Our four-person team at Pristren tracks between 160 and 200 hours of work per week inside Zlyqor. That is 640 to 800 hours per month, across 30 to 50 active tasks at any given time, with every hour linked to a specific task in a specific project. After six months of this data, I can tell you what it actually shows about where knowledge-worker time goes, how AI meeting summaries changed our post-meeting workflow, and what the difference is between having integrated time tracking versus a standalone tool like TimeCamp.
The short answer: integrated time tracking changes behavior, not just reporting. When the timer is attached to the task, people start it. When it requires a separate application, they often forget.
What the Numbers Look Like
Four team members. Eight to ten hours of work per day. Five days per week. On the high end, that is 200 hours per week. On the low end, with shorter days or partial-week availability, it is around 160 hours.
The weekly breakdown across our team, based on six months of aggregated data:
Deep work (development, writing, design): 45 to 55 percent of tracked hours. This is the work that produces the actual product. For us, this is building Zlyqor itself, writing content, and designing features.
Meetings: 15 to 20 percent of tracked hours. For a four-person team, this runs between 25 and 40 hours per week when you include client calls, internal syncs, and planning sessions. That is higher than I expected when I first saw the data.
Administrative and coordination work: 10 to 15 percent. This includes replying to messages, reviewing pull requests, handling billing, and the category I call "work about work" — updating task statuses, writing project summaries, filling out reports.
Client communication and support: 10 to 15 percent. This varies significantly week to week depending on where our projects are in their cycles.
Research and learning: 5 to 10 percent. This is the category most teams undercount because it often happens in fragments throughout the day rather than in dedicated blocks.
The first thing that surprised me was the meeting percentage. Twenty percent of a knowledge-work team's time spent in meetings is considered high by most productivity research, but 25 to 40 hours per week across four people felt accurate once I saw it in the data. It was invisible before because we never tracked meetings as a separate time category.
What TimeCamp Gave Us vs. What Zlyqor Gives Us
I want to be precise about this comparison because the difference is more significant than it might appear.
TimeCamp is a capable time tracker. It has a browser extension that can auto-detect what you are working on based on window titles, a mobile app, detailed reports, and integrations with project management tools like Asana and Trello. We used it for about 18 months before switching.
What TimeCamp could not do: link time entries to tasks that also contained the conversation, files, and decisions related to that work. When I exported a week of time data from TimeCamp, I got a table of hours against project names. I could not click through to see what was actually done during those hours, what decisions were made, or what the status of the underlying task was.
Zlyqor's timer is integrated into each task. You open a task, you press start. The timer runs. You stop it, and the time is logged against that task. The task also contains the thread of comments, attached files, and linked meeting summaries. When I look at a week of tracked time in Zlyqor, I can navigate directly from the hour log to the work it represents.
The behavioral difference: our timer utilization rate with TimeCamp was approximately 65 to 70 percent. People forgot to start the timer, forgot to stop it, or logged time retroactively at the end of the day with rough estimates. With Zlyqor, utilization is above 85 percent. The friction of starting is lower because the timer is already where the work is.
The 15 to 20 percentage point improvement in utilization means our data is more accurate, which means our invoices are more accurate, which means we are capturing revenue we were previously leaving on the table.
How AI Meeting Summaries Changed Our Post-Meeting Workflow
Before Zlyqor, our meeting workflow looked like this: meeting happens, someone types rough notes in a Google Doc, the Google Doc is shared in Slack, most people do not read it because the meeting just happened and they were there, the Google Doc becomes difficult to find one month later when someone needs to remember what was decided.
After Zlyqor: meeting happens, AI generates a summary within a few minutes, the summary is automatically attached to the meeting record in the workspace, everyone can see it in context.
What the summary contains: a list of topics discussed, decisions made, action items with the person assigned, and open questions. The AI does not capture everything, and it makes errors, but it captures about 80 percent of what a human note-taker would produce, in less time, with no one having to volunteer for the job.
The behavioral change: people started referencing meeting summaries in task threads instead of asking "can someone remind me what we decided in the Monday call?" That question happened multiple times per week before. It almost never happens now.
I estimated the time saved per meeting at approximately 15 to 20 minutes of post-meeting documentation work. Across 6 to 12 meetings per week, that is between 90 minutes and 4 hours of documentation time per week returned to the team.
What We Do Not Track (and Why It Matters)
I want to be honest about the gaps in our data. Not all work shows up in the timer.
Async reading and thinking. When a developer reads documentation for 45 minutes before writing code, that often does not get tracked because it does not feel like "work" in the billable sense. Our data almost certainly underrepresents research and learning.
Communication fragments. A two-minute Slack reply does not usually trigger a timer start. Across dozens of short messages per day, this probably represents 30 to 60 untracked minutes per person.
Context rebuilding time. The time spent re-reading previous conversations and task notes to get back into context before starting work is not tracked. This is the Gloria Mark problem applied to our own data: the logged hours show the work, but not the overhead around the work.
If I had to estimate the true work hours including these gaps, I would add 15 to 25 percent to our tracked figures. The tracked hours are useful for billing and project management, but they are not a complete picture of where the team's time goes.
What This Data Is Actually Good For
The data we have collected over six months is most useful for three things:
Project estimation. We now know, from actual data, that a certain class of feature takes between 12 and 18 hours to build, not the 8 hours we used to estimate. That has made our project quotes more accurate and our client expectations more honest.
Team balance. When one person's tracked hours are consistently above the team average, it is a signal worth investigating. In two cases over six months, the data pointed to workload imbalances we would not have caught until someone burned out.
Billing accuracy. On fixed-fee projects, the data tells us whether we priced correctly. On hourly projects, it ensures we are capturing everything we should be billing.
What it is not good for: measuring individual productivity or evaluating output quality. Hours tracked is a proxy, not a measure of value created. I look at this data as a project management tool, not a performance evaluation tool.
Keep Reading
- We Replaced 6 SaaS Tools With One: What 6 Months of Real Data Shows — The full context for why we made the switch and what it cost
- 30-50 Tasks Per Week: How a 4-Person AI Team Actually Structures Work — How the task data maps onto the time data and what the combination reveals
- The Hidden Cost of Tool Switching: What We Measured With Our 4-Person Team — The formula for calculating the real cost of running multiple tools
Pristren builds AI-powered software for teams. Zlyqor is our all-in-one workspace — chat, projects, time tracking, AI meeting summaries, and invoicing — in one tool. Try it free.
