More than a decade ago, Clive Humby, a British mathematician, said, “Data is the oil of the new age, and it holds great value.” But at the same time, he notes that “data is not valuable if it is not refined and analyzed.” We must process and “smelt” the data crude oil, in order to fully extract and release the value of the data, and provide a constant power for the digital innovation of the business.
As a data intelligent enterprise, Daily Interactive relies on advanced big data algorithms, knowledge mining, deep learning and other technologies. Over the years, Daily Interactive has been continuously “managing number” and “refining number”, and has accumulated profound experience in data processing and data value extraction.
While constantly strengthening data capabilities and providing professional data intelligence solutions for vertical fields such as mobile Internet, digital marketing and smart city, the company interactively sorts out and summarizes its own valuable experience in data extraction and data insight every day, and innovatively puts forward the “five-step method of data intelligence”. It aims to provide mature ideas and effective solutions for industry customers to fully mine and efficiently utilize data value.
Five-step method of data intelligence to create closed loop of data value “smelting”
The five steps in the five-step method of data intelligence are “result sampling”, “annotation review”, “similar expansion”, “practical application” and “feedback attribution”. Through these five steps, the processing and extraction of massive data resources can be completed through daily interaction, a rich three-dimensional label system can be constructed, the humanistic meaning behind the data can be interpreted, and the driving effect of data on business growth can be truly brought into play.
Step1 Sample the results
The first step in the five-step method of data intelligence, “result sampling,” refers to receiving data samples, compiling and analyzing the data. The sample data here includes not only the one-party data from the industry customers, i.e. the “Y value”, but also the rich characteristic data of individual data and other third-party data, i.e. the “X value”. Daily Interactive believes that “Y value” plays an irreplaceable role in guiding and leading us to produce accurate and effective data analysis results. However, it is not enough to have only one side of the data. Enterprises also need the support of larger volumes and more dimensions of data to develop comprehensive and deep insights.
For example, in the field of digital marketing, brand owners want to understand consumers in an all-round way, so they not only need to rely on the brand’s own data such as membership data, transaction data and advertising data, but also need to use third-party data to detail the characteristics of consumers from multiple perspectives. In the field of mobile Internet, APP also needs to fully integrate users’ in-app data with out-of-app interest and scene preference data, so as to establish a more complete user portrait system and support intelligent operation.
As a data intelligent enterprise, Daily Interactive has built thousands of portrait labels by virtue of massive online and offline data accumulation, which can provide rich third-party data support for industry customers to improve data assets and carry out in-depth and high-quality data analysis.
Step2 Label carefully
The second step in the Five-Step Method of Data Intelligence, “annotating and examining”, refers to carefully examining and observing the data samples, summarizing the correlation features and internal laws from the numerous data, and uncovering the deep humanistic meaning behind the objective data, so as to provide scientific basis and effective support for intelligent decision-making. This requires not only efficient technical means, but also professional and deep insights from people.
Daily interactions, for example, to help the relevant government units, the analysis of the urban population on the one hand, using technology of data such as time, space, population of multi-dimensional data series and get through, characteristics and algorithm such as population flows model by building a crowd, portraits of urban population, population structure, population distribution and migration trend analysis and forecasting; On the other hand, urban planners are also required to make a comprehensive study and judgment based on the current economic, social and industrial development of the city, mark and interpret the humanistic meaning of various data indicators from a professional perspective, so as to provide scientific decision-making reference for urban development and social management.
Similarly, in the fields of digital marketing and mobile Internet operation, Daily Interactive also emphasizes that data analysts should combine the experience and knowledge of the industry to manually review and continuously adjust the sample data and algorithmic model, so as to ensure that they will not be misled by abnormal and invalid data and truly see the essence through the data accurately.
Step3 Similar expansion
The third step of “similar expansion” in the Five-Step Method of Data Intelligence refers to finding more potential groups with similar characteristics according to the characteristics and models obtained from the insight analysis of data samples in the second step, so as to realize the scene requirements of accurate customer expansion, similar crowd mining and targeted promotion.
Just as every living organism has its own specific biological genes, Daily Interactive believes that every user in the Internet age also has its own internal data genes, which are attributes, preferences and interest characteristics that users display in their online and offline behaviors. In the second step of “labeling and inspection”, the problem to be solved is to find the data genes of sample users, and extract and construct specific labels and models to represent them. The third step, “similar enlargement”, is to accurately screen out groups with similar data genes with sample users from a large number of users.
In this respect, Daily Interactive has a large amount of terminal data, covering the whole network of users, and has established a rich three-dimensional label system. With the help of lookalike and other expansion algorithms, it can effectively help industry customers find a large number of potential target groups in the wide network of users through similar features.
Step4 Practical application
The fourth step in the “Five-Step Method of Data Intelligence”, “Practical Application”, refers to the application of the results of data analysis and the accurate potential customer groups found in the actual business scenarios such as brand marketing and user operation.
For example, in the field of brand marketing, brand owners can screen and match the characteristics of media traffic according to the insight results of consumers. On the premise of satisfying the delivery order, they can select accurate and high-quality groups and give priority to reach them, so as to deliver advertisements to the right people, so as to reduce ineffective waste and improve the delivery efficiency of advertisements. In terms of user operation, APP can group and group users according to label characteristics, and formulate differentiated operation strategies to promote more sophisticated and intelligent operation.
Step5 Feedback attribution
It is worth mentioning that whether the brand owner carries out advertising, or the APP carries out user operation, all kinds of effect data such as advertisement click, user retention, transaction conversion will be generated. The Daily Interactive considers these effects data to be of great value. It is necessary for enterprises to recirculate the effect data generated in the practical application and carry out in-depth summary and attribution analysis to help evaluate and optimize the existing data insight results and algorithm models.
That is, the fifth step of “Five-Step Data Intelligence Method”, “Feedback Attribution”, emphasizes that through the full backtracking analysis of the data of the whole process and link of the business, a “refining” closed loop of constant iteration and evolution is constructed to fully stimulate and release the positive driving value of the data to the business.
Five-step method of data intelligence, enabling the digital upgrading of the industry
At present, Daily Interactive has deeply integrated the “five-step method of data intelligence” into the company’s data intelligence solutions and products, enhancing the efficiency and energy of digital innovation in vertical industries.
The five-step “based on” data intelligence, daily interaction in the field of brand marketing, help from mother to child, beauty makeup, fast, brand integration and the clients through the whole domain data assets, from multidimensional comprehensive insight into the consumer, select high-quality media, make advertising TA concentration of 30% to 50% generally, help customers from the boundless huge crowd to find accurate, To achieve greater marketing effectiveness;
In the field of mobile Internet, daily interaction helps APP to improve the portrait label system, through in-depth data insight and advanced algorithm model, to achieve high-quality new customers and fine operation, to help customers to achieve sustainable growth and efficient realization of traffic value;
In the field of smart city, daily interaction helps governments at all levels to deeply manage data, and helps relevant departments such as planning, transportation, cultural tourism and emergency response to carry out grid analysis and multi-dimensional insight on regional population, spatial flow and industrial development, so as to provide in-depth data support for smart urban management and scientific decision-making.
At the present stage, Daily Interactive is gradually productizing and instrumenting the “Five Steps of Data Intelligence”, and implementing it into the company’s data middle platform product “Daily Numerization Platform” to enable more industrial customers to upgrade their digital process.
Daily for several platforms at its core is the data management and application, not only has powerful data daily interaction and the technical ability to do the backing, plus a company to understand the depth of the vertical industry and expert knowledge, to help customers solve data islands, data quality is not high, difficult to spot problems such as data application, and reduce the threshold of the digital upgrade enterprise and government.
“Five-step Method of Data Intelligence” is a methodology of data application for daily interaction, which has been deposited in the platform of daily data management as an important ability. Through the daily counting platform, customers can quickly complete the convergence, access and fusion of multi-source heterogeneous data, and construct labels and models very conveniently. At the same time, the annotation of the insight results and the evaluation and optimization of the model can also be completed efficiently and intelligently on the daily counting platform.
We believe that with mature methodology and intelligent products, the process of extracting and refining data value will be easier and less effort. In the future, Daily Interactive will continue to output data intelligence practice experience, accelerate technological innovation, polish products, and accelerate the digital innovation of industry customers and government departments.