-time price reductions. The increase in output is relatively obvious (since there is no advertising cost in many cases, there is no ROI to speak of, but daily sales on the day of the event may reach 2-3 times of daily sales). Typical examples include 10,000-person groups (brand/single product), 10 billion single pits/half pits (resource positions for brands exposed within the 10 billion subsidy outlet).

Considering that the output is mainly in the free targeted resource position, we will mainly talk about the operation in this area. The core is to cultivate one or more high-authority flash sale links in advance.

What often happens is that 1-2 days before the targeted resource position goes online, the review clerk suddenly changes the strategy, causing the category clerk to temporarily request that the 10 billion/first flash sale link cannot be used, and a second highest weight flash sale link needs to be temporarily added. If the flash sale link has very little basic sales at this time, or the DSR is low, it will not be possible to quickly convert the strong purchasing power users who click into the resource position.

4. CPS: Duoduojinbao, Kuaituantuan

The scale of Duoduo Taoke is much smaller than that of Taobao Taoke. In order to quickly develop Duoduo Taoke, Pinduoduo also launched different CPS projects internally. This will cause the attention of the store manager to be distracted, the CPS backend to be extremely imperfect, the rules to be unclear and may adversely affect the qualification of the same billion-dollar products. At the same time, it will lead to low output of the CPS port and weak stability.

If you need to increase volume quickly, be prepared to break prices and increase commissions.

4. Pinduoduo’s on-site advertising: Lookalike logic and excessive fee ratio (may affect DSR)

Pinduoduo's commercial recommendation system can be seen as a primary version of Douyin's commercial recommendation system. It is not because Pinduoduo's backend lacks the ability to label users, but because the amount of data labels provided by Pinduoduo's data center for commercialization is very small and not accurate enough, but this is still enough to support the lookalike logic of on-site delivery.

The core of Lookalike is the proportional enlargement of seed user portraits. For example, the initial user portraits are 75% of those born after 2000 and 25% of those born before 2000. Without excessive expenditure, the user portraits after the growth caused by the subsequent system recommendations and the traffic from the recommendation system (scenario promotion) during promotion will still grow into similar proportions.

To briefly describe a certain brand's promotion on Pinduoduo , the cost-to-performance ratio is 40% and the ROI is 1.3, which is almost the limit of the cost-to-performance ratio that does not negatively affect the store's DSR. We tried to lower the cost-to-performance ratio, and the ROI could be brought to 1.7. The daily sales decline was not obvious, but due to KPI requirements, we eventually brought the promotion costs back to a high level.

On other platforms, after deliberately increasing the fee ratio, it may happen that the money cannot be spent, or the ROI declines slightly. However, what happened on Pinduoduo is likely that on the one hand, the ROI dropped sharply, and on the other hand, the store's DSR also dropped rapidly, and there was no obvious increase in daily sales after increasing the fee ratio.

We think the reasons for this may be:

1. The paid sales from the promotion "cover" the sales from the natural traffic, which means that the promotion system has increased the repeated delivery to old customers in order to increase the sales volume. And these users are likely to make a deal when they see the natural traffic.
2. In order to further increase the amount of promotion expenses, the promotion system captured an inaccurate population, that is, the non-brand makeup population (considering that the beauty category is only the second-to-last size on Pinduoduo, the accurate population is very small). As a result, this group of people not only has a low conversion rate (ROI drops rapidly), but also has low category awareness and brand awareness, resulting in a significantly high rate of pre-shipment returns and customer complaints.

Therefore, the DSR was significantly lowered. Even on the day of the strong pull of Febi, the DSR began to show a clear turning point and began to decline rapidly. We believe that the response strategy to this situation is also very simple, which is to still have correct expectations for the cosmetics operations on Pinduoduo. First, the scale will not be very large (due to the size of the category). Second, the recommendation-based advertising system does not recommend that you spend too much money (burning money), but instead uses lower prices to promote transactions (burning goods) and make accurate crowd portraits and after-sales services (DSR).

V. Conclusion

Pinduoduo has a large number of users, but the product prices are generally low. For brands, if the product price is right, they can expand sales through this channel.

Author: Sally Nan

Source: Nankaikou