KPI Dashboard Template for Service Businesses: Free CSV Download (2026)
A practical KPI dashboard template for service businesses, with a downloadable CSV, synthetic worked data and clear definitions for sales and operations decisions.
Co-founder of Ampliflow. Builds AI automation, websites, SEO/AEO, and growth systems for UK SMEs.
- 01What the template includes
- 02KPI definitions
- 03Worked example
- 04Source breakdown
- 05How to adapt the CSV
Most KPI dashboard templates give you boxes to fill in. That is not enough.
The hard part is not drawing a dashboard. The hard part is deciding what each metric means, where it comes from, how fresh it needs to be, who owns it, and what decision it should change.
Quick answer: A useful service-business KPI dashboard template should include the metric name, definition, denominator, data source, freshness, owner, decision and worked example values. Start with sales follow-up, quote conversion, response time, overdue work, capacity and revenue by source before adding department-specific charts.
Last updated: 23 September 2026 · Includes synthetic worked data · Written for UK service businesses
Download: service-business-kpi-dashboard-template.csv
The CSV is synthetic. It is not Ampliflow client data, benchmark data or a claim about normal performance. It is a worked example you can adapt before connecting live systems.
KPI means key performance indicator: a measure that helps someone decide what to do. A denominator is the total used to calculate a rate, such as all eligible enquiries when working out what percentage qualified.
What the template includes
The template has three kinds of rows:
| Row type | Purpose |
|---|---|
| Metric definition | Explains the KPI, denominator, source, owner, freshness and decision |
| Weekly worked data | Shows example values for a small service business over four weeks |
| Source breakdown | Shows how lead source can change qualification, quoting and won work |
Use it in Excel, Google Sheets, Power BI, Looker Studio or a custom dashboard prototype. The point is to agree the logic before worrying about the tool.
The dashboard structure
- 01Define the decision
- 02Agree the metric and cohort
- 03Assign an owner
- 04Set the refresh interval
- 05Show the records to act on
Start with one screen that answers five questions.
| Question | KPI group | First decision |
|---|---|---|
| Are enough good enquiries arriving? | Lead volume, qualified opportunity rate, source quality | Improve source mix or qualification |
| Are buyers getting a fast response? | First response time, leads over response target | Reassign cover or improve routing |
| Are quotes turning into work? | Quotes sent, quote acceptance rate, lost reasons | Fix proposal quality, pricing clarity or fit |
| Is work getting stuck? | Open jobs, overdue work rate, blocked work | Escalate blockers or adjust capacity |
| Which channel creates value? | Won revenue by source, lead source quality | Move attention and spend |
This is enough for most first dashboards. The detail table beneath it matters more than extra charts.
KPI definitions
| KPI | Definition | Denominator | Freshness | Owner | Decision |
|---|---|---|---|---|---|
| New enquiries | New commercial enquiries received in the period | N/A for the count; all inbound commercial enquiries for context | Daily | Sales lead | Decide whether demand is rising or falling |
| Qualified opportunities | Enquiries that match service, budget, timing and fit rules | New enquiries | Daily or weekly | Sales lead | Improve source quality or qualification |
| Qualified opportunity rate | Qualified opportunities divided by new enquiries | New enquiries | Weekly | Founder or sales lead | Shift effort away from weak sources |
| Median first response time | Median minutes from enquiry creation to first useful reply | Eligible enquiries requiring a response | Same day | Sales lead | Reassign cover or improve routing |
| Quotes sent | Quotes issued to qualified buyers | N/A for the count; qualified opportunities for context | Daily or weekly | Sales lead | Check whether good leads are moving |
| Quote acceptance rate | Accepted quotes divided by quotes sent in the same quote cohort | Quotes sent in the matched cohort | Weekly | Founder | Review fit, proposal clarity or pricing |
| Won revenue | Revenue attached to won work in the period | Won jobs with value | Weekly or monthly | Founder or finance | Compare channel value, not just lead volume |
| Open jobs | Active jobs not yet completed or closed | N/A for the count; all active job records for context | Daily | Operations lead | See workload and pressure |
| Overdue work rate | Overdue open jobs divided by all open jobs | Open jobs | Daily | Operations lead | Escalate blockers or reset capacity |
| Unassigned work | Active leads or jobs with no owner | Active records | Same day | Operations lead | Assign ownership before work drifts |
| Capacity pressure | Active work compared with agreed capacity | Team capacity units or available slots | Daily or weekly | Operations lead | Move work, adjust diary or hire support |
Use the denominator column for rates and record the population or period for counts. For response time, define what counts as a substantive reply and record automated acknowledgements separately.
Worked example
This table uses synthetic data from the CSV. It describes a fictional UK service business across four weeks.
| Week | New enquiries | Qualified | Quotes sent | Accepted quotes | Median first response | Open jobs | Overdue work |
|---|---|---|---|---|---|---|---|
| 2026-08-03 | 42 | 25 | 18 | 9 | 46 minutes | 31 | 5 |
| 2026-08-10 | 48 | 29 | 22 | 11 | 38 minutes | 34 | 4 |
| 2026-08-17 | 51 | 28 | 19 | 8 | 74 minutes | 39 | 9 |
| 2026-08-24 | 45 | 31 | 24 | 13 | 33 minutes | 35 | 3 |
The useful reading is not "week three was bad". It is more specific:
- Enquiries rose in week three, but qualified opportunities did not rise with them.
- First response time worsened in the same week.
- Overdue work also rose, which suggests capacity or handoff pressure.
- Week four's higher accepted-quote count coincided with faster response time and lower overdue work.
A dashboard should make that pattern obvious, then show the records behind it.
Read week three without jumping to conclusions
The week beginning 17 August shows a reason to investigate, not a diagnosis. Qualification fell while response time and overdue work increased. The figures alone cannot tell us whether the cause was poorer enquiries, staff absence, harder jobs or incomplete records.
Download the dashboard decision review (CSV). Version 23 September 2026 reproduces three calculations from the existing synthetic template. It adds the next record to inspect and what each number cannot establish. Both files can be reused in your own planning; neither contains client results or market benchmarks.
| Measure | Week beginning 10 August | Week beginning 17 August | Change |
|---|---|---|---|
| Qualified opportunity rate | 29 ÷ 48 = 60.4% | 28 ÷ 51 = 54.9% | Down 5.5 percentage points |
| Quote acceptance rate | 11 ÷ 22 = 50.0% | 8 ÷ 19 = 42.1% | Down 7.9 percentage points |
| Overdue work rate | 4 ÷ 34 = 11.8% | 9 ÷ 39 = 23.1% | Up 11.3 percentage points |
A percentage point is the difference between two percentages. Calculate each rate from the unrounded counts, subtract the earlier rate, then round the result to one decimal place. A percentage-point change is different from a relative percentage change.
Read the number, then inspect the work
- 1
Qualification: 28 of 51
Inspect the enquiry sources and reasons for poor fit. A lower rate does not tell you which source changed.
- 2
Acceptance: 8 of 19
Inspect the matched quote cohort. The eleven other quotes may be pending, declined or expired; the total does not say.
- 3
Overdue: 9 of 39
Inspect the nine overdue jobs at the snapshot time. Assign the next action before adding another chart.
Keep the measurement boundaries visible
The example uses matched enquiry and quote cohorts: groups counted under the same definition. Quote acceptance belongs to the week the quotes were issued, assessed at a common follow-up interval. It is not “quotes accepted this week divided by quotes sent this week”. These synthetic figures have no underlying contact records or observed follow-up dates; choose and record that interval before using real data.
Open and overdue jobs are snapshots, not weekly flows. Do not add the four open-job counts together: the same job could appear in more than one snapshot. Median response times also cannot be averaged to recover a reliable overall median; that needs the individual response records.
For live reporting, show the data's last successful refresh and the number of missing or excluded records. If there were no quotes, display not applicable: no quotes in this cohort, rather than 0% acceptance. If status is missing, show it as unknown rather than treating it as a rejection.
The source totals below reconcile with the four-week totals for enquiries and qualified opportunities. They still cannot explain week three's source mix: that would need a source-by-week breakdown. This is why a useful dashboard links its summary to the records behind it.
Source breakdown
Lead source reporting is where many dashboards become misleading. Volume is not quality.
| Source | Leads | Qualified | Quotes sent | Accepted quotes | Won revenue |
|---|---|---|---|---|---|
| Calls | 52 | 35 | 27 | 15 | £18,400 |
| Website forms | 47 | 29 | 22 | 10 | £11,900 |
| Referrals | 18 | 15 | 13 | 9 | £16,800 |
| Paid ads | 43 | 18 | 13 | 4 | £4,700 |
| 26 | 16 | 8 | 3 | £3,100 |
This is synthetic same-cohort worked data: each source row aligns leads, qualified opportunities, quotes sent and accepted quotes from the same example cohort. The decision it illustrates is real: do not judge a channel by lead volume alone. Referrals may be lower volume but higher quality. Paid ads may need better targeting, landing-page intent or qualification.
How to adapt the CSV
Use the template in this order.
- Replace the synthetic source names with your real sources.
- Define what counts as a commercial enquiry.
- Define what makes an enquiry qualified.
- Decide whether a quote is counted when drafted, sent or received by the buyer.
- Decide who owns stale leads, overdue work and missing next actions.
- Set a freshness rule for each KPI.
- Add the detail table your team will act from.
The dashboard does not need perfect data on day one. It needs agreed definitions and an owner for each visible problem.
What to put on the first screen
Use fewer cards than feels comfortable.
| Position | Widget | Why it belongs |
|---|---|---|
| Top left | New enquiries and qualified opportunity rate | Shows whether demand and quality are moving together |
| Top middle | Median first response time | Shows whether buyers are waiting |
| Top right | Quote acceptance rate and won revenue | Connects activity to commercial outcome |
| Centre | Attention queue | Shows leads, quotes and jobs that need action |
| Lower left | Source quality | Shows where useful work comes from |
| Lower right | Operations pressure | Shows overdue work, blocked jobs and capacity |
If the dashboard cannot show the actual records behind a red number, it is weaker than it looks.
Tool choice
The template works before you choose the dashboard tool.
| Tool | Use it when | Watch for |
|---|---|---|
| Spreadsheet | You are agreeing definitions and testing the logic | Manual updates and version drift |
| Power BI | You need internal reporting, modelling and Microsoft governance | Tenant setup, licensing, permissions and user comfort |
| Looker Studio | You need lighter marketing or website reporting | Connector reliability and operational limits |
| Custom dashboard | Users need actions, roles, portals or AI workflow logic | Scope discipline and ongoing ownership |
For the detailed tool comparison, read Power BI Dashboard vs Custom Dashboard.
Build rules
These rules stop the first dashboard turning into a museum of charts.
| Rule | Why it matters |
|---|---|
| One metric, one owner | A red number with no owner becomes background noise |
| Every percentage needs a denominator | Percentages hide bad data when nobody can see the base |
| Match freshness to action | Live data is wasteful when the decision is monthly |
| Show the records beneath the cards | Summary metrics do not tell the team what to do |
| Separate source quality from source volume | High lead count can still be weak commercial value |
| Keep stale work visible | Old leads and overdue jobs are usually where revenue and trust leak |
The lazy first version is a table with definitions, owners and decisions. Add automation after the table has proved useful.
FAQ
Is this real business data?
No. The CSV uses synthetic worked data. It is designed to show structure, formulas and decision logic without exposing client information or implying benchmarks.
Can I import the CSV into Power BI?
Yes. You can import it as a flat file and model from there. Before using live data, replace the synthetic rows with exports from your CRM, quoting tool, job system, call platform or spreadsheet.
What KPIs should a service business track first?
Start with new enquiries, qualified opportunities, first response time, quote acceptance rate, won revenue, open jobs, overdue work, unassigned work and capacity pressure.
How often should the dashboard update?
Match freshness to the decision. Sales follow-up and urgent operations may need same-day data. Board reporting may only need weekly or monthly refreshes.
Should this be a custom dashboard?
Only if the dashboard needs workflow actions, bespoke roles, customer access, embedded AI logic or a product-like experience. If it is internal reporting, a spreadsheet or BI tool may be enough.
Related reading
- ↑ Business Dashboard Development UK
- ↔ Operations Dashboard for Service Businesses
- ↔ Sales Dashboard for Lead Follow-Up
- ↔ Power BI Dashboard vs Custom Dashboard
- ↔ Internal Tool Development UK
Turn the template into a real dashboard
If your team cannot agree on the numbers, start with their definitions, sources and owners. When those are clear, our business dashboard service can help you decide whether the next step needs custom development.