ROAS is the number every advertiser watches, and the one most misread. A 2x return looks fine until you factor in margins, acquisition cost, and churn. In 2026, with CPMs rising across Meta and Google, getting more from existing spend matters more than scaling budgets. These five strategies cut through the noise and focus on what actually moves the number.
1. Fix the funnel before fixing the ads
Bad ROAS is rarely a creative problem alone. Before adjusting bids or testing new visuals, audit what happens after the click. A slow landing page, a weak value proposition above the fold, or a broken mobile checkout will kill conversions regardless of how good the ad is. Run a friction audit on the full path from impression to conversion. Map every drop-off point. If conversion rate is below 1.5% on a direct offer, the funnel needs work before the budget increases.
Example: An e-commerce brand running Meta Ads to a product page sees a 1.8x ROAS. After compressing images and rewriting the hero section to match the ad promise, ROAS moves to 2.9x with no change in spend. See also: this e-commerce launch guide for conversion fundamentals.
2. Segment audiences by intent, not by demographics
Targeting by age and interest is a starting point, not a strategy. In 2026, the gap between broad audiences and intent-based segmentation is where ROAS is won or lost. Use behavioral signals: people who visited a pricing page, abandoned a cart, watched 75% of a video ad, or engaged with a specific post. These micro-segments convert at a higher rate and justify higher CPMs. Build separate ad sets for each level of intent and allocate budget accordingly.
Example: A SaaS targeting startup founders runs one ad set for cold traffic and a separate retargeting set for users who visited the pricing page. The retargeting set delivers a 4x ROAS versus 1.6x for cold, despite a 40% higher CPM.
3. Use creative fatigue as a diagnostic signal
When ROAS drops over a two-week period with stable targeting and budget, creative fatigue is often the cause. Most advertisers react by creating more ads. A better approach is to treat fatigue as data. Study which creative element held attention longest before performance dropped: was it the hook, the format, or the offer? Use that information to iterate with precision rather than volume. Meta Ads frequency data combined with thumbstop rate gives a clear picture of where attention breaks.
Example: A direct-to-consumer brand notices ROAS declining on a winning video ad after three weeks. Frequency has reached 4.2. Rather than replacing the full creative, they recut the first three seconds with a new hook and restore ROAS to baseline within five days. Meta's own guidance on ad fatigue supports this diagnostic approach.
4. Match attribution window to product cycle
A 1-day click attribution window makes sense for impulse purchases. It makes no sense for a 199-euro software subscription or a high-consideration product with a two-week decision cycle. Mismatched attribution windows make profitable campaigns look unprofitable and vice versa. Set the window to match how your customer actually decides. For most B2B or premium products, a 7-day click window gives a more accurate picture. Then cross-reference with CRM data to validate what the ad platform is reporting.
Example: A productivity tool using 1-day click attribution sees a 1.1x ROAS and considers pausing campaigns. Switching to 7-day click reveals that 60% of conversions happen between day 2 and day 6 after the first click. Actual ROAS is 2.7x. No spend change, just a more accurate read.
5. Align ad offer with margin, not just revenue
ROAS measured on revenue can be misleading when margins vary by product. A 3x ROAS on a low-margin SKU can underperform a 1.8x ROAS on a high-margin one. The more useful metric is MER (marketing efficiency ratio) or contribution-margin ROAS, which factors in cost of goods, shipping, and returns. For subscription products, factor in average retention and LTV before deciding what ROAS target is actually profitable. This reframes budget decisions around profitability, not vanity metrics.
Example: When building the paid acquisition strategy for Sunna Planner, the target was never a raw ROAS figure. With a 3.99/month subscription and known churn patterns, the relevant question was: at what cost per install does the 12-month LTV stay positive? That number, not a platform-reported ROAS, drove budget decisions. For more on growth thinking in this context, see this growth strategy guide.
Takeaway
ROAS improvement is not a creative problem or a budget problem by default. Start with the funnel, sharpen the segmentation, read fatigue as signal, fix the attribution window, and measure against margin. Fix one variable at a time and the number becomes readable.