Measuring results for a client: what to track and how to report it
One number per automation, taken the same way before and after, reported monthly in five lines. Metrics by type and where measurement fails.
3 min read Reviewed 22 September 2026 · AgeBridge Editorial
Measure one number per automation for the client, chosen before the build, taken the same way before and after, and reported monthly in five lines the owner reads. Track two supporting counts for yourself and two health counts (runs, failures) for the support plan. That is the whole measurement system for a small-business automation; more metrics produce reports nobody reads and arguments about which one matters.
Why one number?
Because the owner has one question: is it working? A single number they chose, with a before and an after, answers it. Three numbers invite interpretation; ten invite a dashboard nobody opens. The supporting counts exist so you can explain the primary one, not so the client has to.
| How many metrics per automation (editorial recommendation) | |
|---|---|
| Primary (client reads) | 1 |
| Supporting (you track) | 2 |
| Health (failures, runs) | 2 |
Editorial guidance, not survey data: one primary number per automation for the client, plus at most two supporting counts you track for yourself. More than that stops being read.
The primary number by automation type
| Automation | Primary number | Where it comes from |
|---|---|---|
| Lead reply | First-reply time (median) | CRM timestamps: received vs first reply |
| Appointment reminders | No-show rate | Calendar: booked vs attended |
| Quote follow-up | Quotes answered within 10 days | CRM stage changes |
| Document processing | Documents processed without manual entry | Pipeline log |
| Support assistant | Questions answered without a person | Conversation log + routing count |
| CRM hygiene | Duplicates created per week | CRM search |
Choose it with the owner in the brief. If the system has no way to produce it, add the field; measurement is part of the build, not a report at the end.
Taking the baseline
Before you build, measure the same number the same way for two weeks. Write down the method: "median of received-to-reply for all website leads, from CRM timestamps, 1–14 September". Without a baseline the after has nothing to stand against, and "it feels faster" is all you'll have.
The five-line monthly note
- The number: this month vs baseline, in one line.
- What ran: count of runs, count of failures and what you did.
- What changed: the small changes made, one line.
- What you noticed: one observation from the logs.
- Next: one suggestion, priced separately if it's an add-on.
Same day every month, same format. It is the most-read thing you'll write for that client, and where the next project comes from.
Where measurement goes wrong
- Changing the method. Different filter, different period, different definition: the comparison dies. Freeze the method in the brief.
- Measuring what's easy instead of what matters. Messages sent is easy; leads that booked is what matters.
- Claiming causation you can't show. "Revenue up 20%" when you automated reminders: report no-shows, and let the owner connect it to revenue.
- Hiding a dip. The owner will find out; report it first with a plan.
- Invented or rounded numbers in case studies. Use the measured figure, its period and its method, or none.
Reporting to the public
The same number, period and method go into the case study, with the owner's permission. "Median first-reply time from 9 hours to under 1 minute, measured over two weeks from CRM timestamps" is a sentence buyers trust because it can be checked. The proof ladder is built from measurements like this one.
Best fit and not a good fit
Best fit: every automation that touches customers, leads, bookings or documents. Not a good fit: internal conveniences with no countable outcome; report runs and failures only, and don't invent a metric to look rigorous.
What to do this week
For your current project, write the primary number, the method and the baseline period on one line. If you can't, add the field that makes it measurable before you build anything else.
Questions people ask
What if the number goes down?
Report it first, with the likely cause and what you'll change. A candid dip with a plan builds more trust than a quiet month, and it usually points at the next improvement.
Should I build a dashboard?
Not for a small business. A five-line monthly message that the owner actually reads beats a dashboard nobody opens. Build the dashboard when they ask twice.
How do I measure when the CRM has no timestamps?
Add them: the automation writes 'received at' and 'replied at' fields on every record. If the system can't hold them, a sheet can. Measurement is part of the build.
Sources
- HubSpot Knowledge Base · HubSpot · 2026-06-01
Editorial guidance, not advice. Estimates are labelled and dated; nothing here is AgeBridge marketplace data unless it says so.
Put this into practice on your AgeBridge profile
A free profile with real projects is what businesses look at first.
Build my profileRelated guides
How to write a case study a business owner will actually read
A one-page structure for AI-builder case studies: the problem in the owner's words, what you built in plain language, what changed, and proof a buyer can check.
4 min read · 22 September 2026
Maintenance and support plans for automations: what to include
Automations break quietly. A support plan that keeps clients: monitoring, a response time, a change budget, a monthly note, and what is extra.
3 min read · 22 September 2026
How to run a two-week pilot that turns into a retainer
The pilot that converts: fixed scope, five days on real data, one number, a hand-over the client can open, and the retainer conversation on the last day.
3 min read · 22 September 2026