With the COVID-19 outbreak and its expected influence on the economy,
every manager has to assess the impact on their daily business.
To make this process both easier and based on proper data science,
we are providing a free beta business impact explorer, tailored for logistics.
Please, make sure to read the instructions before using the tool,
in order to achieve the most accurate output:
How To Use Business Impact Explorer
Contact us at info@transmetrics.eu with any comments or suggestions for the tool's improvement. Also, let us know if you would like to get the source code - it is public.
How to use the tool:
1. Prepare Data
The Business Impact Explorer needs as input a time series .csv file, e.g. the daily volumes of your operations or daily revenue. We recommend having at least 6 months of data in the data set to get correct estimates. Your data set should not be larger than 10 MB and should contain the following:
- Download Sample File to see the data format or to test the model...
All you will need to do is to upload this .csv to the tool.
We are not storing your data, nor the output of the model. The tool is free to use. It is based on open-source code. It does not intend to gather any information about the users’ business.
2. Run Model
Once data is ready, you can upload it in the “File Input Box”.
Then, choose the “Date Column” as well as the “Value Column” which contains the volumes or financial data you would like to analyze. The tool does not detect these automatically because they are business-specific. You need to select the columns from the drop-down menus.
After that, you have two dates that you need to fill in: the start of the Business-as-Usual period and the Date of suspected business change due to the "out-of-the-ordinary" event that the algorithm should consider.
Make sure that the "Date of suspected business change" is 1-12 weeks before the end of the dataset in order for the tool to work correctly.
The model can be run on the whole data set or just on a subset (e.g. particular location or service). Use the "Filter Data" if you would like to see the results on a more granular level.
Then, press the “Compute Model” and wait for the output to display.
3. Interpret Results
The output of the model gives you 3 charts together with a short text on how to interpret those results.
Chart 1:
- Actuals are displayed with a thick line;
- The business as the usual situation is based on a machine-learning algorithm and it is represented by a dotted line.
- The more the first differs from the second, the bigger the suggested impact.
Chart 2 shows the estimated difference between the business-as-usual model and the actual values for each date.
Chart 3 demonstrates the accumulated difference between the business-as-usual model and actual value up to a given date.
The blue area in the charts shows the confidence intervals around the estimations. The impact on the business is reported in the text below the charts. It also contains the indications of the statistical significance of the estimations.
Additional Information
Since the tool is developed for use on desktops and laptops, there might be some issues with displaying it on mobile. If you are accessing the page from the Safari browser, in order to download the Sample Data, right-click on the link and select "Download Linked File As..." and then choose .csv.
For more information please check Terms of Use.
Many thanks to R, Shiny, and causal impact package authors for providing their resources and frameworks open source and making this tool possible. Let us know if you would like to get the source code of the tool - we also keep it public.
The project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 945610.