The numbers
don't lie. Real PPC case study results.

A nationwide B2B security staffing company with 437 offices across all 50 states was generating leads at $174 each — and the true cost was even higher. Google Ads was double-counting leads, making performance look twice as good as reality. GCLID capture was at only 15%, leaving the bidding algorithm flying blind, and offline leads weren't tracked back into the platform at all. Compounding the measurement failure: the account was serving event-related searches (concerts, weddings) producing one-night gigs, not contracts — with all paid traffic landing on the homepage.

Fixed measurement first. Improved GCLID capture from 15% to 90%, set up enhanced conversions for all lead types (phone, form, chat, email), and imported offline qualified leads from the CRM as the primary conversion signal — giving the algorithm what it needed to optimize. Then overhauled the account: removed event-intent keywords, eliminated 23 irrelevant Fire Watch search terms (fire inspection, sprinkler, permits), rewrote all RSAs with credential-led copy (437 offices, same-day deployment, BBB A+), built a dedicated landing page with live chat integration, ran a structured A/B experiment vs. the homepage, and shifted bidding to Target CPA with accurate data as its input.

Within 60 days, leads per week went from 54 to 82, CPL dropped from $174 to ~$100, and conversion rate improved 40%. Over 16 months and 5,600+ backend-verified leads, the blended CPL held at $99.83. Better data produced better decisions, which produced better outcomes — at the original rate, those leads would have cost $419,000 more.

This women's fashion brand had been through multiple agencies on Google Ads. The non-brand campaigns — the ones responsible for new customer acquisition — were burning budget without measurable purchase results. The primary non-brand search campaigns weren't optimized for purchases at all: they spent tens of thousands and returned zero purchase revenue. PMax campaigns ran without ROAS guardrails. Combined, all prior non-brand management produced $76k in attributed purchase revenue across their entire engagement — a number we now exceed in a single month. The brand had real demand through Meta and a highly engaged live-shopping community, but Google wasn't converting any of that awareness into measurable new customer volume.

Rebuilt the non-brand prospecting architecture from scratch with purchases as the only conversion objective. Built four pillars: Shopping for product-level search capture, PMax segmented by category to give the algorithm tighter signals, and Search targeting high-intent category queries. Added negative keyword fencing at every level to prevent brand traffic from contaminating non-brand budgets. Scaled non-brand spend methodically as each campaign proved efficiency — reaching $99k+/month in prospecting spend while the overall account grew to $120k+/month.

Non-brand campaigns now drive 2,839 purchases and $294k in revenue every month. Total account revenue scaled to $525k/month at 4.51× ROAS. Google went from a channel that couldn't justify its own non-brand budget to the brand's primary new customer acquisition engine outside of social — exactly what they needed to reduce their dependency on Meta.

This DTC brand had cycled through multiple agencies since 2023. By December 2025 their Google ROAS had dropped to 2.7× — the previous agency had scaled Demand Gen spend without efficiency guardrails, chasing volume and torching margin. Multiple PMax campaigns were competing internally, AI Max was pulling search into uncontrolled territory, and remarketing campaigns were effectively running as prospecting. No documentation of what had been tried or why things changed.

Diagnosed before prescribing. Consolidated competing PMax campaigns, disabled AI Max (re-activated later with proper negative keyword structure as a controlled test), added ROAS guardrails to every Demand Gen campaign, and fixed remarketing targeting. Then scaled methodically: mirrored Meta's top-performing creative formats on Google, concentrated spend in core markets (CA, TX, NY), and ran a 60-day search term audit across 4,000+ queries — surfacing 10 new high-ROAS keywords including terms converting at 541–1,754% ROAS.

ROAS recovered from 2.7× to 3.51× while spend increased 37%. The account grew, not just stabilized. Demand Gen alone drove 1,200+ attributed branded searches in Q1 — evidence upper-funnel investment was creating downstream search demand. Google Ads AOV increased +19.3% year-over-year, signaling higher-value customers finding the account.

A SaaS company brand new to paid search — zero historical data, no conversion tracking, no existing strategy. The goal wasn't just growth: it was driving enough qualified signups to hit a $3M ARR milestone that would unlock a Series A raise. The math had to work before the pitch did.

Built full-funnel conversion tracking from scratch. Conducted deep keyword research across use cases and job roles to understand how different buyer personas search. Prioritized exact-match targeting with tightly themed ad groups — no broad match bleed, no wasted impressions. Wrote tailored ad copy for each segment and collaborated on high-converting landing pages matched to each keyword theme.

518% increase in signups. 75% drop in CPA. 441% lift in conversion rate. All within the first 90 days. The data also surfaced the highest-performing search themes — giving the product and sales teams intelligence on which use cases were driving the most motivated buyers.

A home décor brand with strong product-market fit but no proven Pinterest strategy. The platform was generating interest but not reliable revenue — in-platform data was unreliable, creative testing was scattered, and the team had no clear answer to whether Pinterest could scale without sacrificing ROAS. The question wasn't whether to use Pinterest. It was whether it could be made to work predictably.

Rebuilt the Pinterest strategy from the ground up: implemented Conversion API with enhanced match for reliable attribution, consolidated campaigns by objective and product category, switched to broad open targeting with automatic bidding to let Pinterest's algorithm find buyers. Tested ad formats systematically — static lifestyle images outperformed video, a finding that reshuffled creative priorities. Launched catalog and shopping ads for both prospecting and retargeting, and built automated performance reporting.

Over $6M in attributed revenue at an 8.84× blended ROAS and a $4,287 average order value. Pinterest became a reliable customer acquisition channel — not an experiment — with repeatable results and a creative playbook that eliminated the guesswork from future testing.

A home décor brand that had already found success on Pinterest but hit a performance plateau. Revenue had stalled despite being in one of Pinterest's top-performing product categories. The account had a large SKU catalog and abundant creative assets — but no system for knowing which products or formats to scale. The goal: double annual revenue without sacrificing ROAS.

Restructured the account into a leaner, more focused campaign setup. Audited the full product catalog and concentrated spend on historically top-performing SKUs — cutting the long tail that was diluting budget. Identified that UGC-style short-form video (originally shot for TikTok) was the highest-performing creative format and scaled spend behind it. Launched dynamic prospecting campaigns to expand the buyer pool beyond existing audiences.

Monthly revenue doubled: from $46K to $102K — a 122% increase year-over-year. Monthly purchase volume grew from 319 to 740 units (+132%). ROAS held strong through the scale — this wasn't revenue at any cost. The creative insight (TikTok-native UGC outperforming produced video) became a repeatable playbook for future creative decisions.