Codeit has a feature that enables you to track coders and projects against certain metrics of performance or usage. This article explains how to use this feature.
Orientation Video Guide
Summary of Productivity Reporting terms & metrics
The screenshot below shows the interface of the productivity reporting tool, showing the filters and metrics available. The table below explains each of the filters and metrics to assist in selection and interpretation of the data provided.
| Area | Tool | Description |
| Toggle screen | Projects or Coders | Choose to report on projects or coders as the rows being generated in the report |
| Filters | Date Range | Choose the date range for the report. This could be either for monthly reporting requirements or for the duration of a project, for example. |
| Companies | If you have a structure of companies and sub-companies, you can filter by them here. | |
| Coding Modes | Allows you to filter the report on a specific coding mode (e.g. block coding, short text coding, verify). This is useful if you want to exclude, say, time spent in verify from the overall productivity numbers. | |
| Projects | Choose the specific project or projects here. | |
| Coders | Select specific coders to be reported on. | |
| Metrics | Total Duration (hours) | The total number of hours spent in the coding screens. Note: any time spent logged into any non-coding screens (e.g. project dashboard, administration, import/export etc...) is not included in the duration numbers. |
| Total Changes | Total number of times the coding on a given workload has been changed and saved. | |
| Total Coded / Hour | (First Coded Human + First Coded Human after AI) / Duration | |
| First Coded Human | The number of completely uncoded workloads that had at least one code applied by a person. Subsequent changes to a workload by a person wont be counted in this measure, but will be counted in "Total Changes" | |
| First Coded AI | The number of completely uncoded workloads that had at least one code applied by the AI. Subsequent changes to a workload by the AI wont be counted in this measure, but will be counted in "Total Changes" | |
| First Coded Human after AI | The number of workloads that have codes applied or removed by a person having first had codes applied by the AI. Subsequent changes to a workload by a person wont be counted in this measure, but will be counted in the "Total Changes". This measure is also useful as an indicator of how often people are required to correct the work of the AI | |
| First Coded AI after Human | The number of workloads that have codes applied or removed by the AI having first had coded applied by a person. Subsequent changes to a workload by a person wont be counted in this measure but will be counted in the "Total Changes". This measure is also useful as an indicator of how often the AI is making changes to the work of people. | |
| Totals | Page | Sum total for the page based on the filters selected |
| Overall | Sum total for the entire report based on the filters selected |
Note on 'First coded' metrics
The "First Coded" metric is often used by coding services companies when billing their clients. In these situations, it is common for companies to charge for "Total Verbatims", so it doesn't matter if the coding of a workload is changed a number of times (e.g. by the coder in Block Coding, then by a project owner in Verify) - hence, the importance of the "First" time a workload changes from uncoded to coded.
Differentiating between "Human" and "AI" also allows service companies to offer a reduced rate to clients for work done by the AI rather than by people.