New Mobile and Desktop Traffic: How to Test Conversion Rates and Allocate Your Advertising Budget

The same campaign can deliver different results on smartphones and computers. Mobile traffic may generate more clicks, while desktop traffic may show a higher conversion rate at the form completion or purchase stage. However, the balance depends on the offer, traffic source and funnel.

The device itself is not always the reason. Results are affected by the traffic source and placement, page speed, creative format, landing page usability, the complexity of the target action and the context in which a person sees the ad.

These differences are especially important when working with Mobile and In-App traffic, where ad interactions and user behavior can differ significantly from desktop scenarios.

To allocate your budget correctly, you need to assess the economics of mobile and desktop traffic separately and understand which segment generates more approved conversions at an acceptable cost.

Why Results Differ

Device type affects how users interact with ads, but it does not explain the difference in conversion rates on its own. First, you need to check what actually differs between the segments.

Different Contexts for Ad Exposure

On a smartphone, a user may see an ad while traveling, taking a break or scrolling through a feed. In this situation, they may click the ad but put off a lengthy registration or purchase.

On a computer, the interaction often takes place in a different context: a larger screen and a more familiar way of working with forms can make a complex target action easier to complete.

This does not mean that mobile traffic always has a lower CR. For a short form, a call, an app install or another simple action, a smartphone may be the primary device.

Different Traffic Sources and Placements

Mobile and desktop traffic often come from different placements, formats and audiences.

If the mobile segment received most of its impressions in apps while desktop traffic came from search, the difference in CR may be related to the source or placement rather than the device.

First, you need to break down the data at least by source, campaign and placement. For example, comparing mobile traffic from TikTok Ads with desktop traffic from Google Search will not establish whether the device itself caused the difference in conversion rates.

Landing Page and Form

A page displaying correctly on different devices does not guarantee an equally convenient path to conversion.

On mobile, results can suffer because of small buttons, long forms, overlapping elements, inconvenient input fields and a CTA that disappears while scrolling.

On desktop, users may leave the page because of overly stretched-out content, unclear emphasis or a form that is difficult to find.

That is why you need to check the entire user journey: from clicking the ad to completing the target action. Sometimes the problem is not the cost of traffic but a specific funnel stage where users leave the page.

Speed and Technical Errors

A mobile user may open the page over an unstable connection or on a less powerful device. Heavy images, third-party scripts and slow form loading increase the risk of losing visits before the landing page has fully loaded.

You also need to check browsers, operating systems and device versions separately. An overall mobile traffic report can hide a problem that affects only a specific combination of operating system and browser.

Before launching ads, you should also check that forms, analytics, redirects and server infrastructure work correctly. These steps are covered in more detail in our checklist for preparing a landing page and VPS for a traffic launch.

Which Metrics to Compare

Evaluating segments by CR alone is not enough. A higher conversion rate does not guarantee a lower CPA or greater profit.

For mobile and desktop traffic, you need to check the following separately:

• CPM and CPC to identify differences in traffic acquisition costs;
• CTR to assess the response to the creative;
• the number of clicks and landing page opens, if tracked separately;
• CR at every available stage of the funnel;
• CPA or CPL;
• approval rate and lead quality;
• revenue, profit and ROI;
• available traffic volume.

For example, a higher CR for desktop traffic may not offset a more expensive click. That is why you need to compare CPA, revenue and profit for each segment.

It is especially important to distinguish between initial CPA and the actual cost per approved conversion. If one segment generates inexpensive leads but a significant share of them are not approved, its actual performance may be lower.

If you would like a more detailed explanation of the formulas and relationships between metrics, we recommend our article on ROI, ROAS, CPA, EPC, CR and other traffic arbitrage metrics.

How to Set Up a Fair Test

The comparison should be conducted under conditions that are as similar as possible. If GEOs, creatives, audiences and landing pages all differ between mobile and desktop traffic at the same time, identifying the reason for the results will be difficult.

Check Your Tracking

Before launching, make sure device type is identified correctly and displayed in the report, and that the click ID is preserved throughout the user journey.

Check whether the tracker receives conversion data, its status and the payout amount through a postback. For mobile traffic, test redirects, deep links and pages that open in in-app browsers separately.

If your analytics system tracks landing page opens separately, compare their number with clicks for each device type. A noticeable difference may indicate slow page loading, a technical error or an incorrect redirect.

You should also consider that a user may click an ad on a smartphone and complete the purchase on a computer. In such cases, device-level data may differ depending on the attribution model and analytics capabilities.

You can read more about click IDs, postbacks and data collection in our article “Traffic Arbitrage Tracker: How It Works and What Data It Collects”.

Keep the Test Foundation Consistent

The test conditions depend on its objective.

To check the impact of device type, you should not change the offer, GEO, source, audience, creative or funnel logic during the test.

However, even with identical settings, advertising algorithms may distribute impressions differently between segments. Results should therefore be evaluated with the actual audience composition and placements in mind.

If the team is comparing ready-to-run mobile and desktop versions of a campaign setup, creatives and landing pages can be adapted, but the offer and key message should remain comparable.

Do not change the form, offer and advertising approach at the same time. Otherwise, it will be difficult to determine what actually affected CR.

Do Not Draw Conclusions Based on Clicks Alone

There is no universal sample size: it depends on the expected difference between segments, conversion frequency, traffic costs and the available test budget.

If one segment received 500 clicks and the other received 40, it is too early to compare their CR. Both segments first need to accumulate enough data.

The number of conversions matters just as much as the number of clicks. The difference between two segments may be random, especially if each received only a few leads.

Check Results After Approval

An initial conversion shows only part of the picture. If mobile traffic generates more leads but they are approved less often, its actual economics may be worse.

In lead generation and other models with delayed approval, it is better to reallocate the budget after receiving up-to-date conversion statuses.

It is also important to compare segments with the same amount of time allowed for lead approval, so that new conversions do not appear less effective simply because they have not yet been reviewed.

What to Check Before Reallocating the Budget

Before increasing spend on one of the segments, make sure that:

• the GEO, offer and target action remain comparable;
• the tracker correctly passes conversions and their statuses;
• there are no critical technical issues on the mobile or desktop landing page;
• enough approved conversions have been collected for evaluation;
• budget growth does not cause an unacceptable increase in CPA.

How to Allocate Your Advertising Budget

There is no fixed ratio between mobile and desktop traffic. The allocation depends on traffic costs, CPA, approval rate, revenue and the available volume in each segment.

At the Start

If historical data is available, the initial budget can be allocated based on previous volume and CPA.

If there is no data, you need to set aside an amount that allows both segments to complete a full test.

A mechanical 50/50 split is not always appropriate: one segment may have significantly more available traffic or a higher cost per click.

It is better to link the test budget to your target CPA and the number of conversions needed for evaluation.

For example, if the expected CPA is $20, a $40 budget allows you to expect only two conversions on average. This volume is usually insufficient to compare the results of two segments with confidence.

After the Test

Based on the test results, the budget is reallocated with CPA, approval rate, revenue and available traffic volume in mind.

Before reallocating, you need to answer three questions:

• does the current CPA have enough headroom relative to the target value;
• are approval rate and revenue holding up;
• is there enough traffic volume for the new budget.

Moving the entire budget in one step is risky. A gradual increase lets you check how CPA changes and keep part of the budget in the other segment for comparison.

It is important to evaluate not only the average CPA for the entire test but also the cost of additional conversions after scaling. A segment that was profitable on a small budget will not necessarily maintain the same efficiency as spend increases.

We covered gradual increases in ad spend and maintaining profitability in more detail in our article “How to Maintain ROI When Scaling”.

How to Control Spend After the Test

After changing budgets, actual ad spend needs to be checked against the new allocation across campaigns.

To do this, it is important to compare data from advertising accounts and the tracker with actual charges, taking into account possible fees and differences in transaction reporting times.

With Pay2.House, you can use virtual cards to pay for ads, view transaction history and manage team finances. This helps control ad spend when testing and scaling campaigns.

For teams working with multiple advertising accounts and projects, separate cards and transaction monitoring simplify payment tracking.

However, to analyze the performance of mobile and desktop traffic, you need to use advertising platform or tracker reports, because card transactions alone do not show which device type the spend was attributed to.

Example of Budget Allocation

A team is testing one offer with mobile and desktop traffic.

Mobile traffic generates 40 conversions at a CPA of $18, while desktop traffic generates 20 conversions at a CPA of $15.

At first glance, the desktop segment appears more efficient because its initial cost per conversion is lower.

However, the results change once lead statuses are updated.

Let us look at a hypothetical example:

Although mobile traffic had a higher initial CPA, it turned out to be cheaper per approved conversion once the approval rate was taken into account.

Suppose the payout per approved conversion is $35 in both segments.

The mobile segment then generates $1120 in revenue on $720 in spend, while the desktop segment generates $350 in revenue on $300 in spend.

Under these conditions, the mobile segment generates $400 in profit and the desktop segment generates $50, excluding additional costs and fees.

Therefore, relying on initial CPA alone would have been a mistake.

In addition, the available volume of desktop traffic may be limited, and increasing the bid may cause CPA to rise.

In this situation, there is no reason to move the entire budget to the desktop segment. The team can keep desktop traffic within a stable volume while optimizing mobile traffic by placement and landing page version.

These figures are hypothetical and illustrate the calculation method rather than a statistically proven advantage of mobile traffic.

When It Is Better to Separate Devices

Separate campaigns or groups make sense when segments require different creatives, landing pages, bids, budgets or schedules.

Separation also helps if one segment receives almost all the traffic and prevents the other from collecting enough data.

However, not every advertising platform requires manual separation. Automated strategies may consider device type as one of the signals when optimizing impressions and bids.

For example, in Google Ads, automated bidding strategies may consider device type when optimizing for conversions. Therefore, manually separating campaigns is not always necessary.

TikTok Ads offers targeting settings based on mobile device characteristics, including operating system and version. However, comparing iOS and Android is not a direct equivalent of a Mobile vs Desktop analysis.

Artificially splitting campaigns can reduce the amount of data available for automated optimization. First, you need to check the capabilities of the specific advertising platform and buying strategy.

Common Mistakes

The following mistakes most often get in the way when analyzing mobile and desktop traffic:

• comparing segments with different GEOs, sources or offers without accounting for those differences;
• evaluating results only by CR or initial CPA without considering approval rate and revenue;
• drawing conclusions from a small number of conversions;
• ignoring technical landing page issues and attribution specifics;
• shifting the budget sharply after a short test;
• splitting campaigns in a way that interferes with the platform’s automated optimization.

Conclusion

Mobile and desktop traffic should be compared under conditions that are as comparable as possible and evaluated based on the campaign’s actual economics.

A higher CR or lower initial CPA does not necessarily mean that a segment generates more profit. It is important to consider approved conversions, traffic quality, revenue and scaling potential.

The budget should be increased for the segment that maintains an acceptable CPA, generates approved conversions and has enough volume for further traffic acquisition.