Test the change, measure the calls, then decide

Newsroom

Reading GA4 for HVAC leads: five settings that change what you see

GA4 rarely lies, but it often answers a slightly different question from the one you asked. Five settings and behaviours explain most of the confusion HVAC owners run into.

Written by Ali Sedighi, MBAPublished: Last updated:

GA4 is a useful instrument for an HVAC business if you know what each report is counting. Most confusion comes from five things: which events are marked as key events, data thresholds, the retention setting, the attribution settings, and the gap between GA4 and Google Ads. Each is explained below from Google's own help pages, as loaded in October 2026.

This is the research-and-analysis step of our method, and it is where our conversion research work begins.

1. Key events decide what "conversion" means

Google's key events help describes a key event as any collected event you mark because it matches an action important to your business. Once marked, it appears in reports so you can compare channels.

That means you decide what counts. A typical HVAC setup might mark:

  • connected calls from the website (via your call tracking)
  • quote or estimate requests
  • text-us taps, if you take texts

What should usually not be a key event: page scrolls, video plays or visits to the contact page. Marking soft actions inflates totals and makes every channel look productive. When we audit accounts, an over-long key event list is one of the first things we fix in measurement setup.

2. Data thresholds hide small rows

The data thresholds page explains that GA4 may withhold data when a report includes demographics or when too few users sit behind a row, to stop anyone identifying individuals. Google says these thresholds are system-defined and cannot be adjusted.

For an HVAC company in a small market, that means reports broken down by age or interests can go patchy fast. What helps:

  • widen the date range, which Google suggests
  • remove demographic dimensions when you only need channel and page data
  • read totals, not thin slices; a row of three users means very little anyway

3. Retention limits explorations, not standard reports

Google's data retention help states that the retention setting affects explorations and funnel reports only. For standard properties, the options are 2 months or 14 months; larger paid properties can keep more. Age, gender and interest data are always limited to two months.

The practical consequence: if retention is set to two months, you cannot build an exploration comparing this heating season with last year's. Check the setting now, before you need the history. Changing it later does not bring back what has already expired.

4. Attribution settings change which channel gets credit

The attribution overview lists three reporting models: data-driven, paid and organic last click, and Google paid channels last click. It notes that rule-based models such as first click and linear were retired in November 2023, and that direct visits get no credit unless the whole path is direct.

Two habits keep reporting honest:

  • Note the model on every report. A month-on-month comparison is meaningless if someone changed the model in between.
  • Know who can change it. The settings sit under Admin, and the help page says the Marketer role or higher is needed at property level.

Our call tracking article covers how to choose a model for a phone-led business.

5. GA4 and Google Ads will never match exactly

Google says you can create a Google Ads conversion from a GA4 key event, which helps keep counts consistent, but also that Google Ads conversions do not appear in standard GA4 reports. Each tool counts on its own terms.

Expect gaps from:

  • different attribution models and lookback windows
  • calls from ads that never touch your website, which GA4 cannot see
  • people who decline analytics cookies or block scripts

Do not choose the bigger number. Reconcile both against the office call log once a month.

A 20-minute GA4 health check

  1. List your key events. Remove anything that is not a lead.
  2. Check data retention and set it to the longest option your property allows.
  3. Write down the reporting attribution model and the date it was set.
  4. Run a lead report by landing page for the last 90 days. Do the top pages make sense?
  5. Compare one month of key events against your call log.

If any step surprises you, that is your first test hypothesis. Browse more method pieces in the newsroom, or call, text or email to go through your property with us.

Frequently asked questions

Why does my GA4 exploration only go back a couple of months?

Probably the data retention setting. Google's help says retention affects explorations and funnel reports, with options of 2 or 14 months for standard properties. Standard aggregated reports are not affected, so the same numbers may still appear there.

Why are some rows missing from my report?

A data threshold may have been applied, usually when demographics or small counts are involved. Google says thresholds are system-defined and cannot be adjusted, but widening the date range can bring rows back.

Should I trust GA4 or Google Ads for lead counts?

Neither on its own. They use different attribution and counting rules, and Google notes that Ads conversions do not appear in standard GA4 reports. Treat your office call log as the reference and use both tools to explain it.

Sources

  1. Google Analytics Help: About key events (2026-10-23)
  2. Google Analytics Help: About data thresholds (2026-10-23)
  3. Google Analytics Help: Data retention (2026-10-23)
  4. Google Analytics Help: Get started with attribution (2026-10-23)