's understanding of the value of Internet commercial projects should come from a deep understanding of the business model. The traditional estimation network effect and estimation GMV based valuation methods have become invalid.
Since the advent of the Internet era, how to evaluate the value of Internet commercial projects has become the most difficult puzzle to solve. A large number of institutional investors are actually using "big funds" to bet on "big tracks", looking at macro-industry trends rather than micro-industry logic. On the one hand, this created the myth of early investment such as , JD.com, , Vipshop, etc., but on the other hand, it also led to the failure of countless investors in the track.
In such uncertainty, a large number of researchers hope to build a model to accurately measure the value of Internet commercial projects. But after several efforts, the valuation of Internet commercial projects has not only not become clearer, but has become increasingly blurred. Even the saying that "investment in the early stage is investing in people (entrepreneurs)" has become a consensus in the industry.
I have no intention of denying this statement, but as a researcher of business model, it seems necessary to establish some more "logical" investment principles. From another perspective, the understanding of the value of Internet commercial projects should come from a deep understanding of the business model. Maybe, in this direction, we can do something else.
In 2014, I proposed a "bucket model" of Internet business model in my book "Superposition Experience: Designing a Business Model with Internet Thinking", and believed that user assets, ecological assets, and transformation capabilities (the ability to allocate resources within the ecosystem to users) are the three major elements of the Internet business model producing GMV.
In this issue, I will explore in-depth "how to measure user assets". Of course, this is aimed at the business model centered on "user (C)". For business models centered on "merchant (B)", such as B2B and S2B2C, although the principles remain unchanged, they should focus on "merchant assets".
1. Failed traditional valuation methods
In the Internet business world, there are only two mainstream valuation methods based on users: one is to estimate network effects; the other is to estimate GMV (total transaction volume).
's principle of estimating network effects mainly comes from a marketing point that Metcalf, the inventor of Ethernet, proposed when he sold his network card of 3Com, which he founded in the 1980s.
He believes that the more nodes are accessed in the network, the more connections it may form. If the number of nodes is N, the number of possible connections is N (N-1), that is, it is about the order of magnitude equal to N2. This view was summarized by George Gilder, the publisher of the Gilder Science and Technology Monthly, as " Metcalfe's Law" in 1993, and was briefly described as "the value of the network is proportional to the square of the number of connected devices."
On the basis of considering network effects, the Internet business model is based on Internet information technology and will also follow Moore's Law .
Moore's Law was proposed by Gordon Moore, one of the founders of Intel. That is, when the price remains unchanged, the number of components that can be accommodated on an integrated circuit will double every 18 to 24 months, and the performance will also double. In other words, the performance of computers that can be purchased for every dollar will more than double every 18 to 24 months. The two laws of
almost constitute the underlying logic of Internet business (you can also refer to Gilder's Law). It is based on these two laws that the former chief analyst of Morgan Stanley , famous Wall Street Securities analyst and investment banker Mary Meeker and colleagues proposed the "DEVA model (Discounted Equity Valuation Analysis, Stock Value Discount Analysis)" in the Internet Report published in 1995, namely:
E=MCCh
, where E is the economic value of the project, M is the initial capital invested for a single customer, and C is the value of a single customer. C² in this formula obviously follows Metcalfe's law, while M expresses the meaning of Moore's law.
Overall, once the investment in Internet commercial projects crosses the fixed cost line, subsequent growth will no longer be associated with linear changes in fixed costs, but will be exponential. For example, once the investment in infrastructure such as servers, databases, call centers and other infrastructure exceeds the baseline, the project value will enter an exponential growth track with the network value brought by the increase in users.
It is driven by this concept that American Internet companies have confidence in accumulating users and use the number of users as the biggest bargaining chip to support high valuation, especially for SNS projects. Behind the extremely high valuations obtained by connected companies such as Facebook, Twitter, and Google when raising funds are the logic of the DEVA model. The most typical example of
is the case in which Facebook acquired WhatsApp for $19 billion in February 2014. As a communications software, WhatsApp was founded in 2009 and has only about 50 employees. If we only use the price-to-earnings ratio and cash flow discounting methods, such a high valuation cannot be supported. But from the perspective of the DEVA model, its MAU (monthly active users) has reached 450 million, and Facebook can also obtain greater network effects by sharing users, so there is a reason for this valuation.
However, the bursting of the Internet bubble in 2000 alerted enthusiastic investors.
From the statistical data, the number of Internet users represented by the red line in Figure 1 has been growing, but the trend of the Nasdaq 100 index (blue line) is not proportional to the square (green line) of the number of users.
's cliff-like decline in 2000 makes people wonder - is the DEVA model really reliable?
Since then, researchers have tried to correct this model, but the final application effect is hard to say. Interestingly, in China, the valuation of a large number of Internet companies (especially early-stage companies) is still referring to this method. Perhaps there is no way to do it.
After the early stages of Internet commercial projects, some operational data may more accurately support investors' judgment of the value of the project. As a result, valuation methods based on operational data have gradually become mainstream. In summary, this type of model can be portrayed as:
E=K·ARPU·MAU·LT
where K is a constant, which is related to the industry, ARPU is single customer income, MAU is monthly active users, and LT is the user life cycle. ARPU·MAU is monthly income, while LT is based on monthly units, indicating how many months of such income can be obtained.
summarizes it: This model measures the GMV that an Internet business model can obtain throughout its entire life cycle.
Of course, there are certain problems with this valuation method. In addition to K, several other indicators are dynamically changed, and this change makes the valuation very shaky.
For example, if a company maintains the MAU at a high level by purchasing large amounts of traffic and subsidies, this is a typical false prosperity. If capital blindly enters based on this data, it will become a "taker".
2. Valuation of the essence of business model
We hope to build a valuation model that is closer to the value of the business model and evaluate the actual level of user assets.
1. The biggest problem with the traffic diversion × monetization
DEVA model is that it blurs the roles of different nodes in the network, assuming that the value of generating connections is comparable.
People who blindly use Metcalfe's law ignore the different network effects. Internet companies build platforms between supply and demand (of course, there are also self-operated forms to connect supply and demand), and there are same-side network effects and cross-side network effects.
The former refers to the increase of one joiner on the user or merchant side, and every additional joiner will bring about explosive growth of the value on the same side; while the latter refers to the increase of one joiner on the user or merchant side, and every additional joiner will bring about explosive growth of the value on the other side.
The former business model (mainly SNS) can still be used by Metcalf's law, and the latter can estimate the value of users in this way, which is very far-fetched.For example, WeChat users will increase the value of connections due to the addition of new users, while Meituan users rarely interact with each other. The increase in their connection value mainly comes from the increase of merchants on the other side.
looks further. If the user is acquired, will it be able to generate corresponding value? Can the monetization of different users be really the same? Just from the user's side, factors such as user characteristics, scene strength and other factors obviously may have a huge impact on monetization.
So, I think the basic logic of evaluating user assets should be:
E=V·R
where V (Volume) represents the traffic pool and R (Revenue) represents the monetization power (monetization possibility). This means that if Internet companies want to advocate their own value, they should strive to create a high-quality traffic pool on the one hand, and strive to increase the possibility of monetization on the other hand.
2. The benchmark × point number is more than
. The valuation method based on operational data seems to be too simple, and there is great uncertainty. In addition, the biggest problem is that this calculation is just a lagging result and the early value of the project user assets cannot be discovered.
From the model we gave above, the current status data of V and R can be observed or estimated, but this cannot conceal another question - will the calculation based on the current status data cover up the future possibilities of the project?
So, I expanded the model into two parts:
E=V1·R1+ V2·R2
where V1·R1 represents the current value of the project, and V2·R2 represents the future value of the project. Obviously, compared to the clarity of the former, the latter is vague. And this vague place may be the value under the iceberg of Internet business projects. This value may be an exponential growth based on the status quo, so it is difficult to calculate no matter what traditional valuation method (PE, PB, DCF, etc.).
This means that we need to choose another valuation method, that is:
A project value benchmark project value × Project value A number of projects/Benchmark project value points
Through the model, we can convert the value of project A into points, and then, through the benchmark projects with clear valuations on the market, we can calculate whether Project A is overestimated or underestimated. I believe that for Internet commercial projects, this should be a relatively reasonable valuation method.
also needs to be noted that this method of estimating project value using points is more suitable for companies in the same industry track. The more the track is the same, the more accurate the results are. Of course, considering that the Internet business paradigm will move towards "dir traffic × monetization" in the future, and convergence in scenarios will lead to track convergence, this valuation method may have a wider range of application.
3. VR matrix
Considering the VR model to evaluate traffic pools and monetization power from the current situation and the future levels, we can establish the following VR matrix (Table 1): The biggest innovation of the
matrix lies in considering the possibilities of the business model in the future.
In real operation, we generally use the "Liket Five-Point Scale" method to score. For each positive statement of the indicator, there will be five answers: "very agree", "agree", "not necessarily", "disagree" and "very disagree", and score 5, 4, 3, 2, and 1, respectively.
Of course, for some indicators, we will make the scale more concrete. Then, the weighted average is calculated in each module of the matrix, and the scores of the four modules can be obtained and the value points of the project can be calculated.
3. Traffic pool evaluation parsing
The evaluation of traffic pool actually reflects the quality of the Internet project ports.
I have proposed in the book "Superposition Experience: Designing Business Models with Internet Thinking" that Internet business models have two types of ports, one is a perfect terminal with excellent functions; the other is a value community that follows values. In other words, users may enter the port due to the excellent functions of the product or based on emotional attachment.
Of course, these two types of ports may overlap to some extent. In terms of attraction to users, this will produce a 1+12 effect, even if the user achieves scene immersion.Once the user is immersed in the scene, an irrational consumption impulse will arise.
Here, we need to explain the "irrational consumption impulse", which does not refer to the orientation of deceiving users; it means that in such a scenario, users are separated from simply being petty about prices, and instead understand the products or services they come into contact with on the port as "overall solutions" and are willing to pay corresponding high premiums for it.
For example, Xiaohongshu is a community e-commerce company, which has both the shopping function of e-commerce and a community with unique values. Its users will be more aware of certain products because of the Internet celebrity big Vs and tend to appreciate quality, unlike when shopping on Taobao. Here, what users experience is not just a product, but a way of life.
's existing traffic pool is mainly evaluated from four aspects: active users, occupation time, usage frequency, and user market share.
active users are effective indicators for evaluating the size of traffic pools, and their importance is self-evident; while the occupation time and frequency of use reflect the user's immersion in the ports, these two indicators are often easily overlooked; user market share focuses on the Matthew effect in the Internet business world, and high-occupancy ports will trigger the herd effect of other ports.
In addition to the existing traffic pool, we should also pay attention to the future space.
First of all, different industries have different market sizes, and the natural user base is different. Of course, doing business in heavy vertical fields is not a bad thing. After all, China is a big market, and a small relative number is also a large absolute number. Moreover, choosing a heavily vertical track may make it less difficult to obtain subsequent traffic.
Secondly, the pain points of users are also different. This is probably the most important indicator, if the entire matrix can only leave one indicator, it is it. Here, there is no need to use Likert's five-point scale, but the following original graded scale is as follows to confirm the pain points:
- 5 points - without this product, it is completely unimaginable and cannot bear it;
- 4 points - without this product, there will be huge losses in part, which can be borne, but it is very painful;
- 3 points - without this product, it will be inconvenient;
- 2 points - without this product, I miss it a little;
- 1 points - without this product, there will be no impact.
Here, 3 points are the golden point. If you have a score of more than 3, there is a high possibility of success in the project, but if you have a score of less than 3, you don’t need to do it, and it is impossible to succeed. The most tangled thing about
is that it falls at 3 points. If you want to do this kind of project, you must invest a lot of capital, prepare for long-term unprofitability, develop the habit of users, and spoil them so much that they are not used to losing this product (enter to level 4)! So, from this perspective, the starting point of Didi Chuxing is actually 3 minutes. People still travel before this product. However, they used a lot of capital to "burn" users' usage habits.
Again, based on user pain points, we need to observe the degree of attraction of project ports to users, which of course also includes comparison with competitors.
The positive effect of the same-edge network refers to whether the user can have self-interaction, while the positive effect of the cross-edge network refers to whether the user can be attracted by the solutions provided by the project. The latter is the reason why users enter the port, but even if the project continues to iterate over its own solutions, users may still experience burnout. For example, after shopping for some comprehensive e-commerce platforms for many years, users may turn to some vertical e-commerce platforms.
Therefore, SNS will be warmly welcomed by investors, because this type of product can stabilize user traffic, reduce retention costs, and even allow users to cross their life cycle. This is also why Jack Ma has always coveted the social cake of . From the beginning, Alibaba has personally done communication, to later investing in Weibo, Momo , and then doing DingTalk in person... Every step is to try to enter socializing in various ways. In addition to the two network effects,
also needs to consider irreplaceability. Homogeneous products mean that they do not have core competitiveness. Even if the traffic pool is large, it is a flash in the pan. From this perspective, Didi Chuxing needs to withstand severe challenges.Although it has occupied the position of the market leader for many years, after Meituan crossed the border, it was once torn a lot in the local market, which makes people wonder about its core competitiveness.
Finally, it is the difficulty of obtaining traffic. In the past few years, an industry misunderstanding was that as long as you can connect to the Internet, there will be a large amount of traffic dividends to obtain.
But now, online traffic has reached its peak, and the cost of acquiring customers is extremely high. Some companies even go offline to acquire traffic, and the cost of acquiring customers is even lower. Therefore, this is also a key factor that determines the quality of traffic pools in the future, and this is also the reason why Pinduoduo is favored by investment institutions in the primary market in the US.
does not consider its business ethical flaws, and the traffic injected by Tencent is definitely the key to its success. According to the prospectus, Pinduoduo's customer acquisition cost in 2017 was only 11 yuan per person. In the e-commerce industry, it is unimaginable, and even lower than the traffic majors JD.com and Alibaba.
4. Metamorphosis evaluation and analysis
The evaluation of monetization is not just about evaluating user assets, so we insist on limiting it to a limited range, that is, how much monetization is possible in the traffic pool.
's existing monetization power is mainly carried out from three aspects: paid rate, paid user single customer income, and user life cycle.
where
MAU·Payment rate·ARPPU=MAU·ARPU
Equation is the common calculation method used on the right side. However, if the payment rate is not observed, the observation of user conversion fundamentals will be blurred. The sexy level of a paid user project with only the spire of the tower and a paid user-wide project is obviously different. The ideal state of
is: "User classification and grading, payment from high to low", and achieving the state of "first-level price discrimination" in economics. To put it bluntly, it means that users who can pay will pay.
In addition to this, the user life cycle is also a very critical indicator, and it is also an indicator that most Internet companies are unwilling to disclose. But unfortunately, except for the rare projects, most product users have a certain life cycle. The life cycle of
users is too short, and another round of "new recruitment" and "retention" is needed to maintain the scale of the traffic pool, which is extremely cost-saving for enterprises. More importantly, this will cause the enterprise to monetize the space of the traffic pool to be squeezed. At some point, this forces companies to harvest quickly, further causing user satisfaction to decline, shorten their life cycle, and enter a vicious cycle.
In fact, no matter what type of project it is, the user's life cycle can be relatively controlled. Taking the game industry as an example, it is logical that the life cycle of a game is very short, some even less than one year, but Youzu Network's star product "Young Three Kingdoms" has been prosperous since its launch in 2015. The secret lies in the extreme attention to user experience.
It should be noted that I did not place "retention" in the "traffic pool" for evaluation, considering that it would cross the metrics of active users.
The above current status indicators can be clearly observed, and for future space, it needs to be calculated through the logic of the business model.
Generally speaking, there are three types of charging models for Internet projects:
- One of them is advertising revenue. By accumulating traffic, the project becomes the place of advertising or traffic distributor.
- The second is to trade commissions. This includes the platform charging a certain pass fee and entry fee through matching transactions; it also includes the situation where the self-operated model of JD.com earns trade differences.
- The third is service charges. This includes the service fees earned from both ends of the transaction using SaaS, PaaS and other means to improve transaction efficiency.
From the traditional perspective of the Internet industry, the valuation of SaaS-type projects is not high, but this is just a prejudice from yesterday. In the United States, Salesforce, the originator of SaaS, has approached a valuation of 100 billion US dollars, and the future market is optimistic. The reason is simple. When a large number of companies pour into the battlefield of the first two types of charging, service charges have become a new blue ocean.
In fact, SaaS, PaaS and other services are the data capabilities of enterprises, and this may be a huge space for Internet projects to increase value in the future, and several unicorn-level and even BAT-level "head projects" may emerge. After a short-term traffic dividend, the data dividend may be the foundation of the everlasting growth of the Internet business industry (this part will be highlighted in the subsequent articles to measure the "conversion ability" of Internet business projects).
Overall, in the above three fields, any of them will have a future if they do well.
The three types of charging space will affect the company's profit and loss statement, while another indicator will affect the company's balance sheet, that is, the merger and acquisition value of the project. If a project has strategic value to the investment entity, it will inevitably obtain a higher valuation premium. The premium here comes from the synergy effect of 1+12 that the merger and acquisition entities are optimistic about. A typical example of
is that Meituan acquired Mobike on April 3 this year. This acquisition is not only an inevitable development path for Meituan, but also a consideration for fighting against strong enemies such as Alibaba and Didi.
Alibaba's new retail strategy once proposed the concept of "three-kilometer ideal life circle", aiming to meet users' "eating, drinking, playing, having fun, wearing, living, traveling, and even education and medical services" within the three-kilometer life circle through Internet technology. The day before the acquisition (April 2), Alibaba acquired Ele.me with US$9.5 billion. Obviously, Alibaba hopes to use it to supplement its short-distance transportation capabilities, launch new takeaway business formats, achieve a three-kilometer and a half-hour delivery, and provide users with more abundant solutions.
When Alibaba continues to strengthen offline, Meituan chooses to acquire Mobike to include the shared travel and life scenarios in the three-kilometer living circle and further improve its valuation. Looking further, through the acquisition of Mobike, it can not only expand the traffic pool, but also seamlessly link this type of travel scenario with the newly opened online ride-hailing business, expand more of its business in the travel industry, and increase the bargaining chips to confront Didi.
Here, although in the eyes of many industry insiders, the acquisition price is not amazing, considering the operating conditions of projects like Mobike, this kind of acquisition is also a blessing. Overall, having a good position is enough to attract acquisitions from giants, and it is also a manifestation of monetization.
But it has to be said that Internet business projects should build their own "monetized traffic pool" from the beginning. This means: to consolidate the basis of advertising revenue, transaction commissions, and service charges, which is the true reflection of the value of user assets.
Remember that projects that don’t make money are not good projects. Otherwise, if you blindly aim at capital (institutional capital VC or corporate capital CVC) and wait for takeover, the business model itself will lose its roots.
# columnist #
Mu Sheng, WeChat public account: Mu Sheng Firm (ID: hrm-yun), everyone is a product manager columnist. Well-known management scholar, postdoctoral fellow in business administration, doctoral fellow in management, founder of Musheng Enterprise Management Consulting Firm, and an expert in research on Internet business model and organizational transformation.
This article was originally published on Everyone is a product manager. Reproduction of
The title picture comes from the genuine library Tuchuang Creative
