user Traffic (Traffic), the user viscosity (Traction), user scale Growth (Growth), we all know that these three words is a prerequisite for success. When we had just established a company, the first is to seek our customers, this will attract user traffic, keep the user viscosity, promote the user scale growth (hereinafter referred to as the TTG) these three aspects. In some cases, we are willing to pay for attracting user traffic. In most other cases, we worked day and night in order to let the TTG this development model organically.
when we read the co-founder of success, such as Yelp, Pinterest or WhatsApp, about their story, we may find themselves shocked by their operating ability and wisdom, but we often forget our true should know – how the startup will TTG perfectly combined with the work. Of course I to success for any how a startup method is to don’t have any prejudice, but as a founder of the theory analysis model, described in the “from good to success” approach is unable to meet us. We need more mathematical formula in the form of advice.
this subject in order to better understand business infrastructure is operations research (operations research).
the principle of operational research can not only guide our optimization of enterprise and make it going smoothly, also can through the modeling and simulation provides the foundation for our business statistical analysis. One of the most interesting theory in operational research, as well as provide relevant for today’s business guidance of an application, called “theory of queue” (). In queuing theory, the “little law” is a hidden gem, it is how to keep attracting user traffic, user viscosity, and promote the user scale growth, offers an excellent combination operation, gives us profound influence.
here is in a store, the use of queuing theory is one of the most shallow layer to solve the problem:
to – & gt; Service – & gt; Leave
in a queuing system, the project will arrive at a certain rate system, and then will leave. Referred to in this project can be customers also can be in stock. Let’s think about it carefully, this is as we can see in the middle of the web site or application: a user arrived, with the application or shopping for a site for a period of time, then they will leave. One of the most valuable company is users stay the longest time.
, pointed out that in a steady state system, is equal to the average number of items in the system, project arrived at times at an average rate of a project in the system the average time spent.
L=the average number of items in the system;
W=a project in the system average waiting time required;
lambda=project to reach the average number of per unit time;
the rules are as follows:
“stable system, the customer’s long-term effective long-term and effective mean arrival rate multiplied by the average number is equal to unit customer the time spent in the shop.”
this sentence sounds simple, however, it’s magic lies in its simplicity. It is not subject to distribution services, service order, or other what influence, the color of the site, distribution of content, the price of the product can affect it. It is only care about the client will come and how long it will stay at what speed, everything else is secondary. Litle law applies not only to store the queuing theory, it applies to the website or other systems of current project.
we take a real case to verify. Google as a search engine, with the highest user hit rate, is the lambda. But the user did not stay for a long time, they quickly click on search results link to another site, and then after they come back search another site and leave at once. Google is doing amazing work to have such a high click-through rates, is also for this reason, contributed to today’s Google. But we’ll look at Google’s acquisition, product development or any other plan on tall, you can easily find the laws of their target is a litle part 2: W, namely each customer in Google spend an average time, email, phone, calendar or Chrome.
according to Comscore, Google in March 2014, received 13 billion search, this is equivalent to 433.3 million times a day, each hour 18 million times, 300000 times per minute, only about 5000 times per second. Compared with those in the following table bing:
we can assume that if bing at any point in time than the Google 5000 times a second, if so, it will be good for bing over Google, find out the key to its reason, is the law of the second law, compared to observe the daily or monthly data, survey is higher than usual as the outbreak of short-term traffic can give us more profound reminder.
now take a look at Facebook. Facebook have great views and at the same time in the “store” the amount of time. But its customers hit rate (lambda) without Google is so high, that’s why it’s acquisition and product are all in order to improve the hit rate. We visit Facebook several times a day and stay for a period of time, but we’ll jump to Google to search.
operating managers and entrepreneurs rather than hit rate is more concerned with productivity. But only the user access, productivity is important, not is a very useful dual function, of course, once the customer come, this key indicator monitoring is how quickly the customer to arrive, rather than how many customers.
here is a litle principle of three law, also applies to startups:
1. The startup to the assessment of investors, it is best to check the lowest level of traffic data, even if low monthly unique visitors, but in the small time interval surge of traffic can provide you with more valuable clues to asses