Case Studies
A few engagements start to finish: what was broken, what I changed, and what it did to the numbers. Client names are left out. The numbers are real.
Case Study · Consumer Goods E-Commerce
900% ROAS Was a Lie
Google Ads account restructure for a consumer goods brand on Shopify.
The Problem
The Google Ads account was reporting 900% ROAS. Checked against Shopify purchase data, the real number was 1.04x: about $42K a quarter in spend bringing in $43,649 in revenue, which is break-even before product costs. Conversion tracking was counting micro-events instead of purchases, there was no UTM structure, and nobody could see Shopify revenue next to ad spend.
What I Did
- Fixed the data first. Rebuilt conversion tracking to count purchases only, put a full UTM structure on every campaign, and made Shopify the single source of truth for revenue.
- Cut spend on purpose. Pulled quarterly spend from $42K to $26K to collect clean data before optimizing anything.
- Optimized against real purchases. Rebuilt targeting and keywords around actual orders and paused underperformers as they surfaced. Q3 hit 2.91x ROAS on $19,850 in spend.
- Stabilized and planned the scale-up. Three campaigns drove 86% of revenue at a $51 to $72 average order. Three more were flagged as likely unprofitable (under $38). The account settled at $10K a month with a clear scaling roadmap.
The Results
2.29x
Real ROAS in Q4, up from 1.04x in Q1
+42%
Revenue growth
-36%
Quarterly ad spend ($42K to $27K)
+12%
Average order value
+13%
Orders (948 to 1,071)
$27K
Quarterly swing from loss to profit
You cannot optimize your way out of a measurement problem.
Case Study · Consumer Goods E-Commerce
One Move That Changed Brand Performance
Branded search isolation for a Shopify DTC brand, turning a blind spot into the highest-ROAS asset in the account.
The Problem
The account ran two campaigns, Performance Max and General Search, with brand and non-brand keywords sharing budget. There was no way to see what brand terms were driving or whether they cannibalized organic traffic, and Google pushed spend toward high-volume non-brand queries, starving the brand terms that actually converted.
What I Did
- Gave brand its own campaign. Pure brand terms plus consideration-phase searches like “is it food safe” and “vs mineral oil” moved into one owned Search campaign.
- Capped the budget. Brand held to 25 to 30% of total spend, enough to protect every high-intent branded query without crowding out non-brand or cannibalizing organic.
- Enforced negatives account-wide. Brand terms were excluded from every non-brand campaign, so nothing bled across campaigns.
- Targeted the research phase. Reaching people who were still validating before they buy, along with repeat buyers, is what produced the 71% new-customer rate.
The Results
4.04x
Brand campaign ROAS, verified in Shopify
23.89%
CTR on branded terms (benchmark 5 to 10%)
71%
New customers from the brand campaign
$0.90
Average CPC on brand terms
1.74x
Efficiency: 28% of budget drove 49% of revenue
+20%
AOV vs. account average ($62 vs. $52)
When brand terms share a campaign with non-brand, budget goes wherever Google decides. Isolate them and you control the lever.
Case Study · Specialty Accessories E-Commerce
From Frozen Account to 69% Open Rates
Email platform rescue and rebuild for a specialty accessories brand on Shopify, six weeks before the holidays.
The Problem
The brand migrated email platforms without a warm-up plan and lost its entire sending capability six weeks before the holidays. Before the move, open rates had swung from 0.3% to over 40% with no pattern, and only a fraction of the 50,000+ contacts on file were actually sendable.
What I Did
- Diagnosed the platform. Swings that extreme pointed to the sending infrastructure itself as the root cause, ahead of creative, frequency, or list quality.
- Segmented in an emergency. Suppressed unengaged contacts and set up a phased re-engagement to protect sender reputation before the holiday window.
- Ran the warm-up like a campaign. A six-week domain warm-up, sequenced by engagement tier, so the full list was ready for the holiday send.
- Moved to Klaviyo. An e-commerce platform with deep Shopify integration, which opened up behavioral triggers and real lifecycle automation.
- Built the lifecycle flows. Welcome series, abandoned checkout, and browse abandonment flows, built from scratch to run revenue on autopilot.
The Results
54-69%
Consistent campaign open rates (Klaviyo, bot-filtered)
0.3%
Low end of open rates on the previous platform
~5x
Active list growth after full re-engagement
6 wks
From frozen account to full holiday send
1.41%
Browse abandonment flow conversion rate
1.05%
Abandoned checkout flow conversion rate
Consistent 54% beats an occasional 70% every time, because you can build on it.
Case Study · Specialty Bike Retail & E-Commerce
Turning Paid Traffic Into Inventory Movement
Lifecycle and paid media for a specialty bike repair and parts retailer, without more ad spend.
The Problem
Paid media was driving traffic that stalled at the cart. The retailer runs a hybrid model, high-touch local service plus national e-commerce for specialty components, and had no retention layer to bring hesitant buyers back.
What I Did
- Gave each channel a job. Meta became awareness and retargeting fuel instead of a direct revenue channel. Google was optimized for high-AOV component sales.
- Found the leak in GA4. Funnel exploration reports pinpointed cart abandonment as the primary leak, so the focus moved from buying traffic to mid-funnel throughput.
- Delayed the cart emails on purpose. A three-day delay matched how specialty parts buyers decide. They need time to check compatibility, and patience outperformed urgency.
- Recovered revenue without discounts. The first reminder email, with no discount, drove 68% of recovered revenue. The 10%-off email drove zero conversions.
The Results
42%
Reduction in cart abandonment rate
34.9%
Abandoned cart email CTR
3x
The industry email CTR benchmark (10 to 14%)
$851
Recovered at 100% margin
$216
Average order value from Google Ads
57,768
New users driven by Meta campaigns
The discount email converted zero users. Once we removed the friction, the revenue was already there, waiting.
Case Study · Public Health Advocacy
Beating the Benchmark on an HIV Awareness Campaign
Multi-channel awareness campaign for a statewide public health program, managed at KeyMedia Solutions.
The Challenge
The campaign needed to raise HIV awareness, drive testing, and connect people to care across two audiences: the general public and a priority at-risk audience. It ran on programmatic audio, display, and video, plus YouTube, Snapchat, and a niche LGBTQ+ ad network.
What We Did
- Tested YouTube objectives head to head. “Video views” and “Brand awareness and reach” had the same 0.08% CTR, but video views cost $7.85 per conversion against $22.85. The budget moved.
- Started Snapchat broad, then narrowed. Clicks skewed 32% toward ages 35+, so targeting moved to 35+, Apple and Samsung devices, and separate ad groups for the top interests.
- Carried insights across platforms. Sports fans were engaging on Snapchat and X, so a Sports Fans ad group launched on YouTube.
- Shifted the LGBTQ+ network to web. Web placements were pulling nearly 2x the CTR of app placements, so web got priority.
- Looked past the ads. A landing page audit found copy you could read in under seven seconds, unchanged since 2020, so we recommended richer content. We also caught 8,000+ video-driven sessions showing 0% engagement in GA, a tracking issue, before it wasted more spend.
The Results
68.27%
YouTube completion rate, against a 29.6% benchmark
99.09%
Programmatic audio completion rate
2.25%
Snapchat CTR, against a 0.3 to 1.5% benchmark
+123%
CTR after shifting to web placements
+175%
Programmatic completion rate in one month
$7.85
Cost per conversion, down from $22.85
Launch broad, let the first few weeks collect the data, then follow where it points.
Case Study · B2B Telecom
Every Tactic Had a Job. Not Every KPI Knew It.
Search and programmatic for a regional B2B broadband and cybersecurity provider, run agency-side.
The Problem
The account ran Google Ads search alongside programmatic display and CTV. An awareness tactic was being graded on CPA, search was judged on clicks and calls with no tracking on form fills, and search terms leaned residential for a provider that only sells to businesses.
What I Did
- Re-scored the awareness tactics. CTV carried a CPA target on an awareness campaign. I moved it to vCPM, cost per completed view, and completion rate, and let display retargeting chase the clicks.
- Segmented programmatic by industry. Broad, lookalike, and retargeting groups ran for a few months, then got cut into six core segments: Education, Financial, Government, Health Care, IT Pros, and Manufacturing.
- Pointed search at business buyers. Tested broad vs. exact match, moved core terms to exact, and layered bid adjustments by city, device, and day of week.
- Fixed what the reports were missing. Added form fill and engagement tracking mid-year, and traced customer login pages that were inflating site traffic back to the tag setup.
- Cut low-value inventory. Added an invalid traffic filter and bid down or removed weak sites, while keeping budget pacing in the 90s.
The Results
+145%
Active users from paid media, 2023 vs. 2022
+133%
Sessions from paid media, 2023 vs. 2022
+75%
User conversion rate from paid traffic, year over year
376
Calls from search ads in 2023
278
Contact form submissions from search ads
$42.60
Display CPA in September, against a $175 goal
If a campaign is built for awareness, grading it on CPA just tells you it’s failing at a job it never had.
Get all five in one file
The full set of case study one-pagers as a single PDF, ready to save or pass along.
Want results like these?
Tell me what’s going on with your marketing and I’ll tell you where I’d start.