01 brand live, from barbaric growth to fine operation
2021 is the first year of the brand self-seeding. Douyin, Kuaishou and other platforms have put forward the new concepts of "interest in e-commerce" and "trust in e-commerce". Brand merchants come in response to the policies and traffic. International brands, old domestic products and new domestic products have entered the market one after another.
In the hot e-commerce live broadcast market, many new domestic products with grass marketing, DTC marketing rapidly rise, ride the wind into the broadcast circuit, and those international brand groups with strong brand mentality compete on the stage.
In the past year of 2021, live streaming e-commerce is undoubtedly one of the biggest tuyere at present, and has also brought soil for the savage growth of different types of brands. In this increasingly fierce competition, how should the brand precipitate a unique way to play? In the broadcast room to continue to build brand power at the same time, how to further polish the operation force, more specialized, systematic fine operation?
The direction in which these issues are addressed matters. After all, the era of the extensive development of live streaming may be coming to an end.
02 single broadcast room fine operation in why trapped?
Platform algorithm will be a beauty brand broadcast room blurred into the black box crowd pack; Design strategies for the components to cooperate with the live streaming process; Anchor according to carefully choreographed words, scripts began to burn themselves, introducing product highlights and beauty skills, interspersed with skin care knowledge, with time limit purchase and other inflammatory expression; Potential consumers continue to enter the broadcast room, the high mood and exaggerated actions of anchors affect the whole atmosphere of the broadcast room, but also try to control the rhythm of live broadcast as fine as possible.
After a live broadcast ended, the operation began to measure the performance of the live broadcast according to GPM, ROI and other indicators of the background, while the review of the process mostly stayed on the surface of the existing data. Mainstream live streaming data is often accessed and presented in the background as a "data Kanban", without providing in-depth analysis. In fact, with a "Kanban", it is difficult for brands to find a quantifiable way to explore the combination and performance of people, goods and venues to precipitate a successful model --
You know how a flashy setup for a studio scene or a cat on a table seems to improve the CTR, but lacks valid attribution and data verification;
You know that Hibuy live bungey-dancing brings millions of goods, but you can't know how the look, state and movement range of the anchor imperceptibly affect the stay time and powder conversion rate;
You know the art is king, a set of words fit the brand formula may be able to make the product revitalize, premium ability to rise, but it is difficult to capture the anchor of a word collocation or keyword is in which second clever breakdown of the user's mind, let GVR ride the wind.
Real-time changes in the environment and massive factors in the process of live broadcasting affect the final transaction amount, while the refined operation of the broadcast room requires brands to quickly identify the key variables that can affect the business indicators, and respond quickly to give corresponding strategy adjustment.
We all know the importance of "people, goods and fields". For the "field" operation and polishing of the broadcast room, it is not only about whether the penetration rate is high enough or whether it can bring a steady flow, but also important to analyze the process when it is oriented by the results of business indicators.
And win the broadcast room refined operation, bring the new password of business growth, may be hidden in the single shop broadcast room portrait analysis model.
How to break the Game: Video understanding + NLP + Knowledge Graph
Up to now, the penetration rate of Douyin live streaming e-commerce is less than 20%, and there is still a huge space for increment. However, under the new traffic distribution mode, only by increasing the attention and investment of the brand to the fine operation of live streaming can the long-term operation be established.
Sensing an opportunity, brands have begun to use the algorithm model for digital operation innovation in Douyin single store broadcast room.
Then, how to conduct modeling analysis for Douyin single store live broadcast? Dissect a live broadcast and dissect the key variables that really affect business growth?
Whale, as a wise strategy, launched "DAM Broadcast room Analysis Model". Relying on CV algorithms such as video understanding and knowledge graph and NLP technology, it brought the solution path of Douyin single store broadcast room portrait analysis:
The first step is to enable brands to store each live video stream and corresponding data based on Whale DAM's spatial storage and management capabilities, combined with AI algorithm to process unstructured data. Each live broadcast will be dissected into multiple segments for easy retrieval and rerecording.
The second step is to access and utilize multiple, multiple data sources and information. Such as live broadcast speech script, live broadcast room performance data, unified access of the third party platform index data, etc.
The third step, combined with video understanding, semantic understanding, knowledge graph and other algorithm capabilities, to carry out structural analysis of live video streams. Through correlation analysis and attribution analysis, key variables affecting business indicators (such as powder conversion rate, UV value, etc.) in the process of live broadcasting are mined. For example, it analyzes the correlation between anchor speech speed and volume, expression state, motion range, broadcast content and the change time of service indicators.
The fourth step, release Whale DAM "broadcast room analysis model" intelligent analysis ability. Based on the change of real-time data in the broadcast room and the dynamic change of business indicators, the performance of the anchor and the effect of the speech were evaluated, and the key variables were quantitatively analyzed to influence the final indicators, so as to guide the scientific adjustment of the live broadcast operation plan and the script of the speech, and maximize the effect of the single store live broadcast.
# Knowledge Graph vs. Super talk
It is worth mentioning that Whale DAM is essentially an effective tool for applying knowledge mapping to live streaming. Two of the best-known uses of the knowledge graph are Apple's Siri, which interprets human voice commands semantically and gives intelligent responses; Second, search engines can also find the ontology you really want to search based on fuzzy description. In the current full swing of live broadcast e-commerce, perhaps broadcast analysis will also become a new position to release the magic of knowledge graph.
Through the construction of the knowledge map of the brand's field, combined with the summary of the scene of live broadcast, it can help the brand to effectively polish the successful speech model, and carry out sustainable dynamic optimization of the dialogue model in the practice of live broadcast. While other brands are still honing their language choreography and honing their methods, brands that apply the power of numbers are already pulling out best practice formulas backed by quantitative data through models.
# Analytical model vs. super anchor
From the perspective of anchor training and management, the analysis model based on broadcast room can not only quantify anchor performance and make assessment more reasonable, but also explore the performance mode of super anchor with high ROI and derive the ability model of super anchor, forming a set of efficient and replicable anchor training methodology.
In addition, the status of anchors also affects the penetration rate, stay time and conversion rate. Managers can also adjust the working rhythm of anchors and improve the service quality of broadcast rooms by identifying and analyzing the status of anchors through the model.
Each anchor has different advantages and disadvantages, and their own abilities are difficult to unify. However, if enough strong skills can be provided, after targeted training, anchors will be able to exploit their strengths and circumvent their weaknesses in the execution of work, and give full play to the best effect.
Live streaming e-commerce industry has come to the second half, in addition to marketing strategies and channels, while building brand power and product power, how to long-term improve the broadcast room "operating power", is the brand at the present stage to think about the problem.
On the brand's exploration of the possibility of fine operation of live streaming e-commerce, Whale DAM will help customers gather live streaming data and build algorithm models, so that the evaluation of variables affecting business growth can be clearly quantified. Based on the analysis of the portrait of a single store in the broadcast room, the adjustment of the live broadcast speech technique, the training of anchors and the adjustment of the live broadcast operation plan are more reliable.
About Whale strategizing
As a professional omni-digital marketing operation platform in China, Whale, through key technological innovations in artificial intelligence (AI), large-scale Internet of Things (IoT) and Data model (Data), provides future-oriented retail brands with data-driven, collaboration-first, simple and easy to deploy omni-brand marketing solutions. It aims to enable sustainable fine operation and lean growth of retail brands and open up the "last mile" of MarTech. At present, Whale strategy service system has widely covered food and beverage, beauty makeup and skin care, fashion shoes and clothing, light luxury jewelry, digital electrical appliances, catering and tea drinking, business super convenience, automobile service, medicine and health and other industries. It has accumulated benchmark customers such as Unilever, Watsons, Sibei, Midea, Paopamart, Nextev, Carrefour and more than 300 Top brands in the industry. The company was founded in 2017 in Hangzhou, and has offices in Shanghai, Shenzhen and Beijing.
