What Website Analytics Should a Small Business Track?
Choose a small set of website analytics that connects useful visits with confirmed enquiries and business outcomes without collecting unnecessary data.
Part of our guide: How to Measure Whether Your Business Website Is Actually Working
A small business should track the measures it can use to make decisions: relevant visits, important page journeys, confirmed enquiries, lead quality and technical failures.
You do not need to record every movement or read every report. Start with a clear business question, then collect the least information needed to answer it.
Choose an outcome before a tool
For a service website, the primary outcome may be a suitable quotation request. For a training provider, it may be a completed booking. For an existing-customer resource, it may be successful access to instructions.
Write down the definition and how it will be verified. A click on a booking button is an attempt; a booking-system confirmation is evidence of completion.
The website measurement guide connects these stages.
Thinking about your own rebuild?
Tell us about your business, your current website and what you would like to improve. We will help you work out the next step.
A useful starting set
Track:
- Sessions reaching important service pages.
- Landing pages and broad traffic sources.
- Confirmed form submissions or bookings.
- Telephone-link clicks, clearly labelled as attempts.
- Qualified enquiries recorded by staff.
- Errors that prevent an important action.
- Sales or appointments resulting from those enquiries where reasonably attributable.
Add device category when it helps find usability problems. Avoid collecting dimensions merely because the software makes them available.
Understand users, sessions and events
A user is the analytics system’s estimate of a visitor identity. A session groups activity into a visit according to the tool’s rules. An event records an interaction.
These are not interchangeable counts. One person may visit several times or appear as different users across devices. One session can contain several events.
Write the denominator beside each rate. A session enquiry rate cannot be compared directly with a rate based on individual users or advertising clicks.
Validate important events
Google Analytics documents key events as events important to business success. Configure the relevant event and test the actual implementation rather than assuming automatic form detection proves delivery.
Check a successful submission, a failed submission, a page refresh and a repeat request. Ensure the event does not fire when validation prevents sending.
Where possible, reconcile analytics with the destination system. Exact totals may differ because of privacy choices or blocked scripts, but large unexplained differences deserve investigation.
Keep personal information out
Do not send names, emails, telephone numbers or free-text messages in event names, page URLs or analytics parameters. Google’s policy guidance prohibits sending information it can recognise as personally identifiable.
Store enquiry details in the appropriate business system with suitable access controls. Use aggregated analytics to understand patterns.
Review applicable privacy and storage-access requirements for the actual configuration. A tool’s marketing label does not establish that every feature can run without consent.
Add context to traffic reports
Separate paid campaigns from organic discovery and direct visits where the evidence allows. Use consistent campaign tags without putting personal data in URLs.
An increase in visitors can mean a successful campaign, irrelevant traffic or automated activity. A decrease can follow the removal of low-value traffic while qualified leads stay stable.
Look at outcomes alongside volume. See traffic without enquiries for a practical diagnosis.
Do you need heatmaps and recordings?
They can help investigate a specific interaction, but they add data collection and review work. Do not install them automatically on sensitive pages.
Often a few observed usability sessions, customer questions and error reports reveal the problem more directly. If you use recordings, configure masking, access, retention and any required permissions carefully.
Make the monthly review short
Use one page showing the outcome, trend, confidence limits and next action. Note changes in campaigns, prices, staffing and tracking that affect comparison.
Remove a metric if nobody uses it. Add one only when it answers a question the current dashboard cannot resolve.
For page contribution, see which pages generate leads. For a rate you can interpret, see small-business conversion rates. Contact AfterWorks to establish useful measurement around a website improvement.
