Currently, many AdMaster customers (Coca-Cola, KFC, Haier appliances) are facing problems on Internet marketing. They spend a lot of resources and time on new marketing methods, such as social network marketing. However, they often do not receive the expected marketing results. The main reason for that is that they are not able to understand the influence of these marketing activities in real-time among consumers. Especially, the consumer’s attitude towards the brand compared to competitive products is difficult for them to understand. More importantly, faced with huge amounts of data, AdMaster’s customer tends to analyze marketing campaigns by extracting only a small part of data. The one-sided data cannot bring accurate, comprehensive and valuable analysis and therefore it cannot be applied to marketing activities.
2. Solutions and Challenges
Based on these customer needs, we proposed a complete solution: build a SAAS platform, provide personalized social analysis and consulting for different customers. At the same time, the platform should be equipped with massive data processing capabilities, powerful text processing capabilities, real-time analysis and early warning capabilities.
Personalized Consulting Solutions and Massive Data Processing Capabilities
“Our customers are from different industry backgrounds with various internal and external conditions. Based on these differences, we understand that their custom analysis programs also need to be considered differently. Therefore, in this SAAS platform, we provide the ability to select and expand modules in order to meet the analytical needs from different enterprise customers. At the same time, the platform has mass data processing capability in order to ensure an accurate and reliable analysis.” [Insert where you got that quote from]
Customers want to know the user’s comments and evaluation of the brand as fast as possible. Based on this demand, we need a platform that reacts quickly on user actions, using semantic analysis. The platform must contain a mass of unstructured data, such as marketing-related fan comments, blogging content, and the ability to quickly present them our customers through text mining and visualization techniques.
Whether it is analyzing the effect of advertising, industry trends, or dynamic consumer groups, customers want it to take place “fast”. Therefore, the platform must have powerful performance to be able to deal with data analysis and present real-time data results to the client.
However, those needs brought three major problems along:
- vast amounts of unstructured text data
- enormous costs for building a platform with different architectures that combine customization, multi-language, and globalization
- demand for processing capabilities of the base when doing real-time data analysis
“Facing these challenges, we chose SAP HANA as part of our underlying architecture. Using SAP HANA’s powerful computing analysis capability, we were able to achieve emotion word analysis, text analysis, and semantic analysis, and solve other problems. In addition, we are able to make complex queries on massive, rapidly changing real-time information with SAP HANA, without spending extra time. We can directly react to the native data. Even more importantly, based on HANA, we are able to plan the future with multi-language. We do not need to change the architecture even when we face an international challenge. We can easily expand to a global solution. This will save a lot of costs for this customized platform, especially in the future.” [Insert name of person who said that]
3. Product Architecture
As shown in the figure below, according to the needs of our customers, we can use open API and crawlers to collect data, and save cleaned data in Hadoop clusters. Subsequently, we import the data into HANA, use text clustering, sentiment analysis and text classification to generate results back into the front end through data visualization. That way, we are able to show our customers the functional modules they want, and intuitive data analysis.
4. Product Introduction
Combining our clients Coca-Cola case’s experience
Multi-angle Marketing Effectiveness Analysis
We use this platform to help Coca-Cola build a flexible and convenient Dashboard and get rid of the shackles of traditional fixed rigid statements. Customers can combine a variety of dimensions, such as clicking exposures, user sentiment analysis results, the distribution of hot words, geographic distribution, *** ratio and others to DIY the analysis reports they demand.
Real-Time Early Warning
To help Coca-Cola do real-time warning, we use the semantic text analysis technology to set some sensitive words such as: fake, yield and so on. While monitoring the whole network, those sensitive words or similar words can be discovered. As they are discovered, we promptly tag the content. For example, if someone sends a question about the taste of Coca-Cola, we are able to quickly analyze semantic microblogging content. We can mark this message as negative news and rapid feedback to Coca-Cola’s marketing department.
User Population Attribute Analysis
Monitoring blogging content for Coca-Cola, we focus on many angles such as the crowd, purpose and common user behavior to do customer analysis. This enables the Coca-Cola ads to connect with their target user groups. At the same time, we analyze who are the most efficient, influential opinion leaders for Coca-Cola through the mass media. They are supposed to adapt to the opinion of the general audience, what allows marketing to be more targeted and efficient.
Performance Evaluation of third-party Marketing Agency
Many of our customers will outsource their internet marketing to a third party, though they are unable to measure the effect of third-party’s work better. In our platform, there is a task performance module. With that module our customers can clearly know the ID of each staff member of a third party, task type, number of processing tasks, number of completed tasks, number of responses, completion rates, response rates and average processing time.
5. Business Value
“Based on the real-time community-based consulting platform, customers are able to better understand how to use the social networking platform to obtain greater brand awareness and consumer intentions. For customers, such as Coca-Cola, we customize a complete marketing monitoring program. With this process, we helped Coca-Cola monitor 424.291 marketing related topics, 1.675.325 redirections, 3.370.947.542 exposures and other data. Based on this data, we adopted the text mining, microblogging monitoring, listening to public opinion and other personalized programs. That way we greatly increased the value of our marketing effort. According to a survey, prior to that activity, about 7% of our consumers planned on buying Coca-Cola products; afterwards, purchase intentions reached 25%, which is a strongly increased number compared to the brand’s other marketing programs.” [insert name of a person we can quote]
On top of that, real-timeearly warning analysis can share the consumer’s experience and feedback timely and due to that enhance brand reputation. We use the SCC (social master) System to monitor Windows and provide early warnings. With user experience issued data collected from windows8 and weekly meetings with Microsoft Windows Products team, we help the team to identify main sources of information. Those information should support them in locating where the problem is. We are also able to prioritize different issues (compatibility, update installation issues) for them. In our data support, Microsoft is able to enhance their favorable rate by 50%, reducing assessment rate by 15%.