How to Improve ROAS: The Metrics That Show What to Fix (2026)

Quick Answer: How do you improve ROAS?

To improve ROAS, check tracking and margin first, then fix the one metric that is broken. Make sure platform conversions and values match real orders. Work out your break-even ROAS from margin. Then compare conversion rate, order value and cost per click to find the weak lever, and fix it before you change bids or budgets.

ROAS (return on ad spend) is conversion value divided by ad cost, so 4x means $4 of tracked revenue for each $1 spent. Our ROAS calculator adds break-even ROAS and profit after ad spend.

Which symptom points to which fix?

Match the symptom to its metric, then try the first fix. Sources are platform documentation; row five is Ishant Sharma’s practice.

What is the right order to improve ROAS?

The right order to improve ROAS is tracking, margin, waste, feed, creative and landing pages, bidding, then order value. Each step feeds the next. Google’s Smart Bidding and Meta’s ad delivery both learn from the conversions and values you send. Change bids before you fix tracking, and the system learns the wrong lesson faster.

1. Tracking: conversions and values match real orders (tracking checks).
2. Margin: targets come from your break-even ROAS, not a benchmark.
3. Waste: stop paying for searches you would never sell to.
4. Feed: fix titles, identifiers and disapprovals, and group products by margin.
5. Creative and landing pages: lift conversion rate with message match and speed.
6. Bidding: bid on value once you have enough data.
7. Order value: raise AOV and count repeat purchases, so you can afford new customers.

Our Google Ads audit checklist covers the checks behind the first six.

Which metrics tell you what to fix?

Conversion rate, average order value and cost per click tell you what to fix, because ROAS is built from them. ROAS equals conversion rate times average order value divided by cost per click. If ROAS moves, one of those three moved, or the tracking behind them changed. Check each one before you touch bids.

The math: revenue is clicks × conversion rate × order value, and spend is clicks × CPC, so clicks cancel out. In this made-up illustration, three modest gains more than double ROAS.

How does conversion rate change ROAS?

At the same CPC and order value, ROAS moves in step with conversion rate. Google says that in retail "a 1-second delay in mobile can impact mobile conversions by up to 20 percent" (Google Ads Help). Keep the ad’s promise on the page, and show price, delivery and returns up front.

How does average order value change ROAS?

Bundles, free-shipping thresholds and add-ons lift order value, and ROAS with it. Discounts cut margin instead and raise break-even.

Illustration: a $100 product that costs $60 to make and ship has a 40% margin and breaks even at 2.5x. At $90, the margin falls to 33.3% ($30 ÷ $90) and break-even rises to 3.0x.

Is your click cost the problem?

Rising CPC lowers ROAS even when the page converts well. Google splits lost impression share into time lost "due to poor Ad Rank in the auction" and "due to insufficient budget" (Google Ads Help). Check auction insights for new rivals too.

Quality Score shows the weak part but does not set your price. Google calls it "a diagnostic tool" that "is not an input in the ad auction" (Google Ads Help). Use its three parts, expected CTR, ad relevance and landing page experience, as a to-do list.

Why is my ROAS low or dropping?

ROAS usually drops because tracking or values changed, a bid, budget or target changed, the feed or auction changed, or demand moved with the season. Google’s Shopping troubleshooting page names 12 causes, from "Recent changes to account or campaign settings" to "External market shifts" (Google Ads Help). Check the cheap causes first.

Did tracking, consent or conversion values change?

Compare platform conversions with store orders by day. A sudden gap usually points to a tag, consent banner or checkout change.

Google says "Transaction-specific conversion values are ideal if each sale holds a unique value for your business" (Google Ads Help). Our Google Ads conversion tracking guide covers consent mode and enhanced conversions.

Did a bid, budget or target change?

Google says the bidder reacts at once but needs time to hit a new target, so "give it 1-2 conversion cycles" (Google Ads Help). For budget-limited target campaigns, the August 17, 2026 update may also explain a shift.

Budget increases hide a trap. Illustration: spend rises from $10,000 to $15,000 and revenue from $40,000 to $52,000. Average ROAS falls from 4.0x to 3.47x, but the extra $5,000 brought $12,000, a marginal ROAS of 2.4x. At a 40% margin, break-even is 2.5x, so the new spend loses money.

Is it the feed, the auction, the season or tired creative?

For Shopping and Performance Max, check Merchant Center first, because disapprovals and feed edits change which products serve. Google also names "Auction dynamics" and "Changing user interest and seasonality" as causes (Google Ads Help), so check auction insights and compare with the same weeks last year. On Meta, if click-through rate falls as frequency climbs, test new creative before you change the audience.

How do you fix a ROAS below 1? A worked example

To fix a ROAS below 1, check tracking first, then cut the spend that sells nothing, then refresh creative, then fix the offer page. Below 1, each dollar of spend returns less than a dollar of revenue, so the account loses money before product costs.

Illustration (not a client account): a store spends $10,000 a month on Meta at 0.53 ROAS, so it reports $5,300 of revenue. At a 40% margin, break-even ROAS is 1 ÷ 0.40 = 2.5, or $25,000 of revenue. The $5,300 earns $2,120 of margin (5,300 × 0.40), a loss of $7,880 before fixed costs.

Getting from 0.53 to 2.5 needs 4.7 times the revenue per dollar (2.5 ÷ 0.53 = 4.72). No single fix does that. If all four together cannot get close to break-even, the problem is the product, price or channel, not the campaign.

What changed for ROAS bidding in 2026?

Google made four changes in 2026 that affect ROAS bidding, Meta retired its old shopping campaign API, and Microsoft is dropping maximum CPC caps on new automated-bid campaigns. The biggest is Google’s August 17, 2026 update: budget-limited Target ROAS and Target CPA campaigns now aim closer to the target. Google "won’t automatically adjust your bidding targets or budgets" (Google Ads Help), so check each target against recent results.

Our Google Ads bidding strategies guide covers the August change in depth.

How do you improve ROAS in Google Ads?

To improve ROAS in Google Ads, clean up search terms first, then fix the Shopping feed, then bid on value once the data is there. New accounts should build clean conversion data before they set a target, because Target ROAS has a conversion minimum.

How do search terms, negatives and match types affect ROAS?

Search terms show what you paid for, though Google omits terms that "don’t have enough query activity" (Google Ads Help). Ishant Sharma, who founded Hustle Marketers and has worked in digital marketing since 2013, starts new or low-volume accounts on exact match. He runs Search on exact for roughly the first 15 days to build conversion data, then lets Performance Max handle broad reach once that data exists.

When he runs broad match, he builds the negative keyword list upfront, in phrase and exact, at account level. Google confirms that account-level negatives "automatically apply to all Search and Shopping inventory, including in your Performance Max campaigns" (Google Ads Help). See our Performance Max negative keywords guide.

Split brand from non-brand too. In our data, brand Search had a median ROAS of 16.44x against 3.09x for non-brand (Google Ads benchmarks), as our branded vs non-branded search guide explains.

How do you improve Shopping and Performance Max ROAS?

Shopping and Performance Max ROAS starts in the feed. In one client account, "Dog Harness" became "No-Pull Dog Harness for Medium Breeds, Reflective, Adjustable Chest Strap" (pet accessories case study). Google says a GTIN "can increase the performance of the product" (Merchant Center Help).

Before he launches Performance Max, Ishant adds account-level negatives, excludes irrelevant placements, and excludes blog and other non-selling URLs. Then group products with custom labels by margin tier, keep the product feed current, and add brand exclusions if brand search runs separately.

Google’s August 7, 2025 update added campaign-level negative keyword lists to Performance Max (Google Ads Help). AI Max for Search adds "search term matching" and "final URL expansion" (Google Ads Help), so review the terms and pages it picks.

Which Shopping price badges can help ROAS?

Sale price annotations need a discount "greater than 5% and less than 90%". For US Shopping ads, the base price must also have run 5 days in the past 30 or 15 days in the past 200 (Merchant Center Help). Price drop annotations need a price "lower than the average price that you listed in the past 60 days" (Merchant Center Help).

Should you raise your target ROAS?

Raise target ROAS only when ROAS is stable and has room above break-even. Google warns that "setting a target that’s too high may limit the amount of traffic your ads may get" (Google Ads Help). Profit can fall even as reported ROAS rises.

Ishant’s sequence: accounts with enough conversion data run Maximize Conversion Value with no target until ROAS has been stable for about 30 days. Without a target, it "tries to fully spend your average daily budget" (Google Ads Help). Then he switches to target ROAS, raises it by 15 to 20%, and judges the result against break-even ROAS.

What do Google, Meta and Microsoft require before ROAS bidding works?

Google, Meta and Microsoft all bid on the conversion values you send, and each gives its own volume guidance. Google asks for 15 conversions in 30 days on Search and Shopping. Meta says its ROAS goal works best with 50 to 100 weekly conversions per ad set, and Microsoft suggests 30 conversions before you judge a target.

How do you improve ROAS on Meta Ads?

To improve ROAS on Meta Ads, fix the purchase signal first, then creative, then budget. Meta recommends the Conversions API "in addition to the Meta Pixel" (Meta) and says it helps "optimize ad targeting, decrease cost per result and measure outcomes" (Meta).

Why does Meta show fewer purchases than your store?

Meta counts the purchases it can tie to your ads, so its numbers rarely equal store orders. A large gap usually means missing signal. Send purchases from the pixel and the Conversions API, deduplicated with the same event_name plus an event_id, or an external_id with fbp (Meta).

Then watch event match quality, scored out of 10, and event coverage, "the 7-day average percentage of Meta Pixel events that are covered by the Conversions API" (Meta). Our Meta CAPI setup guide covers the build, and our tracking checks for Google Ads, Meta and GA4 compare what each counts.

How do creative testing and scaling affect Meta ROAS?

Test creative on a fixed routine. One of our clients, a pet accessories brand, ran a four-week cycle (case study). Week one launched 5 to 8 new creatives per ad set, and week two cut anything below 2x ROAS after 50+ link clicks.

Week three scaled winners with 20% daily budget increases, and week four refreshed angles. That cut-off fits that store’s margin, not every store.

Retargeting also flatters blended ROAS. In the same account, a dynamic product retargeting campaign returned 7.72x and broad prospecting 2.93x. Judge prospecting on its own numbers.

How should you set a Meta ROAS goal?

Set a Meta ROAS goal as a ratio, not a percent: Meta’s own example is a 110% return entered as 1.100 (About ROAS goal). Meta suggests starting 10 to 20% below your average ROAS, running at least 3 weeks before you scale, and changing bids only after 1 week (Best practices for ROAS goal).

Meta calls 1.00 "break even", but that only means purchase value equals spend. Your real floor is your break-even ROAS, which depends on margin. With no history, Meta suggests running Highest value for at least 2 weeks and starting from the ROAS it reaches.

How do you improve ROAS on Microsoft Ads?

To improve ROAS on Microsoft Ads, send real revenue with each conversion, then bid with Maximize Conversion Value and a ROAS target. Microsoft offers that for Search, Shopping and Performance Max. Since August 4, 2025, new campaigns set a ROAS target inside Maximize Conversion Value, because stand-alone Target ROAS "can’t be set up for newly created campaigns", while existing campaigns run as usual (Microsoft Advertising Help). Microsoft treats the target as a 30-day average. Conversion goals default to "NoValue" (Microsoft Learn), so set fixed or variable revenue before you bid on value.

For one of our clients, an epoxy flooring brand, Microsoft Advertising returned 1,084.68% ROAS (10.8x) from March 1 to August 15, 2026. That came from US$17,476.19 of spend and US$189,561.04 of revenue (case study). See our Bing Ads agency page.

For Amazon Ads, Walmart Connect and other retail media networks, where attribution windows and halo sales work differently, see our guide to improving ROAS in retail media.

How do you improve ROAS for lead generation and B2B?

For lead generation and B2B, ROAS only works when each lead carries a value and the platform learns what happened after the form. Google’s offline conversion import lets you "measure what happens in the offline world after your ad results in a click or call" (Google Ads Help). Bidding then learns from qualified leads and sales, not raw form fills.

Save the Google Click ID (GCLID) with every lead in your CRM. Upload outcomes as qualified leads and converted leads, the goal types that replaced "imported leads". Lead journey mapping links "conversion delays and values with each stage" (Google Ads Help). If your CRM uploads by API, those uploads moved to the Data Manager API on June 15, 2026 (Google Ads Help).

Value each stage by what it is worth. Illustration: if 25% of qualified leads close at an average deal value of $8,000, a qualified lead is worth 0.25 × $8,000 = $2,000. Avoid zero values, which Google calls "generally not recommended" with value-based bidding (Google Ads Help). Microsoft says to use its offline conversion goal "if you have lead generation as an objective" (Microsoft Learn), and Meta’s Conversions API accepts CRM events (Meta).

What is the quickest way to improve ROAS in B2B?

The quickest way to improve ROAS in B2B is usually to stop paying for searches that never become pipeline. Then send qualified-lead values back, so bidding stops chasing cheap form fills. Start by excluding job, free and training searches if you do not sell to them.

Why is my ROAS high but my profit low?

ROAS can be high while profit is low because ROAS counts revenue, not margin. Break-even ROAS is "1 ÷ Gross profit margin" (Amazon Ads), so a 4x ROAS only breaks even at a 25% margin. A store with 20% margins needs 5x just to cover product costs.

Contribution margin, after product, shipping and payment costs, gives the most honest figure. Our break-even ROAS guide shows the math, and contribution margin vs ROAS explains why 4x can lose money.

Brand search, retargeting, returning customers, view-through credit and late refunds can all make ROAS look better than the new profit behind it. Split new and returning buyers in reports. Google’s New Customer Value mode will "Bid higher for new customers than existing ones" (Google Ads Help).

How do you set ROAS targets across a large catalog?

Set ROAS targets by margin tier, not one store-wide number. A single target overspends on low-margin products and underspends on high-margin ones. Group products with custom labels and give each tier its own target.

How do AOV and lifetime value change the ROAS you can afford?

Repeat purchases let you accept a lower first-order ROAS. Illustration: a customer places three $60 orders a year at a 40% margin. The first order earns $24 of margin, so first-order break-even is 2.5x. Over the year they earn $72 (3 × $60 × 0.40), so you could spend up to $72 to win them, a first-order ROAS of 0.83x, and still break even.

Use real repeat-purchase data before you lower targets this way. For MER, POAS and incremental ROAS, see our ROAS vs ROI guide.

Is my ROAS good enough?

Your ROAS is good enough when it beats your break-even ROAS by enough to cover fixed costs and leave a profit. Break-even ROAS is 1 divided by margin, so a 40% margin breaks even at 2.5x (Amazon Ads).

Our own study covers 22 Google Ads accounts we managed from October 1, 2025 to September 30, 2026: 14 ecommerce and 8 lead generation. Together they had 46,901,558 impressions, 960,836 clicks and 57,228 conversions (Google Ads benchmarks by campaign type). ROAS is "conversion value divided by cost, as reported by Google Ads", and Zenodo hosts the dataset.

The break-even column is our calculation. As the study says, "This is one agency’s portfolio, not the whole market," and each campaign type has three to fourteen accounts. Our PPC statistics page has wider figures.

What ROAS gains have we seen in real accounts?

Our published case studies show ROAS gains from the same order of work: tracking and feed first, then structure by margin and intent, then bidding. Figures are copied from each case study and are not a promise, because Ishant never promises clients a ROAS figure.

Read the pet accessories split by channel. Search and Shopping meet existing demand while Meta creates it, so judge each against break-even, not against each other.

How long does it take to improve ROAS?

Improving ROAS takes several conversion cycles. Google asks for "1-2 conversion cycles" after each target change (Google Ads Help) and at least 2 weeks or 3 conversion cycles before you judge value-based bidding (Google Ads Help). A realistic plan has three stages: clean the data, bid on value, then tighten targets.

The plan uses only Ishant Sharma’s sequence and Google’s timings. Move on only when each step’s condition is met.

If month one finds broken tracking, the clock restarts from the fix. That is why we never quote a ROAS figure upfront.

If you want this done for you, our PPC management costs $600 to $800 per platform, month to month with no lock-in, and clients own their ad accounts. See ecommerce PPC management or Google Ads agency services.

ROAS Improvement FAQs

Is 2.5 a good ROAS?

A 2.5 ROAS is break-even at a 40% margin, because break-even ROAS is 1 divided by margin (1 ÷ 0.40 = 2.5), as Amazon Ads shows. With a higher margin, 2.5 makes a profit before fixed costs. With a lower margin, it loses money on every sale.

Is a 4 ROAS good?

A 4 ROAS is good if your margin is above 25%, since 1 ÷ 0.25 = 4 is the break-even point. At a 40% margin, 4x leaves $0.60 of margin per $1 of spend after ad costs (4 × 0.40 − 1). At a 20% margin, the same 4x loses $0.20 per $1.

Is 5x ROAS good?

5x ROAS breaks even at a 20% margin and makes a profit above it. In our data from 22 accounts, the ecommerce median was 5.62x for Standard Shopping and 4.95x for Performance Max (Hustle Marketers benchmarks), so 5x sits near the middle. Judge it against your break-even, not the median.

Why does last week’s ROAS look low in Google Ads?

Recent days look worse because Google Ads has always "reported conversions by the date your ad was clicked" (Google Ads Help). Sales from buyers who convert days later land back on last week. Add the Conversions (by conv. time) columns, and judge recent weeks after your usual conversion lag.

Why did my Google Ads ROAS jump overnight?

A sudden jump with no extra orders usually means the conversion value changed, not performance. Check for duplicate tags, a second purchase action set as primary, or tax and shipping added to the value. One merchant found the value sent to Google Ads had "suddenly started including VAT" (Shopify Community, October 2025).

Should I increase my budget when ROAS is low?

Not until you know why ROAS is low. Extra budget buys auctions you skipped before, so new spend often returns less than your average. Fix tracking, waste and conversion rate first, then add budget only where Google reports impression share lost "due to insufficient budget" (Google Ads Help).

Can I improve my ROAS without increasing my ad budget?

Yes. Cutting wasted search terms, fixing conversion values, raising conversion rate and lifting order value all raise ROAS at the same spend. A 50% lift in conversion rate lifts ROAS by 50% if order value and CPC hold.

Why don’t my Google Shopping ads show star ratings?

Google requires "a minimum of 50 reviews across all of your products" and matches reviews to products mainly through GTINs (Merchant Center Help). Too few reviews or missing GTINs are the usual reasons stars do not appear.

How we created this guide

We built this guide from Google, Meta and Microsoft documentation, our benchmark study and our case studies. We checked every rule, date and figure on November 3, 2026, re-checked the Meta and Microsoft help pages on November 4, 2026, and each links to its source.

Ishant Sharma’s methods are his stated practice, not platform rules. Worked examples are illustrations with the math shown. Hustle Marketers publishes this guide and sells PPC management.

Last updated: November 4, 2026.