Recommend a BI tool - pony BI
There is a consensus in the workplace: data-driven business value. In the product business, operations, development, technical support, marketing and other aspects have a lot of data needs, the market there have been many BI tools BI visualization tools, but if we can also have the following characteristics, it can be called a good:
-
Easy access to the data
-
Generate charts drag formula
-
Quick calculation data
-
Regularly sent weekly
-
+ PC support mobile end end
-
free of charge
Combining these characteristics, to introduce a team built by Tencent TEG lightweight data visualization tools - pony BI .
Briefly explain this product.
0 threshold, be absurd to do it
By simple drag and drop you can use the data has been accessed and edit your data billboards, WYSIWYG:
Pony provide up to 26 kinds of components, covering "text, tables, index cards, line charts, bar graphs, pie charts, area charts, maps, radar, funnel, scatter, heat map, word cloud" and all mainstream visual diagrams.
Almost everything you can to meet the needs of data visualization, mining connotation arbitrary business data, quickly build your business billboards.
Effective and efficient, more than happy to reduce duplication of effort
-
It supports multiple data access methods
-
Direct database, Excel file uploads, SQL tables built, cloud database, API data access, just fill out a form, you can instantly access data to begin with.
-
Easy data cleansing
Desired data by creating a bonding association table; filtered through a filter, removing the invalid data; a data formatting, data is read to ensure ease of use.
-
Second-level computing efficiency in the process of creation, all the charts are generated in seconds, quick results, timely response and timely adjust, implement agile way.
-
The report template supports a push system of automatic periodic sending mobile terminal, mail report, a delivery report, liberate your Friday.
Even support dynamic variable text (that is, even text descriptions are values which can be automatically updated .)
Cross-platform, anytime, anywhere to enjoy the beauty of data
Introducing an end user PC, access data, perform simple page design, the user can publish to have permission to view the corresponding; on the movable end, also supported by the support push public micro channel number of the received reports.
The following details the use of this tool for data processing visualization process.
A Data Access
Pony BI data access table is divided into "Excel upload", "SQL to build," and "cloud database", "API data access," "multi-table associated with", "data aggregation", "SQL to create a table together." seven ways.
One of the most common was undoubtedly the "SQL table to build" and "build cloud database table" in two ways. With "SQL build tables" as an example. First, choose to fill out basic information and links to the database, then you can play for free.
Pony support include:
"MySQL", "PostgreSQL", "SQLSever", "Oracle", "Elasticsearch", "Hive" and other mainstream database types.
Real-time direct database approach does not store data, so no need to worry about data leakage. Of course, if the sensitivity is not authorized because the data does not matter database connection, pony also support for data access through the API mode.
II. Data processing
After the completion of the data access can be a simple ETL needed.
Pony support for data table field name, type of editing operation, and can use the "calculated field" function processing is required to field some of the analysis, such as we want to calculate profit margin merchandise, you can use the original table "profit" / "sales" get.
In addition pony can also support the correlation dimension tables and data tables, such as the original table in the sales area is the numeric codes, only the last dimension table relationships and a code area, select and correlation dimension table field types you can complete the association operation.
III. Visual analysis
After the completion of data processing can begin to visualize analysis, first included in the "Page Design", click on the upper left corner of the page to add "new page" or "New large screen pages", if desired. After the page has been created it can be used up to 28 chart components and filters.
Step1. To make your page a text explanation of the components and annotation capabilities.
Step2. Configure dynamic time / volume of core fixed time information indicator table assembly, real-time data at a glance.
Step3 . Configuring a time line chart (according to hour, day, month, year polymerization) sales trend for regional / time range filter through a filter, to understand trends overview.
Step4. Understand sales trends after the merchandise category to refine the analysis, summarize the category of sales situation, you can use the histogram and set in the chart to drill insight into the details of sub-categories.
Step5. Then look at the maps component sales in each region, using the same drill down to switch provinces.
Step6. In addition, you can also further combined, by setting the linkage assembly to analyze sales performance in different categories in different regions, auxiliary customized strategies such as the inventory cycle
Step7. Here have an understanding of the company's business conditions generally, then we look at the sales performance of a single SKU. Use the table component selection and value dimensions need to analyze, sort of key indicators. You can also set metric thresholds or warning, when the condition or push highlight show.
Step8. After reading the article sales data, we analyze the performance assessment of sales staff, where you can use the chart component to display performance ranking (area / time) at various dimensions of the sales staff.
Step9. Similarly, we can also suppliers were analyzed according to the rate of supply, delivery, yield, account periods of time. Here word cloud components used by sales to find good suppliers.
Step10. In order to facilitate screening of view, the page can also add a global component screening, screening for page-level filtering from multiple dimensions.
At this point we are from "real time data", "overall trend", "category analysis", "Regional analysis", "commodity analysis", "sales analysis", "vendor analysis" of several dimensions learned operation status.
