Should I be watching average ticket or median ticket?
Both, side by side. The average is what pays the bills and is the only version that works in revenue arithmetic. The median describes what a typical job looks like and is not moved by one large replacement. When the average moves and the median does not, a handful of big jobs did it. When both move together, something changed about the whole job population.
Two numbers, two different questions
Ticket values in a service business are heavy-tailed. Most jobs cluster in a fairly narrow band and a small number are many times larger. That shape means the average sits well above the median and is highly sensitive to the top of the distribution.
So the average answers a financial question — what does a job contribute on average to the total — and the median answers an operational one — what does a normal job look like for a technician on a normal day. Confusing them produces goals nobody can hit and comparisons that mean nothing.
The spread is the actual diagnostic
Track the gap between average and median over time. A widening gap means the business is becoming more dependent on large jobs, which is usually a good revenue story and a bad predictability story. A narrowing gap means either the large jobs stopped or the small ones came back.
Add one more column: the share of revenue coming from the top decile of jobs. When that share climbs, month-to-month volatility climbs with it, and forecasting from an average becomes progressively less reliable.
Where the average is the only correct choice
Anything additive requires the average. Revenue per lead, cost per acquired job, expected value of an open pipeline, capacity planning — these all multiply a count by a mean, and substituting a median understates the total.
Anything comparative and human-facing is usually better served by the median. Comparing technicians, comparing job types, or setting a realistic expectation for a shift are median questions. Using the average for those tells a technician who ran ten ordinary calls that they underperformed a colleague who happened to catch one system replacement, which is a coaching conversation built on a statistical artifact. See rep coaching for how performance data should support a human decision rather than substitute for one.
A layout worth copying
For each job type and period, show completed jobs, median ticket, average ticket, and the top-decile revenue share. Four columns, and together they say more than any single headline number.
Adding source to that table is where it becomes genuinely useful, because ticket distributions differ by channel in ways that blended numbers hide entirely. That join is the core of revenue intelligence and the reason it needs real data integration rather than an export.
Topics: average ticket · median · distribution · reporting · diagnostics
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.