Five years of spend. No measurement.

1Where the audit stands

Thirteen systems in scope. Four reviewed, one partial, eight still ahead.

2The target funnel

From your own figures, the current or desired funnel is:

The critical point: we currently have no way to verify any of these numbers from the platforms themselves. They exist in your team's knowledge and in HubSpot. Making them visible and measurable is the core outcome of this work.

3Three periods

To make performance legible, the account's history is split into three periods, derived from the Google Ads change history — which records who made which changes, and when.

Periods B and C overlap in July 2026. Altitude's last substantive change was 15–16 July; the two main campaigns were paused on 4 August. July is a transition month and appears in both.

4Google Ads Reviewed

How it is set up

Account 648-805-2521. Owned through NetValue's manager account, consistent with NetValue's ownership of Addurre Group.

The HubSpot manager link was added on 31 July — the native HubSpot–Google Ads integration, added at the same time Ben applied HubSpot UTM parameters across all campaigns.

Campaign structure — five of six paused

Conversion tracking

• 18 conversion actions exist. 17 have recorded zero conversions, ever.
• The one action with any value (request_demo_page_click, 603.28) is set to Secondary and excluded from account-level goals — so even that counts toward nothing.
• The only natively configured action is call-based. There is no working website conversion action.
• Customer data terms: not accepted. Enhanced Conversions: not configured.
• One action is sourced from Universal Analytics, discontinued by Google in 2023.
• Two campaigns carry Google's own warning: "conversion tracking setup is incomplete."

Period A — pre-agency baseline

Effectively one campaign ran — Website Traffic — on entirely generic keywords ("farm management software", "agriculture management system", "farm accounting"). No contractor-specific language, no NZ/AU differentiation.

This period is a cost benchmark, not a performance benchmark. It gives us roughly NZ$2.00 per click and a 1.43% CTR, and nothing else. Any claim about what Google Ads historically delivered for JobFlow cannot be substantiated from the account.

Period B — agency-managed

chelsea@altitude-marketing.co.nz was invited as Admin by Ben Meehan on 13 March 2026 and accepted the same day.

March 2026 spend not yet isolated — TBC. Conversions in period: 0.

What the agency actually did. 199 of the recorded changes came from one user, and the overwhelming majority were micro-adjustments to the DemandGen daily budget, moving in cent-sized increments around NZ$4–5/day. That is the signature of an automated bid strategy drifting, not of active management.

Substantive work in the period was limited to adding Queensland and Tasmania as regions (30 June), switching final URLs from jobflow.nz to jobflow.com (June), some keyword edits, and pausing a campaign (15 July).

What was not done: conversion tracking was never fixed. The account ran for four months, at a reported ~NZ$3,000/month agency cost, with Google's own interface displaying "conversion tracking setup is incomplete" throughout.

Period C — post-agency

On 4 August the Agriculture and Competitor campaigns were paused. Since then the account has been effectively dormant — the search keyword report for 15 August to 13 September shows near-zero impressions and clicks account-wide. Conversions in period: 0.

Targeting and creative

No AFIA or fodder-industry vocabulary appears anywhere — no fodder, pit or bunker silage, custom harvesting, TMR. For the Australian market that is the difference between contractors recognising themselves in the product and scrolling past.

A cluster of Competitor keywords sit at Quality Score 1 out of 10, the lowest possible, driving costs per click as high as NZ$17–23. "Low quality" appears as a status reason across the majority of keywords with recorded spend.

Competitors being bid on: Ag-drive, FarmBackup, Tradify, Figured, ServiceM8, Simpro, Planfarm, Cashmanager Rural, EasyFarm, Farm Focus. Several of these are general trades software, not agricultural — a broader comparison set than the competitive picture we had been working from.

Competitive visibility

Search impression share is under 10%. Two genuine agricultural competitors appear that were not on our radar:

Monday.com and Capterra appearing at all confirms the keyword problem independently: JobFlow's terms are broad enough to enter auctions against generic project-management software and software directories.

Outstanding issues

5Google Analytics Mostly reviewed

Property jobflow.co.nz – GA4, ID 331082986. Three data streams:

Configuration issues

The retention setting has a practical consequence: any analysis in GA4's exploration tools currently cannot look back further than about two months. Raising it to 14 months will not recover lost history — it only helps from the day it is changed. Worth doing today regardless of anything else here.

What is actually being measured

Tracking is not entirely dead. A full-year export (1 Sep 2025 – 31 Aug 2026) shows real activity:

Three things stand out

ScoreApp is generating more leads than the official demo funnel. The quiz tool at quiz.jobflow.com produced 20 completed leads over the year — more than contact_form_submit (9 users) and demo_landing_lead (9 users) combined. This tool appears in none of the stack documentation provided.

A probable naming mismatch is breaking conversion counting. Events fire as demo_booked_thanks and demo_meeting_engaged, but the conversion actions and tags are configured against demo_booked and meeting_engaged. If the live event name does not exactly match the configured name, GA4 counts nothing. This may be a direct, fixable cause of part of the zero-conversion problem — to be confirmed in testing.

Outstanding issues

6Tag Manager Mostly reviewed

Container JobFlow Website (GTM-WFKHTTMV), quality rated Good by Google's own scoring. A single Google tag object named "jobflow.co.nz – GA4" carries the NZ measurement IDs and distributes data to both GA4 destinations. This is where the confusing property naming originates.

Issues found

A scoping bug in the "Global" triggers

Triggers named Global are hardcoded to NZ-specific values — Global Support Email fires only on mailto:support@jobflow.nz, Global Sales Email only on mailto:info@jobflow.nz.

If jobflow.com uses different addresses, these cannot fire for jobflow.com visitors at all. If it reuses the NZ addresses they will work — but that raises a separate question: does jobflow.com have its own email aliases? If not, that is worth setting up regardless, both for tracking clarity and for how the Australian market reads the business.

Four tags are explicitly paused

• Custom Event – Contact Form Submit
• GA4 – demo_booked
• GA4 – hubspot_meeting_interaction
• GA4 – meeting_engaged

Two are HubSpot meeting-tracking tags. Someone previously built HubSpot meeting tracking into GA4 and it is currently switched off — directly relevant to the "HubSpot not properly connected" concern.

A Google-flagged technical defect

Tag Diagnostics reports "Config command out of order" — some pages send events before the configuration command runs. Google's own description warns this may result in unexpected behaviour while routing and processing those events. A plausible contributor to data loss beyond simple misconfiguration.

Also confirmed: the Meta Pixel is live and firing on all pages, giving a cross-check point for the Meta review. The container has a single administrator — a continuity risk flagged by Google itself.

7Search Console Partial

Four structured data alerts received:

The bad escape sequence is a genuine code defect — malformed markup, not just incomplete data — and Google flags it as preventing the page from appearing in relevant search results.

8Meta Scheduled next

9HubSpot CRM Scheduled

10Other systems TBC

11Decisions needed from JobFlow

These block progress and can be resolved quickly.

12Action log

Consolidated and ordered across all platforms. Sequencing matters — the immediate block unblocks everything below it.

13Seven-month holding plan

Draft for discussion. Costs and sequencing to be firmed up once the remaining platforms are reviewed.

Indicative budget

Working figures for discussion, not a quote.

A note on the arithmetic. If the target is 20 leads a week (~87/month) and paid channels supply roughly half of those, at NZ$2,500 combined monthly ad spend that implies a cost per lead around NZ$55–60. Against your model — NZ$1,500–3,000 onboarding, then NZ$250/month plus operator fees — that is comfortably viable if conversion rates hold. It is not a promise. It is arithmetic worth testing, and we will not know the real numbers until tracking works. Which is precisely why Phase 1 comes first.

14Growth pathway

This audit was deliberately scoped to stabilise and optimise. Sustained growth in the AU/NZ market warrants a deeper piece of work, scoped separately. What the current engagement is not covering:

• Positioning and narrative. The market research shows Australia is structurally different from New Zealand — horticulture rather than dairy, a feedlot forage supply chain worth AU$5–8b with no NZ equivalent, and contractors who self-perform far more. The current advertising treats both markets identically.
• Segmentation. Which contractor profiles to pursue first, in which geography, with what message. The AFIA vocabulary work in the targeting guide is a strong start that has never reached the advertising.
• The ScoreApp question. A quiz tool is quietly outperforming the official demo funnel. Understanding why may be more valuable than anything in the paid accounts.
• Channel expansion. Your own marketing scorecard shows 4 of 11 channels active. Podcast, editorial, YouTube sponsorship, industry awards and self-hosted events are all unused.
• The 83% form abandonment. The largest single measurable leak found so far, sitting upstream of every channel.