5 things you need to think about before building an Internet platform!

People who have been in the Internet circle for some time will often hear a saying: "We want to build an XX platform." This sentence is both fashionable and powerful, and is much better than those who say "I want to create a public account". But every time I have in-depth discussions with friends who are "building platforms", I find that these business experts often don't understand what "building an Internet platform" actually means.

▍The Internet platform is essentially a collective contract

Generally speaking, there must be suppliers and demanders in a business. In recent decades, if you want to do a business of a certain scale, most suppliers have chosen to set up a company, manage people through contractual relationships, have people manage productivity and means of production, and then assemble these things into products and finally sell them to the demand side.

In a theoretical market economy environment, if all suppliers in a society are freelancers and all demanders can obtain sufficient information to find these freelancers, then we will not need any companies. Of course, this is just a theoretical state, just like the "spherical chicken in a vacuum", and information asymmetry is still a common phenomenon. But with the development of information technology, more efficient production models have emerged.

This model has many names: P2P, O2O , C2C, sharing economy - what they have in common is that they use some technical means to directly connect the information of suppliers and demanders. For example, Airbnb connects landlords and renters through the Internet, and Uber connects idle drivers and car users through its app. We call entities that exist in these models Internet platforms - they use technological means to solve the problem of information asymmetry between suppliers and demanders, connect them, and create value in the process.

As mentioned earlier, in a company, the production process is managed through labor contracts, or employment contracts. It is relatively easy to establish and maintain an employment contract because the process is essentially buying "normal operation" with money. If you get paid by a company, you have to work according to that company's rules. However, if you need to spend money to purchase the rules, then as the business grows, the cost of maintaining the rules will become higher and higher.

Correspondingly, the advantage of Internet platforms is that they often do not own the means of production and do not need to hire labor to manage these means of production. For example, Uber does not own any cars and does not directly employ drivers; Douyu Live does not need to purchase computers for every live broadcast, nor does it need to establish an employment relationship with the anchors. Not owning the means of production means that there is no need to directly maintain the means of production, so visible costs are greatly reduced.

But Internet platforms need to protect those "who maintain the means of production." Due to the lack of a binding employment contract, the task of the Internet platform is to formulate rules that are acceptable to both suppliers and demanders, and then use some means to make them abide by these rules. This is an invisible, collective contract.

▍Is this market suitable for a platform?

People often say that no one has built a platform in a certain market, or no one has built it yet, so if I do it, there will definitely be a chance! Unfortunately, this reasoning is not reliable, because if there is no one building a platform in a market, it is possible that it is really a blue ocean, but it is more likely that this market does not really exist.

If we want to correctly understand the reason why "no one is building a platform in a certain market", there are several key words to consider: demand, frequency, standards and profit. If we only focus on one of these points, we will make the wrong inferences as above.

We need to study whether there is a certain product or service in this market that has rigid demand, whether the frequency of this demand is high, whether the transaction process has been standardized, and whether the profit margin for each transaction is large enough.

If there is no rigid demand, then it is very likely that the platform will be built but no one will use it; if the transaction frequency is not high enough, then the scale of the entire market will probably not be large; if the transaction process has not formed standards, then the cost of educating users will be relatively high; if the profit margin of the transaction is not large, then it will be difficult for the platform to get a share of the low-profit business.

When we believe that a market does exist, we need to find a way to determine whether we can support an Internet platform under existing conditions. A very important criterion for judgment is operating cost, that is, the cost of establishing and maintaining rules. If its cost is higher than the cost of directly hiring labor, then there is not much point in building an Internet platform.

Generally speaking, if an industry has become highly specialized and standardized, the chances of building an Internet platform are much greater. The rules for renting houses are relatively mature (there are market prices, payment standards, and grades are differentiated by location and luxury), and the rules for using cars are also relatively mature (priced by time and by distance, and grades are differentiated by car model), so relatively successful Internet platforms have emerged first in these fields.

However, if an industry has not reached a high level of specialization, the cost of building an Internet platform will be relatively high. For example, equity crowdfunding, which was popular in the past two years, has not yet shown any success. Each company is currently in the stage of educating users, but educating users is a long process: first, users must accept the general rules of this field (which are only mastered by a few professionals), and then find ways to make users accept the special rules of their own platform - the most important thing is that these things have to be done for both the supply side and the demand side.

Why are highly specialized and standardized markets more likely to produce Internet platforms? Because you can borrow existing rules, making it easier for target users to accept them. The human brain works by using existing things to explain new things. If we want to describe a wolf to someone who has never seen one, we will most likely say "it looks like a dog, but it eats people" instead of starting with details such as fur color, body shape, and habits. It is easier for people to accept new rules by improving the old rules than to build a set of rules out of thin air.

In addition, when considering whether you are competitive enough in this industry, you can also think about this: Do you have enough resources on either the supply or demand side?

Judging whether an industry is suitable for an Internet platform is a long-term process, because we live in an era where new situations occur at any time. "Entering the market at an inappropriate time" and "Not entering the market at the right time" are both failures.

▍A good solution should shorten the chain

Once you have a clear understanding of the market/industry, you need to estimate the specific costs. Before estimating, you must first have a clear solution. It is a simple truth that a good solution should be able to shorten the chain. The success of the entire business requires that every link in the chain can operate normally; the shorter the chain, the greater the chance of success. However, this link often goes wrong in practice because of the influence of the Internet, and we often ignore the chain beyond the technological products.

We know that Zaixing and Fenda are both platforms for online interviews with experts, and their business models are similar: connecting experts and consultants, improving the efficiency of light consultations, and thus creating value; their solutions are different. Zaixing does it through online appointments and offline interviews, while Fenda directly asks questions online and gets answers (essentially a very granular interview).

On the surface, the chain in action looks like this:

However, the actual chain, at a certain stage of operation, may be like this:

Correspondingly, the chain of Fenda is roughly like this:

As the platform operation stage changes, the above chain will also change at any time. The complexity of a solution can easily be underestimated if the complete business process, including operations, is not considered. In fact, in the process of building an Internet platform, the difficulty is often not in developing technical products, but in formulating a mature solution based on the existing business model. Sometimes we will throw this task to the product manager and let him/her think hard and solve it by himself/herself; but many times, this is beyond the ability of a junior product manager.

▍Once you have a solution, estimate the cost in detail

Let’s continue to use Zaixing and Fenda as examples and use hypothetical numbers to estimate the operating cost of “expanding experts” - assuming that expanding a new expert requires an operations staff to make 4 phone calls, each call lasting 5 minutes, plus 20 minutes of written work, a total of 40 minutes, then the maximum workload of each operations staff member per day is to expand 60 × 8 ÷ 40 = 12 experts. Assuming that the success rate of expansion is 70% and the efficient working time of operators is also 70%, then the number becomes 12 × 70% × 70% ≈ 6 experts.

Assuming the operational goal is to bring 5,000 experts online within half a year, then if spread over 180 days, 5,000 ÷ 180 ÷ 6 ≈ 5 full-time operational staff will be needed. Assuming that the labor cost of each operator is RMB 10,000 per month, then the amount to be spent on expanding the number of experts in the next six months is 10,000 × 5 × 6 = RMB 300,000.

With a solution, we can calculate the approximate cost within a certain period of time; with exact numbers, we can know whether we can support such an Internet platform.

Switching from "Zaixing" to "Fenda", and from offline interviews to answering questions online, is also an effort to expand the work of experts, and the results of platform staff per unit time can be greatly improved. For example, the answerer of Fenda only needs a very brief introduction, so there is much less paperwork; the settlement of Fenda is completed when the question is asked, so there is no need for a separate transaction completion confirmation and settlement link; the answerer of Fenda no longer needs to come out and sit for an hour, so there is no need to provide cash subsidies in the early stage...

Interested readers can estimate for themselves what the approximate cost is in the entire process of Xing and Fenda to achieve the same operational goals.

For the same business model, applying different solutions will result in significantly different operating costs, and the same amount of money will produce completely different results.

▍Generating user value can sustain platform development

This may seem like a simple statement, but it is actually the core competitiveness of the Internet platform. User value and user experience are often confused. Many people tend to think that creating user value means making good technical products, but if you think about it carefully, you will find that this proposition is not true. Where real user value is created lies in how to deconstruct the industry chain to come up with a solution that is better than the existing process, thereby improving the efficiency of industry participants.

Therefore, one of the key points in building an Internet platform is to find quantifiable key indicators of user efficiency and then find ways to improve them. For example, the key indicator of user efficiency for a taxi-hailing platform may be the time it takes to call and wait for a taxi. So the way they generate user value (at least in the early stages) is to expand the supply of taxis. For the equity crowdfunding platform mentioned above, the key indicator of user efficiency may be the return on investment per unit time, so we need to find good projects and convince them to choose equity crowdfunding for financing.

In the past two years, there has been a common saying that "making apps is no longer a good idea". The popular explanation is that super apps like WeChat have taken up most of users' time. Another reason is that many people have figured out that generating user value does not necessarily need to be solved through an app.

Many startups choose to release their products on WeChat first, verify the business processes, and then develop more complex apps with better experiences. The "product first" approach does not refer to a simple technical product, but to the entire business/service process that generates user value.

▍Improving matching efficiency is the best way to increase user value

Generating user value does not necessarily mean making a good technical product, but this proposition may be true at a certain stage. A typical example is the catering O2O platform, especially the food delivery platform, which is at this stage. For example, Baidu Takeout spends a lot of effort on developing delivery algorithms. Since both the supply side (restaurants) and the demand side (diners) are relatively sufficient, improving the matching efficiency becomes the best way to increase user value.

But in essence, in a business's production → dissemination → purchase → consumption chain, complex technology products represented by apps often meet consumer needs, that is, the scenarios for use after purchasing a product. For example, I hope to be able to seamlessly fast forward and rewind, speed up and slow down when playing audio, or I hope to use a smoother native app to read e-books. At these times, the intervention of the app may be required.