382 inbound calls. 5 weeks. $7.02 each.

The Challenge

Two franchise locations. No paid search presence. Phones need to ring.

This client operates two auto repair franchise locations in separate markets. The business model is local and transaction-driven — someone's car needs service, they search, they call, they come in. The conversion window is short. Intent is high. The challenge: build campaigns that generate phone calls immediately, efficiently, and at a cost that makes the math work on a typical repair ticket.

Both accounts were built from scratch. No historical data. No conversion history. No previous campaign learnings to fall back on. Week one was the first data point.

The Structure

One campaign per service. No cannibalization. No wasted spend.

We built service-specific campaigns for each location rather than one broad "auto repair" campaign. The logic: different services have different CPCs, different seasonal patterns, and different average ticket sizes. Lumping them together makes optimization impossible. Separating them gives you a dial for each.

The data surfaced winners and problem areas immediately. Generic "mechanic near me" terms and inspection queries converted at under $2.50 per call at Location B — that's exceptional. Tires and alignments delivered high volume efficiently at both locations. A/C was generating calls but at 3–5× the account average, flagged for keyword refinement heading into summer.

Location vs. Location

Same playbook. Different markets. Different results.

Both locations ran the same campaign structure, the same bidding strategy (Maximize Conversions), and the same service-specific architecture. The cost-per-call gap between them — $9.25 vs $4.42 — is almost entirely explained by market-level competition and CPC differences, not campaign quality.

Location B's performance is genuinely exceptional — $4.42/call for local auto repair is a benchmark number. It's also accelerating: March spend at Location B was on pace to exceed February despite the month being only one-third complete at the time of this review.

The Testing Layer

Every campaign had a B-test running from week one.

Rather than running single ad variants and waiting, we launched structured A/B tests in every campaign — a base RSA, a messaging-variant RSA (safety, urgency, trust), and in several campaigns a promotion variant (50% off brakes, $25 synthetic oil change, $59.99 four-wheel alignment).

• Brakes "Safety Urgency" variant — converting at $29.48/call vs. $38.04 for the base RSA. The winner was clear within weeks.
• Generic "Brand & Neighborhood" variant — $14.58/call vs. the campaign average. Messaging that mentioned the local area and the brand name together was significantly outperforming generic service copy.
• Oil "Included Services" variant — $11.43/call, outperforming the base by 2×. Leads that knew what was included in the oil change were more likely to call.
• Promo variants — Mixed. "$59.99 alignment" and "$25 synthetic oil" showed potential. "50% off brakes" was expensive and underperforming — likely attracting overly price-sensitive callers.