Enterprise data-driven new paradigm: Volcano Engine V-Tech Summit, shared by Tan Dai

September 19, 2023 Volcano Engine V-Tech The Data-Driven Technology Summit officially opens. At the summit, Tan Dai, President of Volcano Engine, explained the newly upgraded "data flywheel" concept of Volcano Engine. He said that building a data flywheel with "data consumption" as the core will become a new paradigm for enterprises to achieve data drive.

Tan Dai said that the "data flywheel" advocated by Volcano Engine consists of two parts: "business application" and "data assets", which is different from the past data middle platform. It is different

Among them, "business application" focuses on making business decisions more scientific and strategy implementation more agile through tools and BP mechanisms; while "data assets" make system construction more targeted. Through more frequent data consumption in the upstream, data will be continuously accumulated, data assets will be enriched, data quality will be further optimized, and data research and development efficiency will be improved.

At the meeting, Tan Dai also confirmed the above theory by sharing three best practices in the construction of "data flywheel"

The first thing is to enable everything to be Metrics

For this, the experience inside Byte can be summarized in four numbers: "0987". Among them, "0" represents "zero data incidents", which places high demands on the company's technical capabilities, operation and maintenance, and governance; while the number "9" means that 90% of the needs are met. The data team needs to be very familiar with the specific business and be able to interact in depth with product and business personnel. Only in this way can 90% of business needs be truly met

The third number "8" refers to 80% Analysis - Judging from a large number of business practices, 80% analysis coverage is a relatively reasonable goal; the final "7" refers to 70% NPS, which is the business team's favorable opinion of the data team. In the industry, an NPS of 70% is a very high standard. Based on this indicator, we should discover various problems in the data service link to improve business satisfaction.

In terms of measuring the level of "data consumption" of enterprises, Byte's experience is "two 80%"; the first is "80% of people in the enterprise use various data every day Tool usage and data consumption" - which includes not only data engineers and data analysts, but also products, operations, marketing, and even administration, HR, and UED, which are traditionally "people who are far away from the data."

The second 80% refers to the unified construction of analysis indicators, which can cover 80% of daily analysis and business scenarios. This means it analyzes and uses data efficiently in most situations. At the same time, it also reserves enough flexibility for data analysis and application in special scenarios

Through two 80%, enterprises can well measure their own " Is the “data flywheel” functioning well?

The second point of Byte practice that Tan Dai shared is that in daily business operations, "it is very important whether the boss can see the number or not." He said that data-driven is a top-down thing and a culture. If the department leader has the habit of reading numbers, then the department will most likely be data-driven; if the boss of a company can develop the habit of reading numbers, then the company will most likely be data-driven.

The third point is to ensure good tool construction. Corporate culture has been formed and goals and processes have been quantified. Without good tools, the company's "data flywheel" will still not be able to spin.

Tan said that the original intention of Volcano Engine to launch the digital intelligence platform VeDI was to provide byte-based data tools to help more companies build their own data flywheels. He also mentioned that new technological changes will also bring new capabilities to the data flywheel. Among them, large model technology will bring a major upgrade to the "data flywheel": through large model technology, enterprises can better process unstructured data and help enterprises collect and process more data sources; the blessing of large models not only reduces The threshold for enterprise employees to consume and apply data also improves the efficiency and accuracy of R&D personnel in the data development, data governance and data analysis processes. (Author: Liu Yuan)

The above is the detailed content of Enterprise data-driven new paradigm: Volcano Engine V-Tech Summit, shared by Tan Dai. For more information, please follow other related articles on the PHP Chinese website!