Book Your FREE Strategy Session
If your business invests $50K+ a month in paid search, you can claim a complimentary deep-dive audit with one of our senior PPC specialists.
In this 1-on-1 session, we’ll:
- Pinpoint exactly where spend is being wasted
Share proven strategies to cut wasted spend and scale ROI
Benchmark your performance against top competitors in your space
Others charge $1,000+ for this level of strategic insight, we provide it free to qualified advertisers.
Clients and Results
”Launching a startup airline comes with big challenges. Partnering with Search Machines was the best decision we made. Their expert team and ATLAS software helped us scale efficiently with multilingual ads reaching markets as far as Brazil and Canada. Thanks to their expertise, Arajet has experienced 40% growth in revenue year on year
Tania MancillaVP of Marketing, Arajet
Separating desktop and mobile campaigns increased revenue by 22% in a mature OTA account
Search Machines · May 2026 · 4 min read
We noticed something strange while reviewing a mature OTA account.
Desktop had quietly become the most profitable segment. Its ROAS was almost double that of mobile. But at the same time, desktop traffic was falling.
That combination didn’t make much sense. If desktop was consistently returning more value, you would expect Google’s value-based bidding system to send more traffic there, not less. Yet month after month the trend was the same – stronger performance on desktop, declining volume.
Nothing else in the account suggested a problem. Targets were being hit, and overall performance looked stable. On the surface, everything was working. But underneath that stability, the most profitable segment in the account was slowly shrinking.
We took a closer look.
What we found
Desktop and mobile were running inside the same campaign structures, sharing a setup, a target, and a single learning system.
In theory that shouldn’t prevent the system from favouring the better-performing device. Value-based bidding has access to device signals and is designed to allocate spend toward the highest-value traffic.
But inside consolidated structures, those differences can become diluted. The algorithm optimises toward the blended performance of the campaign rather than aggressively scaling the strongest segment inside it.
In this case, the gap between desktop and mobile was being absorbed into that blended optimisation.
Desktop continued to outperform mobile, but the system wasn’t prioritising it. Over time, that meant the account was gradually underinvesting in its most profitable device.
The intervention
We decided not to touch targets or budgets – the problem wasn’t a bidding issue. It was structural.
If the system was optimising toward a blended outcome, then our theory was that separating the devices might allow the stronger segment to stand on its own.
So we split the campaigns by device.
One campaign was dedicated to desktop. One campaign was dedicated to mobile.
Nothing else changed. The ROAS targets stayed the same. Total budgets stayed the same. Ads and landing pages were left exactly as they were.
The change went live on January 12th.
What changed
The response was immediate.
- Desktop clicks increased by 72%.
- Desktop conversions increased by 46%.
- Desktop ROAS settled back toward target levels.
Mobile continued to behave much as it had before. It followed its normal seasonal pattern and didn’t show any negative impact from the change.
The main difference was what happened to desktop. Instead of declining after peak season, it continued to scale.
Across the campaigns affected by the change, that additional desktop traffic resulted in a 22% increase in total revenue compared to the previous structure.
No additional budget was introduced.
The demand had always been there. The account structure had simply been preventing the system from fully acting on it.
Explore more case studies, or book a strategy session to talk through your account.
