Structure bid adjustments for mobile versus desktop by measuring conversion rate, cost per acquisition, and average order value per device, then applying modifiers that match each device's actual performance. Increase bids where a device converts above your target ROAS and decrease bids where it underperforms. Never set device modifiers blindly—wait for at least 30 conversions per device before adjusting.
Start With Device-Level Performance Data
Most teams get this wrong by guessing that mobile always underperforms. It doesn't. Pull a device segment report in Google Ads (Campaigns → Segment → Device) and look at three metrics per device:
- Conversion rate — the percentage of clicks that convert
- Cost per acquisition (CPA) — total spend divided by conversions
- Return on ad spend (ROAS) — revenue divided by ad spend
If mobile delivers a $40 CPA against a $35 target, that's a candidate for a negative bid adjustment. If desktop hits a $28 CPA, you can afford to bid up.
Set a minimum data threshold
Don't act on thin data. Wait for at least 30 conversions per device over a 30-day window. Below that, the numbers are noisy and you'll chase statistical ghosts. For low-volume B2B accounts, extend the lookback to 60 or 90 days.
How to Calculate the Right Bid Adjustment
Use a conversion-rate ratio to size the modifier. The formula:
Bid adjustment % = (Device conversion rate / Baseline conversion rate − 1) × 100
Example: desktop converts at 4%, mobile at 2.4%. If desktop is your baseline, mobile's adjustment is:
(2.4 / 4 − 1) × 100 = −40%
That's a starting point, not a final answer. Layer in CPA and downstream lead quality before committing. A mobile lead that converts to a closed deal less often deserves a deeper cut than conversion rate alone suggests—similar to how you'd score lead quality in an inbound versus outbound pipeline analysis.
Device modifier ranges by goal
| Goal | Mobile adjustment | Desktop adjustment |
|---|---|---|
| Lead gen (long forms) | −20% to −40% | 0% to +20% |
| E-commerce | −10% to +30% | 0% to +15% |
| Local / click-to-call | +20% to +50% | −10% to 0% |
| B2B SaaS demos | −30% to −50% | +10% to +30% |
B2B buyers usually research on mobile but convert on desktop, which is why complex sales workflows—the kind covered in MEDDIC and BANT qualification—often justify aggressive desktop weighting.
Account for Campaign Type Differences
Bid adjustment logic shifts depending on the platform mechanics.
Search campaigns
Device bid modifiers apply at the campaign or ad group level. Set a base bid for your highest-performing device, then adjust the others relative to it. As of recent Google Ads versions, you can set modifiers from −100% (exclude entirely) to +900%.
Performance Max and Smart Bidding
If you run Target CPA or Target ROAS bidding, manual device modifiers are mostly ignored—the algorithm already optimizes per device. You can still apply a −100% modifier to fully exclude a device, but partial adjustments lose effect. Google's documentation on bid adjustments spells out which strategies honor manual modifiers.
Tablet handling
Tablets sit in a gray zone. Many advertisers fold tablet performance into desktop because user behavior is closer. Check your segment report—if tablet CPA tracks desktop, group them mentally and bid accordingly.
Common Mistakes to Avoid
- Setting modifiers before conversion tracking is verified. Garbage data produces garbage adjustments.
- Ignoring landing page experience. A −50% mobile bid often masks a broken mobile checkout. Fix the page before cutting spend.
- Stacking conflicting modifiers. Device, location, and audience adjustments multiply together. A −30% device modifier plus a +50% audience modifier nets to +5%, not +20%.
- Forgetting to revisit. Device performance drifts. Audit modifiers every 30 to 60 days.
How modifiers stack
The combined adjustment multiplies, it doesn't add:
Final bid = Base bid × (1 + device mod) × (1 + audience mod) × (1 + location mod)
$2.00 base × (1 − 0.30) × (1 + 0.20) = $1.68. Track this so you don't accidentally overspend.
A Practical Workflow
- Run a 30-day device segment report.
- Confirm at least 30 conversions per device.
- Calculate conversion-rate ratios against your best device.
- Cross-check CPA and lead quality.
- Apply conservative modifiers (start at half your calculated number).
- Wait 14 days, re-measure, then refine.
- Audit every 30 to 60 days.
This disciplined loop matters most for accounts feeding a sales team, where a cheap-but-low-quality mobile lead wastes rep hours—the same downstream cost logic that drives SDR build-versus-outsource decisions.
Key Takeaways
- Base every bid adjustment on device-specific conversion rate, CPA, and ROAS, not assumptions.
- Require 30+ conversions per device before acting.
- Use the conversion-rate ratio formula to size modifiers, then halve it as a safety margin.
- Smart Bidding strategies override manual modifiers except full exclusions.
- Remember modifiers multiply across device, audience, and location layers.
- Re-audit every 30 to 60 days because device behavior shifts.