Quick answer: Use ten clearly defined operating numbers to review the year: completed installs, repeat customers, active service agreements, referrals, cameras per project, callbacks, first-response time, genuine reviews, larger projects and permission-based marketing contacts. Compare like with like, investigate the causes and choose three practical actions. These measures complement your financial accounts; they do not replace them.
On this page
Set the review boundary before counting
Start with the year and the unit you are counting. A company may have several sites and several jobs; those are different things. Give each customer, site and job an identifier so a repeat visit does not become a new customer by accident. Keep installation jobs separate from service work.
Use the installation documentation guide to keep consistent customer/site/job IDs and installation handover dates. Record service visits, cancelled jobs and open work separately. Count physical cameras separately from optical views or NVR channels. If a definition changes, explain it and restate comparable data where possible; a different counting rule is not an improvement on site.
| Measure | Definition for this review | Record needed |
|---|---|---|
| Completed installs | Installation jobs handed over during the year. | Job ID and handover date. |
| Repeat-customer rate | Distinct customers served this year who also bought before this year ÷ distinct customers served this year. | Customer ID and earlier completed work. |
| Active agreements | Agreements in force at the review date; show renewals separately. | Start, end and renewal status. |
| Referral share of new customers | New customers attributed to an agreed referral source ÷ all new customers. | First-source record, including unknown. |
| Cameras per project | Cameras installed ÷ completed installation jobs. | Consistent included camera count. |
| Callback rate | Eligible installs with at least one qualifying callback ÷ eligible installs. | Handover, issue, cause and visit date. |
| First-response time | Median elapsed time to a useful first human response. | Request, response and coverage hours. |
| New reviews | New genuine reviews received during the year. | Platform and review date; no review filtering. |
| Larger projects | Jobs meeting a size or complexity threshold set before review. | Camera count, buildings and scope type. |
| Marketable contacts | Unique contacts with a recorded, currently valid marketing basis. | Consent/basis, jurisdiction and suppression status. |
Make callback and response figures comparable
A December installation may not yet have completed a thirty-day callback observation period. Either wait until its window closes or mark it pending and exclude it from both parts of the calculation. Count an affected job once for the rate, then track its repeat visits separately. Otherwise a single difficult site can distort the picture.
Define qualifying causes: setup mistakes, missed testing and incomplete handover might belong in the install-quality measure. Customer damage, approved additions and scheduled maintenance need separate categories. Keep unresolved causes visible rather than forcing them into the most flattering category.
Use the service-agreement outline to define the response clock and covered hours, then measure the actual agreed service. Show uncovered after-hours requests separately. An automatic acknowledgement is not a useful first human response. Compare the median with the oldest unresolved requests so an apparently good median does not hide a neglected customer.
Use a simple job sheet and check the totals
Keep controlled job IDs, type, handover date, camera count, source and callback status, with a separate request/response log where needed. Restrict access and keep passwords, footage and unnecessary personal information out of review exports. A bookkeeping or CRM field does not by itself prove the job, permission or service event occurred.
When filtering a spreadsheet date column, use the start of the year through, but not including, the start of the following year. This includes records with a time on December 31. Use ordinary straight quotation marks in formulas, and check your spreadsheet locale and column types before copying any expression.
Use the Google Sheets CRM guide to retain source counts beside each rate. For example, five eligible jobs with a qualifying callback out of forty eligible jobs gives 5 ÷ 40 × 100 = 12.5%. This is arithmetic, not a benchmark or a claim about your business. If the denominator is zero, report N/A rather than a fabricated percentage.
Turn the review into three actions
Read pairs of measures before drawing conclusions. More installs and more callbacks may indicate rushed commissioning, but the job mix or newly reported issues may also explain the change. More cameras per project does not prove higher profit. Review revenue, direct costs, overhead and cash collection with your accountant alongside these operating measures.
| Measure and finding | Action / owner | Review date and evidence |
|---|---|---|
| [Measure, counts and likely cause] | [One specific change; named owner] | [Date; result to inspect] |
| [Second measure] | [Action and owner] | [Date and evidence] |
| [Third measure] | [Action and owner] | [Date and evidence] |
For example, investigate repeat setup errors with a bench-test and handover check, rather than asking the crew merely to “reduce callbacks.” Review service-agreement renewal records before assuming customers rejected the offer. Use documented maintenance visits to understand what customers actually need.
A quarterly review can reveal process issues while records are recent. Anabon’s security camera installer program describes a route to professional account support; confirm available help for the actual product/documentation question. Do not send the company’s entire customer database or private financial workbook, and do not treat account access as a guarantee of business results.
FAQ
Which three KPIs should a small installer start with?
Choose measures you can define and act on, such as eligible callback rate, first-response time and agreement renewals. There is no universal target for every company.
Does a larger project mean a better business?
Not necessarily. Compare complexity, delivery capacity, cost and payment risk as well as camera count. Growth that cannot be commissioned or supported properly needs investigation.
References
- SBA — business bookkeeping and financial management
