Health & Fitness · D2C · United States
How Forty Forty Cut Its Cost Per Purchase by 66% and Tripled Revenue in 90 Days
An untrained algorithm, wasted spend, and a $31 cost per purchase. Ninety days of disciplined weekly optimization later: 543 purchases in a single week at $10.87 each.
Forty Forty had a strong product and a broken paid acquisition engine
Forty Forty is a faith-based health and fitness program targeting adults, particularly women over 40, who want a structured and values-aligned approach to wellness. The product had clear demand, a defined audience, and strong mission alignment. The paid acquisition strategy did not.
When ads.expert came on board in November 2025, the Google Ads account was running without a trained algorithm, spending budget inefficiently across campaigns, and delivering a cost per purchase of $31.85, a figure that made profitable scale impossible. Branded and non-branded search campaigns were cannibalising each other's traffic. The Performance Max campaign had no clear purpose and was not suppressed after its urgency window had closed. Audience insights existed in the data but had never been acted on.
The account was generating purchases every week, but an underoptimized campaign structure meant a large share of potential buyers were never reached, never converted, and never attributed.
Systematically reduce cost per purchase to a level that makes scaling profitable: through bid strategy discipline, audience refinement, and campaign structure cleanup, and without sacrificing purchase volume.
Three structural problems keeping CAC elevated and growth capped
-
01
Algorithm spending without direction
Campaigns were running without a fully trained bidding strategy. Without sufficient conversion history and properly configured Target CPA or Target ROAS signals, Smart Bidding was allocating budget based on weak signals, inflating CAC and buying impressions that were unlikely to convert. At $31.85 per purchase, the account was structurally unprofitable at any meaningful scale.
-
02
Branded and non-branded campaigns cannibalising each other
Search term overlap between Branded and Non-Branded campaigns meant both were bidding against Forty Forty's own traffic, driving up CPCs on queries that should have been captured cheaply. Without regular search term filtering and negative keyword discipline, the account was paying a premium for clicks it should have owned at minimal cost.
-
03
High-value audience segment identified but never activated
The account data clearly showed that women aged 45 to 64 were the highest-converting demographic in Non-Branded Search. This insight had never been used to inform bid adjustments, ad copy, or audience targeting. The campaigns were serving generic messaging to a broad audience when a clearly defined high-value segment was sitting unused in the data.
Six levers, nine weeks, zero learning phase resets
The approach was disciplined and sequential. Every week had a clear performance review, a specific optimization action, and a decision framework for when to hold spend versus when to scale. The result was nine consecutive weeks of improvement with no panic adjustments.
Target ROAS Configuration by Campaign Type
Set differentiated Target ROAS targets across campaign types to give the algorithm clear, achievable goals without over-constraining volume. Non-Branded Search was set at 250% ROAS to prioritize efficiency; Christian Fitness campaigns were set at 90 to 100% to allow volume at a slightly looser threshold where audience overlap justified it. This prevented a single blended ROAS target from optimizing toward the wrong campaigns.
Weekly Negative Keyword Cadence
Implemented a weekly negative keyword addition cadence across all campaigns, systematically filtering irrelevant search terms before they could waste budget. Simultaneously ran regular search term audits to identify and block cross-campaign cannibalisation between Branded and Non-Branded Search, ensuring each campaign owned its intended query segment without competing against itself.
Budget Pacing During Learning Phase
Made a deliberate decision to cut spend during early optimization weeks, reducing budget by approximately 54% to allow the bidding algorithm to re-optimize without burning through budget on low-quality conversions. Budget was only increased again once CAC had stabilized below the target threshold, preventing the common mistake of scaling before the algorithm was ready.
PMAX Lifecycle Management
The PMAX campaign was built for a specific urgency window and paused once that window closed, rather than left running and diluting budget efficiency. Keeping a campaign live past its useful purpose increases algorithm confusion and cost per lead. Pausing it at the right moment preserved budget quality for the always-on campaigns heading into January.
Audience Intelligence and ICP Activation
Identified women aged 45 to 64 as the highest-converting demographic segment in Non-Branded Search and activated this insight across bid adjustments, ad copy, and audience targeting. This allowed the algorithm to prioritize the highest-value traffic segment while messaging was refined to speak directly to this audience.
Structured Weekly Performance Reviews
Established a consistent weekly performance review covering spend, CAC, purchase volume, search term quality, and bidding behavior, with a documented win, action, and next-step framework for each week. This prevented drift, ensured every change had a clear hypothesis, and created an accountability layer that meant no optimization opportunity was missed.
Google Ads
Google Analytics
Google Merchant Center
Shopify
Notion
Google Sheets
A 66% reduction in CAC across 9 weeks, achieved through bid strategy discipline, weekly negative keyword cadence, budget pacing, and demographic targeting, without a single learning phase reset.
See what else we have built
B2B SaaS · Search Infrastructure · France
How Meilisearch Stopped Buying Noise and Started Building an Enterprise Pipeline
EdTech · Higher Education · United States
How QuantHub Turned $20K in Ad Spend into $842K in Qualified Pipeline at a 38:1 Return
E-commerce Logistics · Turkey · UK · Germany
How OPLOG Turned a Fragmented Ad Presence into a $21.5M Revenue Engine Across Three Markets
Paying too much per purchase?
Let's bring that number down.
We help D2C brands build Google Ads systems that scale profitably, with weekly discipline, not one-time fixes.
Book a Strategy Call