Skip to content
Proxpat

(18) — Work

Proof, when we can evidence it.

We do not publish results we cannot show you the account for. Client-permissioned case studies are being prepared. Until they are ready, here is the next most useful thing: exactly how we work, and the checklist we run before recommending anything.

STAGE0103The auditFUNNELUNDIAGNOSED

The checklist, published.

Every agency says it runs an audit. Here is ours, in full. Run it yourself if you like — if you find nothing, you do not need us, and you have saved a phone call.

Measurement

  • GA4 installed once, not twice, with key events marked as conversions
  • Form submissions and calls recorded as events, not thank-you page views
  • Google Ads conversions imported and de-duplicated against Analytics
  • Consent handling that does not silently discard the data it claims to collect

Technical

  • Indexability: robots rules, canonicals, noindex left on by accident
  • Sitemap that matches the site, without noindexed or redirecting URLs in it
  • Core Web Vitals measured on field data, not a lab score
  • Mobile rendering and tap targets on a real handset

Content & search

  • Whether each page matches what the search actually wants
  • Pages competing with each other for the same term
  • Thin or duplicated service and location pages
  • Which searches are winnable against who currently ranks

Local

  • Google Business Profile completeness, per location
  • Name, address and phone consistency across directories
  • Review volume, recency and whether anyone replies
  • A landing page per location that is not a copy of the last one

Paid

  • Search terms report — what you are actually paying for
  • Match types and the negative keyword list
  • Ad group to landing page relevance
  • Budget pacing and where spend is being wasted this month
CHECKS20 PUBLISHEDDIAGNOSTICNO CHARGE
STAGE0203ScenariosSIGNALLOW

Four situations we recognise.

Illustrative, not client work, and every card says so. They are here because the pattern is the useful part: what we look at first in each, and what we would measure. No figures, because there is no account behind them to check them against.

STAGE0512ProofSCENARIOS / 04

Proof, when we can evidence it

SCENARIOS04
0104ILLUSTRATIVENOT CLIENT WORK

Real estate

A brokerage losing to the portals

The situation
Listings sit on Bayut and Property Finder, so the portals rank for every property search and the brokerage’s own site gets nothing. Ad spend goes to portal leads shared with four other agents.
What we check first
Which searches the site could realistically win — area guides, building-specific pages, off-plan payment plans — rather than the head terms the portals own outright.
The approach
Concede the listing searches. Build the pages the portals do not: honest area comparisons, service-charge realities, what a specific tower is actually like to live in. Route paid spend to those pages instead of generic lead forms.
What we’d measure
Enquiries arriving direct rather than via portal, and cost per enquiry compared against the portal lead price.
0204ILLUSTRATIVENOT CLIENT WORK

Clinics & healthcare

A multi-branch clinic invisible in the map pack

The situation
Three branches, one Google Business Profile between them, and a booking flow nobody has opened on a phone. Most patient searches are on mobile and near-me.
What we check first
Profile completeness per branch, NAP consistency across directories, review volume and recency, and the booking path timed on an actual handset.
The approach
A separate, fully completed profile per branch with its own landing page. Fix the booking flow before spending anything on traffic — sending paid clicks into a broken form is paying to lose patients.
What we’d measure
Map pack visibility per branch, direction requests and calls from the profile, and completed bookings rather than page views.
0304ILLUSTRATIVENOT CLIENT WORK

Professional services

A B2B firm paying for the wrong clicks

The situation
Google Ads running on broad match with no negative keyword list. The search terms report is full of job seekers, students and competitors. Cost per lead is high and nobody knows which leads closed.
What we check first
The search terms report before anything else, then whether form submissions are tied back to the campaign that produced them.
The approach
Tighten match types, build the negative list properly, and align each ad group with a landing page that answers that specific search. Import closed deals so bidding optimises toward revenue rather than form fills.
What we’d measure
Cost per qualified enquiry, and the share of spend on searches with genuine commercial intent.
0404ILLUSTRATIVENOT CLIENT WORK

Any sector

The engagement that starts with broken tracking

The situation
The most common one. Conversions double-counted, thank-you page views recorded as leads, Ads and Analytics disagreeing by a wide margin. Every reported figure is wrong in the same optimistic direction.
What we check first
Fire the events manually and compare what arrives in each platform against what actually happened.
The approach
Rebuild measurement before touching campaigns or content. It is the least impressive work we do and the reason everything after it holds up.
What we’d measure
Platforms agreeing with each other, and a lead count that matches the inbox.
CHECKS20The checklist, published
SCROLL TO ADVANCE — 04 SCENARIOS
STAGE0303When they are readyMETHODFIXED

What each case study will carry.

Three sections, in this order, with nothing that cannot be checked against the client’s own accounts.

  1. The starting point

    What the diagnostic found: broken tracking, thin pages, wasted spend — wherever the money was leaking before we touched anything.

  2. The system we built

    Which services we activated, in what order, and why. What was built, what was refined, and what we turned off.

  3. The numbers

    Enquiry quality, visibility, campaign performance and cost — reported the way the client’s own dashboard reports them, so they can be checked.