Are Google Local Services Ads Worth It Follow the Lead Past the Dashboard

Google Local Services Ads are worth testing when an eligible business can answer enquiries reliably and turn the right contacts into profitable work. They are worth scaling only when those jobs survive the maths after suitability, scheduling and delivery costs enter the picture.

A dashboard full of leads does not answer that question. Google has to decide whether a contact qualifies under the product’s rules. Your business has to decide whether the resulting work is worth doing. Those are different decisions with different denominators.

That means a specific service, sensible area, workable response model and known job economics, not “LSAs seem to be going well”.

Google sells the contact, not the customer

Local Services Ads place eligible providers in a paid local-search product. Google says the ad may show information such as services, service area, hours, reviews and verification details, with potential customers making contact through the product (Google’s Local Services Ads overview). The advertiser is buying contact with demand. The platform is not guaranteeing that the person fits the company’s minimum job value, accepts its price, can be scheduled efficiently or becomes profitable work.

That distinction sounds obvious until a report uses valid lead and qualified lead as though they mean the same thing. A contact may be valid for billing purposes and commercially unsuitable for the business. Equally, a relevant contact may be lost because the phone went unanswered. One result concerns targeting and fit; the other concerns the operation that received the opportunity. I would not call either result a verdict on the channel until they are separated.

Google’s product controls, charging rules and available lead-resolution processes can vary by market, service and account. Its guidance explains how leads are charged and assessed, but the commercial judgement remains with the advertiser (Google’s guidance on Local Services Ads leads). Check the current controls visible in the actual account rather than building a UK forecast from an old tutorial recorded for another region.

Define useful work before the first call arrives

If the team cannot classify a contact consistently, “lead quality” will become whatever protects the opinion of the person presenting the report. Before launch, define a suitable opportunity in operating terms. The definition might include:

• a service the business currently provides;
• a location it can serve economically;
• a job type and likely value that fit the commercial model;
• a prospect with a genuine need and a usable contact route;
• a timescale the operation can meet;
• no obvious duplication, solicitation, spam or irrelevant request.

Then keep the outcome stages separate: received, answered, contacted, suitable, booked, completed and paid are not synonyms.

My preference is to define the Commercial Survival Line: the furthest stage a contact must reach before it earns the label used in the budget decision. For a capacity-constrained trade, the line might be completed work above an agreed contribution threshold. For a longer professional-service sale, it might be a sales-accepted opportunity with a defined need and authority. The exact line belongs to the business, not to a universal agency benchmark. The point is not to manipulate the definition until the campaign looks good. Set it before the result arrives and record exceptions rather than moving the goalposts.

Follow every charged contact through the Lead Survival Map

I use a Lead Survival Map to prevent the platform event from inheriting a commercial outcome it has not earned. Each stage asks one question and exposes one kind of failure:

Take a bounded cohort from a stable period and inspect every contact. Record the lead type, service requested, area, charged amount, response outcome, suitability reason, booking, completion and the commercial fields the business is authorised to use. Use controlled reason codes for rejection: outside service, outside area, below threshold, duplicate, unreachable, spam or another specific reason the team can apply consistently. Keep a short note for edge cases. A single broad label such as “bad lead” is emotionally satisfying and analytically useless.

The first useful output is not an average. It is the distribution of failure. Which services produce suitable work? Which areas consume travel? Which contact types go unanswered? Which requests repeatedly fail after quotation? That pattern tells you what to repair, narrow or stop.

Cost per lead is the beginning of the calculation

Cost per lead is easy to calculate and dangerously easy to admire. Two periods can produce the same figure while creating entirely different businesses. In one, the jobs are local, bookable and above the minimum value. In the other, the team spends hours returning calls for marginal work across a wide service area. The dashboard calls them equal because it is counting the charged event, not living with the job. I want at least four separate measures:

1. Cost per received contact: Local Services Ads spend divided by received contacts.
2. Cost per suitable opportunity: spend divided by contacts that meet the pre-agreed fit rules.
3. Cost per completed job: spend divided by completed work attributed to the cohort.
4. Contribution after acquisition: completed revenue minus direct delivery costs and Local Services Ads spend, using the business’s accounting definitions.

Do not compare gross revenue with advertising cost and call the remainder profit. Labour, materials, travel, subcontractors, refunds, finance costs and payment fees do not disappear because Google supplied the introduction. Nor would I use speculative lifetime value to rescue a weak first-order result. If repeat purchase or referrals materially change the economics, use observed behaviour from an appropriate cohort and keep the uncertainty visible. A heroic spreadsheet assumption is not a retention strategy.

There is no universal good Local Services Ads cost per lead. A contact price that works for a high-value emergency service may be absurd for a low-value domestic task. The ceiling comes from the business’s conversion stages and job economics.

Responsiveness is part of what you are buying

Paid demand exposes operational friction quickly because every missed opportunity arrives with a visible acquisition cost. Google says Local Services Ads ranking considers the likelihood that an ad will result in a lead. Its documented factors include responsiveness, relevance to the search, enabled contact routes and profile quality. Missed calls may negatively affect responsiveness; reviews, images and completed verification checks may also contribute to profile quality (Google’s Local Services Ads ranking guidance).

That does not mean every company should answer at all hours or enable every contact method. It means the advertised availability should match the operation that exists. If technicians cannot answer while working, decide who can. If messages are enabled, define a response standard the team can meet. If a service can be delivered only in a narrow window, do not advertise as though capacity is infinite.

Audit missed and delayed contacts separately from unsuitable requests. A handling failure points towards staffing, routing, contact settings or opening hours. Repeated poor fit points towards services, area, profile information, bidding or market demand. Blending them allows the channel and operation to blame one another indefinitely.

Trust signals cannot repair the wrong job

Local Services Ads place reviews and verification information close to the contact action. Those signals can help a potential customer assess a provider, and Google sets screening and conduct requirements for participation (Google’s Local Services Ads provider requirements). Keep the profile accurate. List services the operation genuinely provides, cover territory that makes economic sense and publish hours consistent with the response model. Ask for reviews through a normal, policy-compliant process.

But do not ask reputation signals to repair unit economics. More reviews do not shorten a two-hour drive. Verification does not make a low-value task commercially suitable. Faster response does not improve work the company should never have advertised. I treat profile strength as an amplifier. Before increasing it, ask what is being amplified. A well-presented wrong offer can acquire unsuitable contacts more efficiently.

Test one commercial cell before widening the map

Start with a controlled cell: one meaningful service group, a defensible area, stable operating hours and a budget the business can afford to learn with. The cell creates a decision boundary. If several services, postcodes, contact modes and budget changes are mixed together, more leads may arrive while the business loses the ability to explain why. Set the possible decisions before launch:

• Scale when suitable and completed-job economics clear the agreed threshold and the operation can absorb more work.
• Repair when demand fits but response, scheduling, profile accuracy or sales handling breaks the route.
• Narrow when particular services, areas or contact types repeatedly create poor-fit work.
• Hold when the cohort is too small or disrupted to support a decision.
• Stop when the controlled cell repeatedly fails the economic threshold after proportionate corrections.

Google provides bidding guidance and notes that automated systems need time and data to adjust (Google’s Local Services Ads bidding guidance). Platform learning time is not the same as commercial sample sufficiency. A business with sparse, high-value work may need longer to observe completed outcomes than the interface needs to adjust bids. Change one material variable where practical, record it and inspect the next bounded cohort. If you expand the area, services and budget together, the campaign may grow while the learning collapses.

Make the budget decision from work worth repeating

Before increasing spend, reconcile the charged contact list against handling, suitability, bookings, completed jobs and contribution. Name the service-and-area cells that earned more budget. Name the failure patterns that need repair. Leave uncertain cohorts uncertain. Then ask three questions: are we buying contacts for work we deliberately want? Can our operating model respond and deliver without destroying the margin? Which observed result gives us permission to spend the next pound?

If the report cannot answer those questions, I would not scale the dashboard. I would repair the classification and hand-off first. Google can sell contact with local demand. Your business still has to turn that contact into work worth completing, and prove that the same commercial cell deserves to run again.