I was reading the Algorithm book last week and one concept hit different: touch time vs cycle time. Most companies optimize the wrong one.

Touch time is the actual work. The 30 minutes of hands-on technical effort. The value-adding activity when someone is genuinely producing output. Cycle time is the total elapsed time. The 2 weeks from start to finish. That includes the work plus everything else: queues, approvals, handoffs, waiting, transport, setup, inspection. The ratio here is 670:1.

Most organizations try to make people work faster. Cut that 30 minutes to 15. Push harder. Add more shifts. All aimed at touch time. But touch time is usually less than 5% of cycle time. Even if you cut it in half, you only reduce total cycle time by 2%. The 95% nobody looks at is where the real opportunity lives. The waiting. The handoffs. The approvals stacked on approvals. The batching.

Elon's approach at Tesla and SpaceX repeatedly attacked this gap. The question was always: why does this take so long? The answer is almost never "because the work itself is slow." The answer is "because it waits."

Here's what changes when you understand this: You stop asking how do we work faster. You start asking why is this waiting at all. Measure touch time honestly. Time the actual work as if it were continuous. That's your floor. The absolute minimum the cycle could be. Compute the gap. Cycle time minus touch time equals waste. Every minute of that gap is queuing or transferring. None of it creates value.

Attack the gap, not the work. Batch sizes go down, often to one. Handoffs collapse. Approvals get questioned and deleted. Inspections move inline. Use touch time as the target. If touch time is 10 minutes, ask why cycle time isn't 12 minutes.

We're doing this right now with our Quave ONE Connect plan setup. The actual technical work to onboard a customer is maybe 45 minutes. Provision their access to our platform, add their VMs to our cluster, configure the load balancers, do a first test deploy, etc. That's the touch time. But the cycle time? Often 3 to 5 days. Why? Because of back-and-forth. We schedule a call, walk through the process, then wait for the customer to create credentials in their cloud provider. They send us the wrong permissions. We ask again. They don't know how to set up IAM correctly. Another email. Another wait. We finally get access, start the provisioning, realize we need one more permission. Another round.

So we're collapsing it. Now on the first call, we walk the customer through creating the credentials right there. We share screens. We check the specific permissions in real time. We provision the user together on the call and verify access immediately. No follow-up emails. No waiting. No assuming they know how to do it. If something is wrong, we fix it in the moment. The touch time stays 45 minutes, but the cycle time drops from 5 days to 90 minutes.

Next step is automating the VM integration completely. Right now there's still manual configuration when adding VMs to the cluster and setting up load balancers. We're building automation that takes the credentials, scans the customer's infrastructure, and provisions everything with one click. Touch time and cycle time become nearly identical.

This is how the Connect plan is built. The leverage isn't in making engineers work faster. It's in eliminating every wait, every handoff, every moment where something sits idle. Cycle time is what the customer experiences. Touch time is what the work actually requires. The space between them is almost entirely waste. That's where 10x improvements live. Companies that achieve hypergrowth don't have faster workers. They have processes where touch time and cycle time are nearly identical.

Have a great week and think about your touch time more than your cycle time, please!