Easy Interchange Shangan: Application and practice of data in multi-layer programmed marketing

2014 Zhongguancun Large Data day on December 11, 2014 in Zhongguancun, the General Assembly to "aggregate data assets, promote industrial innovation" as the theme, to explore data asset management and transformation, large data depth technology and industry data application innovation and ecological system construction and so on key issues. The Conference also carries on the question of the demand and practice of the departments in charge of the government, the finance, the operators and so on to realize the path of transformation and industry innovation through the management and operation of data assets.

In the afternoon of the operator @big Data Forum, from easy to exchange Shangan Mr. Bring keynote speech, profiling data in the multi-level program marketing application and practice.

Shangan: Hello everyone, very fortunate to be able to receive an invitation today as a partner of the Asia letter to share with you. Leisurely easy to exchange is a kind of company, first give us a brief introduction and a word summary, we use data, through the way of programmed to help advertisers do marketing companies.

Refers to marketing, a word of summary marketing is four aspects, one aspect is the technical advertisement plan, the second aspect carries on the advertisement purchase, the third aspect carries on the advertisement optimization, the fourth aspect is carries on the advertisement appraisal. These four aspects have been in the way of the past 20 years throughout the traditional purchase mode to the digital purchase way throughout. The past purchase to the evaluation can be said to be patted on the forehead to do, for example, I just arrived at easy to exchange, the circle in the face of like leisurely and easy to do brand advertising company has a view, you are mainly to fight, rely on a single, than return point, who can get more customers. But this was four years ago. But with the addition of data, this has changed dramatically. For example, I have just entered the leisurely exchange of our advertisers talk to us about the plan, will tell us how many CPM ads to make how much money, or a few clicks, we know this is very difficult to do, but because we do not have data to prove, so all this is not comparable. So we see 2013 began to today a full two years later, more and more large data operators of data to join our advertising industry, so that the industry to buy optimization, can be measurable effect. How does this work? I will give you an introduction.

First of all I come to easy interoperability after the first Data bank products, what is called Data Bank? As the name implies, each advertiser has its own first data, this data can not be opened with other data sources. The most important thing in the last week is the integration, and we have a lot of data companies outside, like our next lecture, Admaster is also a typical data provider, and there are many more companies that provide information about the data on human manipulation, providing a lot of company to browse for interest, We can actually combine all of these data sources across the industry, and we'll combine the sex data, age data, and advertisers ' first data. We had an ad before the Lord is a doctor of the hidden glasses, perhaps everyone has worn Dr. Lun recessive glasses, but who is the audience we may be difficult to pat the head to say. We advertisers and we make plans, he said he cast a Japanese-style hidden glasses, said you give us 25 years old to 30 years old female users, this How do we position? We first collect the data on the customer's official website who also through Baidu or Sogou 36 search your site of these people, we put his hidden glasses advertising channel All these people gathered in, do a management, we put the first party and third party data source to do a comparison to us to get a conclusion, You can look at the bottom right corner of the page, we see that the yellow curve is a sample of the natural internet people, the blue is the first group of advertisers, when we integrate the first party and third party people, it is true that the 25 to 30 years old people to the proportion of hidden glasses is heavier, Then the month is 5000 yuan more people use the day to throw the glasses a higher proportion, at the same time 5000 yuan monthly salary below the recessive glasses will use low rate. We will feel that this ratio is particularly high, is the mother and child type of label. Our company marketing Director 3-year-old mother, I said you hold children often wear hidden glasses? She said I wore hidden glasses when I went out with a baby, which is a better combination of the first party data and the third party data.

We use this data in the launch process, it can be understood that we will have a lot of first-party people, and we will have a lot of third-party data sources, such as Asiainfo is also a data source, when judged its advertisers the first crowd is strong mother and child, makeup, 25 to 30 years old female proportion of the crowd characteristics, We can use all the third party data to expand the population, which means that the industry terminology we call is the people who look for the RKLT, we collect the concentrated people into the programmed platform, and then the place where the technology is embodied, we have a lot of data sources, Not every data source is capable of generating value by the final result. For example, he is black-haired, because today we all have black hair, no yellow hair or blonde hair. When we spread this, we found that the effect might not help. Because in fact, easy to exchange in this area of technology can be put into the data, and the results of data collection into our system, and then form a feedback, and then we know that we use which data labels for us is the most effective.

This is the figure behind this, even we can do data collection, which data people he used to have a further step in the behavior. For example, Dr. Lun, we let him go to the Internet to participate in the game, he will be limited to the site, enter the verification code, and then obtain the verification code, and then apply for success. The more urgent it is to prove their demand for the doctor's glasses or the need for the event. In fact, there is a collection method is to add our code on the official website, to collect each user to visit the customer site after the requirements of what. All of us are inseparable from the big data, and we see how these people behave differently. Take the example of Dr. Roentgen, on the first floor of the official website, their users did not have an obvious sulky between the ages of 25 and 35, but more to the end of the marketing funnel, but just as we said 25 to 35 years old, more than 5000 of the monthly income will use the characteristics of Dr. Lun.

Another change is that we find that the trend across the screen has become possible, such as the time you spend on mobile phones and 2013 than the changes, may be some people are not very good means, there are people who have a bad habit, in the car will use the mobile phone. Advertisers do marketing when he does not want the PC, this how to achieve? We've found some precise crowds that we can do to collect, we call the way to cross screen ID, for example, we and Baidu as our cooperation platform, they will embed our cloud disk, this time your mobile phone and computer, computer is a cookie, such as the Apple phone is if, this can be synchronized. There is another one, for example, we are here today, I do not know if there is a connection to WiFi today, or the office is connected to WiFi, we also visit the IP address, we judge you for a long time, the office's IP export and your home export, You can capture in several cases that you have two devices on the Internet, we may consider you an accurate customer. So we embrace big data, want more data, and hope that the results of integration are more and more, so we are very welcome, especially the Asia letter just mentioned in the unified ID can provide more support in the industry.

Then we can get a very interesting thing for advertisers through the cross screen ID, so that we could simply share it, for example, when we found out that when we were doing marketing on the mobile side, it was not very good with the PC. For example, Mobile has a lbs information, there is a time-oriented active cycle with the PC is the opposite, and then we locate the phone model, and then look at the network orientation, including for individual applications of some orientation, this is our service advertisers very good way. But for information about sex, age, and occupation, we feel that there is a lot of exposure on the PC and very little exposure on the PC. You may be a student at school or a white-collar worker, you use a very similar characteristics and expertise, we will find in the PC side often exposed to you often surf the Internet when you browse some of the interests of browsing some of the family situation, at the mobile end can not find, so we have more across the screen some ID here to get through, We will be able to more specifically outline the whole of each individual advertising behavior on the Internet, thus for advertisers to do better marketing services.

So with more PC-side data, more moving data and more data across the screen, we can put all of a person's entire interest, such as his age, gender, income, including all preferences, in the mobile phase of marketing. For example, we'll find that the preferences shown on the mobile end we will find that the mobile end on the mobile phone to read the article, see the Circle of friends articles, watch the news, look at some other aggregate class information, we will find that in fact, few people in the car such products will use the mobile phone, because the car article has a feature, It's usually a very dense picture, at the same time the article may be very long, will introduce car characteristics and so on, this is our mobile phone whether it is from the speed or pictures and so on will find characteristics, so we will find the blue area in the car on the PC side, so we will be aware of everyone to do very good precision. So for everyone we have a precise understanding of what kind of product you are suitable for, or in turn we know for advertisers what kind of person is your audience. The other is that we can send you the right information at the right time in the right place, which is the purchase area. This is an area where you can easily communicate with one of our services. So when we have more data, we can find our audience more subtly for our advertisers.

At the same time we have from the PC to the middle of the mobile end of the crowd through, including from the mobile end of what I saw, The advertiser's material, I point in after, to customer ldpage, this data and customer data basically through, we from the plan to the ad purchase to follow-up browsing behavior, To the final results the feedback forms a complete data loop in the data feedback, thereby better correcting the correctness of the logical model that we are describing to the crowd, including the purchase.

We have just talked about some of the methodologies we use in our advertising campaigns, which are some of our clients and some of the awards we have received from our company. So in general, in our view as partners, we are very welcome to have more large data to join the entire biosphere, because the use of large data there is a word, we do not require all the data its label is unified, we would like the entire range of data, or the content of the data will be richer, the better, At the same time we want to let the data itself emit sound, just as we used a lot of data labels to make our doctor-lun invisible glasses crowd ourselves, like the crowd, we welcome more people to embrace the big data. Thank you!

(Responsible editor: Mengyishan)