Why trades ads live or die on intent
Most wasted trades spend comes from treating every search the same. "Emergency electrician" and "how much does rewiring cost" are not the same person. One wants a van at the door in the next hour; the other is six weeks from a decision. Pay top dollar for the first and almost nothing for the second.
So the account isn't organised by keyword theme — it's organised by intent. The structure exists to route budget to the searches that turn into booked jobs today, and to stop the research traffic from quietly eating the budget that should have caught the emergency call.
The Search account structure we run
A trades account needs fewer moving parts than most agencies build. One Search campaign per service line, ad groups split by intent tier, a cheap brand campaign, and a single contained Performance Max cell. Over-segmenting starves each campaign of the conversion data Google needs to optimise.
Here is the shape we run on almost every trades account, before tuning to the specific trade and market:
| Campaign | Example ad groups | Match & bid | Why it is split out |
|---|---|---|---|
| Emergency | "emergency plumber", "24 hour electrician", "no power" | Exact + phrase, top-of-page bid | Highest intent and margin — never shares budget with research traffic |
| Repair / fix | "leaking tap repair", "ducted AC not cooling" | Phrase, mid bid | Real intent, slightly longer decision — own it without overpaying |
| Install / quote | "new hot water system", "switchboard upgrade quote" | Phrase + broad with audience signals | Higher ticket, longer cycle — send to a quote page, not the home page |
| Brand | "[your business name]" | Exact, low bid | Cheap insurance against competitors bidding your name |
| Pmax (test cell) | Asset groups per service line | Capped budget, target ROAS | Finds new queries and fills gaps — contained so it cannot cannibalise Search |
Performance Max as a test cell, not a crutch
Performance Max is good at finding queries you didn't think to bid on, and bad at being controlled. Left uncapped, it will happily spend your emergency budget on cheap, low-intent placements because they convert something. So we run it as a test cell: a capped budget, a target ROAS, and a clear job — discover, don't dominate.
Anything Pmax proves out — a new query, a new service line — graduates into a structured Search ad group where we can actually steer it. Pmax scouts; Search closes.
Call tracking and the missed-call backstop
In trades the money comes in over the phone, and that breaks most ad accounts. Without call tracking, Google can't see which clicks became jobs, so it optimises toward form-fills and cheap clicks instead of booked work. Dynamic call tracking feeds those conversions back so the algorithm bids toward revenue.
The second leak is the call nobody answered. A tradie on a roof can't pick up — and that caller rings the next result. A missed-call text-back agent closes the gap: it texts back within seconds, holds the lead, and books or routes them to a human. The same structure that wins the click has to survive the moment the phone rings.
Where to start
If you're already running Google Ads, the fastest win is usually to split your one catch-all campaign by intent tier and put call tracking in before you touch budgets. If you're starting fresh, build the emergency campaign first — it's the highest-margin traffic and the quickest to prove the unit economics.
That's the build we ship in a sprint: the account structure, the tracking, and the missed-call backstop wired together. The audit tells us whether the numbers justify it before anyone spends a dirham.