01 Overview
In 2025, the debate over which advertising platform delivers the best results continues to evolve. Search, social, and retail media each offer distinct advantages, and the winner depends on your goals, audience, and measurement capabilities. Search remains a powerhouse for capturing existing demand, social excels at generating interest and building brand, and retail media bridges the gap by reaching shoppers close to the point of purchase. Understanding the strengths and limitations of each is crucial for allocating budgets effectively.
The choice is not about a single victor but about how these platforms complement each other within a holistic strategy. Search ads often close sales from users who are already searching for solutions, social ads introduce products to new audiences and nurture consideration, and retail media leverages retailer first-party data to target shoppers with high purchase intent. A balanced mix can drive both short-term conversions and long-term growth. However, without proper measurement, it's easy to misattribute success and overspend on channels that appear to perform well on the surface.
As privacy regulations tighten and third-party cookies fade, the ability to measure true incrementality becomes more important than ever. Each platform provides its own metrics, but these can be misleading if taken at face value. The key is to adopt a test-and-learn approach, using controlled experiments to understand the unique contribution of each channel. In this article, we'll explore how to evaluate search, social, and retail media using a common metric—return on ad spend (ROAS)—and provide guidance on interpreting and scaling performance.
02 What this metric really measures
Return on ad spend (ROAS) is a ratio of revenue generated to the cost of advertising. It is calculated by dividing the revenue attributed to ads by the ad spend. For example, a ROAS of 5:1 means that for every dollar spent, five dollars in revenue were generated. While this seems straightforward, ROAS measures correlation, not causation. It tells you how much revenue was associated with users who saw or clicked on your ads, but it does not prove that the ads caused those conversions. Many of those users might have converted anyway through other channels or organic means.
Moreover, ROAS does not account for profit margins, which vary significantly across products and industries. A high ROAS on a low-margin product may be less profitable than a lower ROAS on a high-margin item. Additionally, ROAS is often calculated using platform-reported conversion data, which can be inflated due to view-through conversions, click attribution windows, and modeled conversions. Each platform has its own definition and attribution model, making cross-platform comparisons challenging. For instance, a social platform might count a conversion if a user viewed an ad and later purchased within a 7-day window, while a search platform might only count clicks within a 30-day window.
Therefore, ROAS should be viewed as a starting point for optimization, not a definitive measure of success. It is most useful when compared against your own historical data and when combined with other metrics like customer acquisition cost (CAC), lifetime value (LTV), and incremental lift. To truly understand the impact of your ads, you need to measure incrementality—the additional revenue that would not have occurred without the ads. This requires rigorous experimentation, such as geo-holdout tests or randomized control trials, which are becoming more accessible with advanced measurement tools.
03 How it moves with your levers
ROAS is influenced by a variety of factors that you can control. Audience targeting is perhaps the most significant lever: reaching the right people with the right message at the right time increases the likelihood of conversion. On search, this means bidding on relevant keywords and using negative keywords to filter out irrelevant traffic. On social, it involves leveraging interest, demographic, and behavioral targeting to find users who are likely to engage. On retail media, it's about using the retailer's first-party data to target shoppers based on purchase history and browsing behavior. The more precise your targeting, the higher your ROAS tends to be, though overly narrow targeting can limit scale.
Creative quality also plays a critical role. Compelling ad copy, eye-catching visuals, and clear calls to action can dramatically improve click-through and conversion rates. On social platforms, where users are scrolling quickly, creative must stop the scroll and resonate emotionally. On search, ad copy should match the user's query and highlight unique selling propositions. On retail media, product images and ratings are key. Testing different creatives and iterating based on performance is essential to maintain high ROAS.
Bid strategy and budget allocation are other levers. Automated bidding algorithms can optimize for conversions or revenue, but they require accurate conversion data to perform well. Manual bidding gives more control but demands constant monitoring. Landing page experience is often overlooked; a slow, confusing, or irrelevant landing page can kill conversions even with great ads. Ensuring that your landing page aligns with the ad promise and provides a seamless path to purchase is vital. Finally, pricing and promotions can temporarily boost ROAS, but they may erode margins. Balancing these levers is an ongoing process of testing and refinement.
04 Reading the number without fooling yourself
One of the biggest mistakes advertisers make is taking ROAS at face value without considering its limitations. A high ROAS might look great, but if it's driven by retargeting users who were already going to purchase, it may not represent incremental growth. Conversely, a low ROAS on prospecting campaigns might be acceptable if it brings in new customers who will have high lifetime value. Therefore, it's essential to segment ROAS by campaign type, audience, and stage in the customer journey. For example, separate ROAS for brand vs. non-brand search, or for prospecting vs. retargeting on social.
Another pitfall is comparing ROAS across platforms without adjusting for attribution differences. Each platform measures conversions differently, and some may claim credit for conversions that were influenced by other channels. This can lead to overinvestment in the platform with the most generous attribution model. To avoid this, use a unified measurement approach, such as marketing mix modeling (MMM) or incrementality testing, to understand the true contribution of each channel. Additionally, be wary of ROAS calculated on a last-click basis, which ignores upper-funnel influence. Multi-touch attribution can provide a more balanced view, though it also has limitations.
Finally, remember that ROAS is a ratio, not an absolute profit measure. A ROAS of 4:1 might be profitable for one business but not for another, depending on margins and operating costs. Always calculate your break-even ROAS based on your profit margin. For instance, if your profit margin is 25%, you need a ROAS of at least 4:1 to break even. This simple calculation can prevent you from scaling unprofitable campaigns. Regularly reviewing ROAS alongside other metrics like CAC, LTV, and payback period will give you a more complete picture of your advertising effectiveness.
05 Benchmarks to sanity-check against
While exact benchmarks vary by industry, business model, and campaign objectives, rough ranges can help you gauge whether your ROAS is in the ballpark. For search advertising, a ROAS between 3:1 and 6:1 is common, though highly competitive industries may see lower, and branded search can be much higher. Social advertising often yields a ROAS between 2:1 and 4:1, with retargeting campaigns typically performing at the higher end. Retail media, which benefits from high purchase intent, often sees ROAS between 2:1 and 5:1, depending on the retailer and product category. These ranges are not guarantees but rather general observations from aggregated data.
It's important to note that these benchmarks are not one-size-fits-all. A luxury brand might have a lower ROAS due to higher price points and longer consideration cycles, while a fast-moving consumer goods brand might see higher ROAS due to frequent purchases. Similarly, B2B companies often have longer sales cycles and may measure ROAS over a longer window. Therefore, use these ranges as a starting point for comparison, but always benchmark against your own historical performance and industry peers. If your ROAS is significantly below the range, it may indicate issues with targeting, creative, or landing pages. If it's significantly above, you might be underinvesting or missing opportunities to scale.
When evaluating benchmarks, consider the attribution window used. A 7-day click window will yield different results than a 1-day view window. Also, be cautious of platform-provided benchmarks, as they may be skewed to highlight the platform's strengths. Third-party studies or industry reports can offer more objective insights, but they may not reflect your specific situation. The best benchmark is your own data over time, analyzed with consistent methodology. By tracking ROAS trends and correlating them with changes in strategy, you can identify what works for your unique business.
06 Scaling what works
Once you've identified campaigns with strong ROAS, the natural next step is to scale them. However, scaling often leads to diminishing returns. As you increase spend, you exhaust the most responsive audiences, and you may need to broaden targeting or accept lower ROAS. This is known as the efficiency frontier: there is a point where additional spend yields less incremental revenue. To scale effectively, you must monitor ROAS closely and be willing to accept a gradual decline as long as the incremental profit remains positive. The key is to find the optimal spend level where marginal revenue equals marginal cost.
On search, scaling might involve expanding keyword lists, increasing bids on profitable terms, or entering new geographies. But beware of bidding on broader terms that attract less qualified traffic. On social, scaling could mean increasing budgets on high-performing ad sets, expanding audiences, or testing new placements. However, broader audiences may dilute performance. On retail media, scaling might involve adding more products to campaigns or increasing bids on high-converting keywords. Each platform has its own scaling levers, and the approach should be tailored.
To scale without sacrificing too much efficiency, consider these strategies: optimize your funnel to improve conversion rates, which can offset lower ROAS from higher spend; use creative refreshes to combat ad fatigue; and implement frequency caps to avoid overexposure. Additionally, test incrementality at higher spend levels to ensure that the additional revenue is truly incremental and not just cannibalizing other channels. Finally, diversify your media mix so that you're not overly reliant on one platform. By balancing search, social, and retail media, you can scale overall revenue while maintaining healthy ROAS across channels.
07 Frequently asked questions
Which platform has the highest ROAS? There is no universal answer. Search often has high ROAS for bottom-funnel conversions, social can be strong for retargeting, and retail media excels for products sold on retailer sites. The best platform depends on your product, audience, and objectives. Test each to see what works for you.
How often should I check ROAS? ROAS should be monitored regularly, but not obsessively. Daily fluctuations can be noise. Weekly or monthly trends are more meaningful. However, during active campaigns or tests, daily checks can help catch issues early. Use automated alerts for significant changes.
Can I compare ROAS across platforms? Direct comparisons are tricky due to differing attribution models and measurement methodologies. It's better to compare ROAS within a platform over time or use a unified measurement framework like MMM to assess cross-channel impact. Always consider incrementality.
What is a good ROAS? A good ROAS is one that exceeds your break-even point and aligns with your profit goals. Calculate your break-even ROAS based on your profit margin. For many businesses, a ROAS of 4:1 is a common target, but it varies widely. Focus on profitability, not just the ratio.
How do I improve ROAS? Improve targeting, creative, landing pages, and bid strategies. Test continuously. Use audience segmentation and exclude irrelevant traffic. Optimize for conversions, not clicks. Ensure your measurement is accurate and consider incrementality. Small improvements across these areas can compound into significant ROAS gains.