Executive summary
Three things at once
- Budget allocation is working against us. Over the last 30 days, spend rose A$7.88K but conversions rose only +4. Blended cost-per-conversion went from ~A$271 to A$359 (roughly +33%). Most of the new money flowed into the two ad groups the March 3 action plan flagged for reduction.
- The Tuesday site-configuration issue made the numbers worse than they look. Until the fix on Tuesday afternoon (Apr 21), the site wasn't catching ads properly. Some conversions weren't reported back to Google Ads, which means smart bidding on Generic Search has been learning from incomplete signal.
- Retargeting readiness is mostly unfinished and the audience clock just restarted. With the tag now firing, the remarketing pool builds from Tuesday afternoon forward. Used well, the next two weeks become a setup window, not a wasted one.
30-day snapshot
Google Ads account overview, Mar 23 – Apr 21, 2026 vs. prior 30 days (Feb 21 – Mar 22).
Put simply: a ~40% increase in spend produced a ~5.5% increase in conversions, with CPA rising a third. That pattern is the signal to pay attention to, though as we'll explain, the Tuesday tracking fix means some conversion volume is likely still missing from the report and the true picture is slightly better than the dashboard shows.
Where the new budget went
Looking at the ad-group cost deltas vs. the prior 30 days, this is the clearest view of how the increase was distributed.
| Ad group | Campaign | Δ Spend (AUD) | Δ % | March 3 plan said |
|---|---|---|---|---|
| Construction Camera | Generic | +A$3,515.81 | +122.7% | 1.3x ROAS — reduce |
| Construction Site Time Lapse Camera | Generic | +A$3,295.19 | +59.3% | 0.7x ROAS — reduce |
| Oxblue | Competitor | +A$464.83 | +57.9% | 21.5x ROAS — expand |
| Brinno | Competitor | +A$459.16 | +158.9% | 55.6x ROAS — expand |
| Solar Powered Time Lapse Camera | Generic | −A$249.11 | −27.1% | 11.2x ROAS — expand |
| Brand | Brand | +A$247.70 | +10.8% | Match-type cleanup first |
The structural problem
Construction Camera and Construction Site Time Lapse Camera absorbed roughly A$6,811 in new spend combined, which is ~86% of the total budget increase. Both were flagged on March 3 as under-performers to starve. Meanwhile, Solar Powered Time Lapse Camera — one of the strongest ROAS ad groups at baseline — was cut by A$249.
Campaign-level breakdown
| Campaign | Spend | Δ Spend | Conv. | Δ Conv. | CPA | Δ CPA |
|---|---|---|---|---|---|---|
| Generic Search | A$22,654 | +A$6,622 | 42.5 | −4.5 | A$533 | +A$192 |
| Brand Search | A$2,540 | +A$248 | 26.5 | +4.5 | A$96 | −A$8 |
| Competitor Search | A$2,440 | +A$1,006 | 8.0 | +4.0 | A$305 | −A$53 |
| Buildcam YouTube (discontinued) | A$0 | — | 0 | — | — | — |
Brand and Competitor are scaling efficiently. Generic is the problem: more money, fewer conversions, CPA up 56% at the campaign level.
The Tuesday tracking fix, and what it changes
On Tuesday afternoon, April 21, a site-configuration issue was identified and fixed. Before the fix, the site wasn't catching ads properly. The specific tags affected are still being confirmed, but the symptoms visible in both Google Ads and GA4 all line up with the same underlying cause.
Three independent signals, one root cause
| Observed signal | Source | Most likely tie-in |
|---|---|---|
| 12 conversion actions flagged "no recent conversions" (only 2 actively recording) | Google Ads → Conversion tracking status | Conversion tags not firing reliably prior to Tuesday |
| GA4 "Unassigned" channel +3,600% (7-day view ending Apr 21) | GA4 home | UTM parameters or landing-page redirects dropping source data |
| No retargeting audience accumulating despite increased spend and visitors | Inferred from tag status + GA4 state | Remarketing tag not present or not firing across pages |
Note: items #1 and #2 are directly observed in the reporting tools. Item #3 is a logical consequence of items #1 and #2, not a metric we can read back directly until the audience has rebuilt.
What this means for the 30-day numbers
Some conversions that occurred during the broken window almost certainly never reported back to Google Ads, which means: (a) the dashboard CPA above is a worst-case view rather than a clean read, and (b) smart bidding on Generic has been optimizing against incomplete signal. The budget-allocation problem is still real and still visible from spend distribution alone, but we should wait 7 to 14 days of clean post-fix data before judging whether the corrective actions are working.
Three other signals worth flagging
1. Google Ads and GA4 are not linked
GA4 is actively prompting the account link. Until this is in place, Ads-based attribution in GA4 is partial and no GA4 audiences can feed back into Google Ads for bidding or remarketing. This is a one-click, one-minute task on the GA4 side.
2. Brand search impression share is 75.7%
Roughly one in four branded searches is going to a competitor's ad instead of Buildcam's. Impression shares on Competitor (62.6%) and Generic (70.3%) are also well below ideal. Brand impression share is the easiest lift in the account and should sit above 90%.
3. Two active Google recommendations that should not be accepted yet
The account is currently showing two prominent recommendations:
- "Set a target CPA, raise your budget +17.5%" — accepting this before fixing allocation will simply pour more money into the under-performing Generic ad groups.
- "Use Display Expansion" — worth evaluating only after retargeting is in place, so Display benefits from warm-audience signal rather than cold.
Recommended next moves
Ordered by what moves the numbers fastest and what needs to happen in sequence.
- Reallocate Generic budget. Cap Construction Camera and Construction Site Time Lapse Camera back to pre-increase levels. Redirect the freed budget to Solar Powered Time Lapse Camera, Oxblue, Brinno, and the Brand match-type cleanup.
- Audit the 12 stale conversion actions. Confirm which ones stopped firing, and verify the
/instant-quoteconversion tag is firing cleanly after Tuesday's fix. Prune any that are obsolete. - Link Google Ads to GA4. One-click fix from the GA4 home prompt.
- Reconcile Mar 23 – Apr 21 conversion data. Compare Webflow form submissions and Pipedrive deals against what Google Ads reported to quantify how much credit was lost during the broken window. This gives the "real" baseline we measure the fix against.
- Tag-health pass. Tag Assistant or GTM preview across homepage,
/instant-quote,/use-cases,/industries, and/featuresto confirm all necessary tags fire post-fix. - Raise Brand impression share. Move Brand keywords to exact / phrase match. No budget change required.
- Decline the two Google recommendations currently shown in the account until steps 1 and 2 are complete.
Retargeting readiness checklist
Status as of today. This is where "ensure we're ready for retargeting" actually lands.
| # | Readiness item | Status | Owner |
|---|---|---|---|
| 1 | Google Ads ↔ GA4 account link | Not linked | Noah (1-click) |
| 2 | Remarketing tag firing site-wide | Fixed Tue PM — verify | Noah to confirm via Tag Assistant |
| 3 | Audiences built (e.g., visited /instant-quote but didn't submit) |
Pending #1 | Local Nerds |
| 4 | Audience size thresholds (typical Display minimum) | Rebuilding from Tue PM | Time-gated |
| 5 | Customer Match list from Pipedrive (hashed email / phone of past buyers and qualified deals) | Not uploaded | Local Nerds to prep, Noah one-line green-light |
| 6 | Lookalike / audience-expansion layer seeded from Customer Match | Ready to prep after #5 | Local Nerds |
Items #4 and part of #3 are direct consequences of Tuesday's fix. The clock on building a usable audience effectively restarted on Tuesday afternoon.
Lookalike / audience expansion
Once the Customer Match list is uploaded (item #5), we can use it as the seed for a lookalike-style layer via Google Ads audience expansion. Given we already have enriched Pipedrive data (matched deals, lead sources, ad groups) from the February and March imports, we're in a strong position to build a high-quality seed audience — not just a generic email list. That gives the expanded audience a much sharper match signal than a cold lookalike.
Timeline & how we'll know it's working
Dates that depend on external factors (audience size thresholds, creative turnaround) are estimates, flagged below.
Primary metrics we'll track
- Blended CPA — target: return toward the pre-increase range of ~A$270.
- Generic-campaign conversion volume — target: reverse the current −4.5-conversion decline.
- Brand impression share — target: > 90%.
- Conversion recording coverage — target: 12+ conversion actions active, > 95% tag-firing health.
- Ratio of Webflow form submissions to Ads-reported conversions — target: > 90% reporting back.
- Remarketing audience size — target: cross Display threshold before mid-May.
What we need from Noah
Kept deliberately light. Everything else on the action plan is Local Nerds' to run.
-
1Link Google Ads to GA4.
-
2One-line reply on the scope of Tuesday's fix.
-
3Green-light Customer Match and the lookalike layer.
Data provenance
So you can audit any figure in this walkthrough:
What is directly observed
- All 30-day Google Ads numbers (spend, clicks, conversions, CPA, per-campaign and per-ad-group figures, impression shares, conversion-tracking health) read on April 22 from the Google Ads account Overview for Buildcam, window Mar 23 – Apr 21, 2026 vs. prior 30 days.
- All GA4 figures (active users, channels, pages, events, form_start) read April 22 from the Buildcam GA4 property home.
- ROAS baselines for each ad group (Construction Camera 1.3x, Construction Site TLC 0.7x, Solar Powered TLC 11.2x, Oxblue 21.5x, Brinno 55.6x) are pulled from the March 3, 2026 Buildcam PPC Action Plan prepared by Local Nerds.
What is inferred
- Claim that "some conversions weren't recorded" — logical consequence of the Tuesday tracking break, not directly measured. The reconciliation against Webflow and Pipedrive (action plan step 4) is what will quantify it.
- Claim that "no usable retargeting audience was accumulating" — inferred from the tag state, not directly readable.
- Retargeting go-live dates (~May 5 for RLSA, mid-May for Display and the audience-expansion layer) — estimates based on typical audience-build timelines, contingent on traffic volume and Google's audience-size minimums.
What is still to confirm
- Exactly which tags were affected by the pre-Tuesday site-configuration issue.
- Whether the March 7 action-plan items (
/instant-quotedestination routing and{adgroupname}→{adgroupid}template fix) actually shipped. - Readiness of Pipedrive data for Customer Match upload under Google's hashing requirements (EU consent obligations do not apply — Buildcam does not serve the EU).