OCP (Oregon Catholic Press)
OCP needed a paid advertising partner to audit 4 brands, fix broken tracking, and build campaigns that drove real purchases and qualified license requests.
The Challenge.
The dashboards showed thousands of conversions per month, and almost none of those conversions were actual purchases. Tracking was broken, duplicated, or measuring the wrong things entirely.
One campaign was targeting every country on Earth. The algorithm had been bidding toward conversions that were not purchases.
Our Role.
We restructured campaigns, fixed location targeting, switched bid strategies, and cut every keyword that had spent money with nothing to show for it.
Then we launched and managed ongoing campaigns across both platforms with weekly reporting and adjustments.
The Results.
Across both platforms that was hundreds of purchases per week. In February that meant $35,815 in tracked revenue on $2,818 of spend: 1,080 purchases, a 17.58% conversion rate, 1271% ROAS on Google, plus 302 Meta purchases on $479 of spend at a $2.32 CPM.
By the March report it was $28,839 in revenue on $2,620 of spend, 1,222 purchases, a 20% conversion rate, 1100% ROAS on Google, and 646 Meta purchases on $915 of spend at a $1.97 CPM. Those are ecommerce numbers.
Music Publishing runs on license requests, so on the publishing side we count real, qualified license requests, and on the store we count purchases. OCP Worship and Music Publishing joined the account in April, the second of 2 engagements that ran across the first half of 2026. OCP came back in September 2026 for a third project.
The Audit
We started where we always start: the data.
OCP had 4 brands running paid ads across 5 websites: OCP.org, OCPChoral.com, OCPWorship.com, OCPMusicPub.com, and EchoHarborMusic.com. They hold 70% saturation in Catholic churches with their Breaking Bread hymnal. Their audience is music directors, choir directors, and anyone involved with liturgy planning. SEO was already their key traffic driver. Paid ads were meant to supplement that.
The dashboards reported that everything was working, and the conversion counts did not match actual purchases.
Google Ads: 10,001 Conversions That Weren’t
The main ecommerce account was reporting over 10,000 conversions per month. Add-to-cart and begin-checkout were set as primary conversion actions, so Google was bidding toward people who browsed. Those conversions did not match revenue.
We fixed the tracking, switched bidding from Maximize Conversions to Maximize Conversion Value, and cut every keyword that had spent $66 and $133 with zero actual sales.
1,400 Conversions. Real Number: 70.
OCP Music Publishing had a conversion set up to track page visits to their licensing page. That same conversion was duplicated, so it counted twice. The average visitor viewed that page about 10 times per session.
The duplicate doubled the count and page reloads multiplied it by 10. 1,400 divided by 20 equals 70.
Google had been bidding toward 1,400 conversions when the real number was 70, and it spent real money doing that for months.
We rebuilt the setup with one form submission trigger that fires only on actual submissions and counts one conversion per person.
After the Rebuild
After the rebuild, clicks on the Music Publishing side cost $0.15 to $0.32 on average across May and June. The data also showed something we didn’t love. In June, 132 ad-driven visitors reached the licensing page and zero of them filled out the form. The recommendation was to add the same form to that URL, since something between the click and the form was losing every visitor.
We also assigned real dollar values to each conversion action that same month, so Google could bid toward the actions worth more to the business. A license request outranks a video play. By the final report, Google was already prioritizing the higher-value conversions.
Targeting Every Country on Earth
One campaign was targeting every country on the planet, and the client sells music licenses in the United States.
Their other campaign was narrowed to the US and Canada, with a setting that showed ads to anyone Google considers “interested in” those countries. Someone in New Zealand researching a trip to Cleveland could get served an ad for church music licensing. Both location settings were wrong.
We narrowed the targeting to the US only and changed the setting from “interested in” to “presence in.” The fix took 5 minutes, and the old settings had been spending budget for months.
The Lent Campaign
Most ad platforms don’t know what a missal is, let alone when Lent starts. We built a campaign that runs on the liturgical calendar, live for roughly 40 to 50 days around Lent and Easter, then shut off.
On Google, the top keyword in that window spent $375 and returned $4 for every $1, a 400% return. On Meta, the Missals ad set got paused 2 weeks before Easter, on purpose, because a missal that arrives after Easter helps nobody.
Our recommendation for next year is to keep the campaign and repurpose it, since it holds a full season of real data.
OCP Worship: Starting From Zero
OCP Worship had no Google Tag Manager container, no Google Ads tag, and no conversion tracking on the site. The single Search campaign targeted all countries, used only broad match keywords, and had no negative keywords at the campaign, list, or account level. It spent $40/day on ads shown to searchers in any country.
Meta Ads
Meta was in better shape than Google. The Conversions API and Pixel were set up correctly. Campaign structure and settings were mostly in line with best practice. Audience lists were segmented logically.
The open items were about growth. 35% of spend in the last 7 days was allocated to ad sets still in the learning phase. Over 30% means the budget is too low to push campaigns out of learning fast enough. There was no systematic approach to testing hooks, messaging angles, or creative themes. Underperforming ads were not being turned off quickly enough.
We Teach The Account, Not Just Run It
Our specialist recorded Loom walkthroughs OCP’s team gets to keep, covering how the account is structured, how to build a campaign from scratch, when a new campaign is the right call versus a new ad group, how to prune a keyword list, and how to set geo-targeting correctly.
Training calls and a checklist PDF went with it. OCP’s team ended the engagement running more of the account themselves than when it started.
September 2026: OCP Choral gets its own Google Ads account
OCP asked us to split the OCP Choral store’s advertising out of the main OCP Google Ads account. Our specialist set up the new account and moved the store’s Shopify Google & YouTube app to it, so the tag and purchase conversions reach the right account. In Merchant Center, we removed 3 old or empty feeds. We recommended keeping the automated Shopify feed of around 5,000 products, which has fewer product disapprovals, and deleting the older file-based feed of 10,000+ products, which keeps Google Ads from combining them. We also set up a feed rule that appends search terms to product titles, so the titles carry phrases people search, such as “Catholic choir music” and “choral sheet music.”
OCP’s budget was flat, and the Choral campaigns spent $8,462.02 in 2025, about $23 a day. We recommended 2 campaigns: Brand at about $3 to $4 a day and Performance Max starting at $12 to $15 a day. The Non-brand campaign is built and stays off until OCP has more budget.
Our specialist excluded “Religion” as a content theme so the ads would not appear beside unrelated religious content, and asked OCP to review it given what OCP sells. The campaigns came with a Loom walkthrough, and OCP’s team is finishing 2 steps: deleting the old feed and completing Google’s advertiser verification.
View More Projects From Our Team
Our name has a story.
Briefcase Marketing wasn’t founded
out of our love for briefcases.
