Where the data comes from
- TimeCamp’s desktop app. During a 2-week pilot it records which apps and websites are in use, for how long and when people switch between them. Employees are informed and give consent before any data is collected.
- Work items from your tools. Projects, tasks, tickets and deals are imported from up to 22 tools, such as Asana, Jira, Zendesk and Salesforce. AI matches the activity to these items.
- Your hourly cost. Costs use the figure you give us. Without one, we use the average U.S. private industry employer cost per hour worked, $46.89 in June 2026 (U.S. Bureau of Labor Statistics, Employer Costs for Employee Compensation, June 2026), rounded to $47.
Where the sample report data comes from
The sample report describes 38 people in 4 teams (support, sales, finance and engineering) over a 14-day pilot, the same scope and format as a customer report.
Today the figures are illustrative. They were built to follow the definitions and formulas on this page, so every total can be checked, but they are not measurements. They will be replaced with anonymized data from TimeCamp’s own teams, collected with TimeCamp’s desktop app and with the employees’ consent, after the anonymization is approved.
What each number means
| Metric | How it is measured |
|---|---|
| Active time | Time an app or website is in use on the computer. Idle time, when there is no keyboard or mouse activity, is left out. |
| App switches per hour | Changes of the app or website in use, divided by active hours, for each person, then averaged for each team. The overall average weights each team by its number of people. |
| Time in apps | The share of active time in each category: core work tool, communication, docs and spreadsheets, other browsing and other apps. Each app or website belongs to one category. |
| Repeated sequence | The same apps used in the same order for one piece of work, for example a helpdesk, then Jira, then Slack, repeating across the team. |
| Time per task, ticket or deal | Active time that AI matches to an imported task, ticket or deal, averaged for each type of work item. |
| Handoffs | Moves of the same work item from one tool to another, counted per week. |
How hours and cost are calculated
| Result | Formula |
|---|---|
| Hours a month for a sequence | times a week × minutes each ÷ 60 × 4.33 |
| Hours saved by an automation | hours a month of the steps the automation removes |
| Cost a month | hours × hourly cost ($47 in the sample report) |
| Cost per work item | average hours per item × hourly cost |
4.33 is the average number of weeks in a month (52 ÷ 12). For example, a sequence that runs 41 times a week and takes 6.5 minutes each time uses 41 × 6.5 ÷ 60 × 4.33 = about 19 hours a month.
How automation candidates are ranked
A task becomes a candidate when it repeats often, takes significant time and follows predictable steps. Candidates are sorted by hours saved per month, so the list starts where the return is largest. Each one also gets an effort level and a tool type:
- Low effort: data moves between tools that both have an API, with no judgment in between. Usually a no-code workflow.
- Medium effort: a step needs reading or writing free text, or working through screens of a tool without an API. Usually an AI agent or RPA.
- High effort: the work changes several systems, needs approvals or has many exceptions. Often a process change comes first.
How the ROI calculator works
The automation ROI calculator estimates the same value before any measurement:
time back = people × repetitive hours a week × share automated × working weeks
value = time back × team cost per hour
first-year cost = tools and AI × 12 + build hours × team cost per hour
net value = value − first-year cost
payback = build cost ÷ (monthly value − monthly tools and AI)It starts at 4.5 hours of repetitive work a week, what office workers said in the UiPath 2021 Office Worker Survey, and assumes a quarter of it is automated, because survey answers are what people think, not a measurement.
Limits
- Two weeks can miss work that happens monthly or quarterly, such as a month-end close. Name such work on the first call.
- Some activity cannot be matched to a work item. It still counts in time in apps, but not in the cost per task, ticket or deal.
- Hours saved assume the automation removes all steps of a sequence. Usually a person still checks the result, so treat each value as the most it can give back.
- All results are estimates, not guarantees.
Privacy
Results are shown for teams, tasks and work types, never as scores for individuals. Read the privacy policy.