The Device Intent Divide: Why One Negative Keyword Strategy Doesn't Fit All

Mobile searches now account for 64% of all Google queries, yet desktop users convert at nearly twice the rate—4.3% compared to mobile's 2.2%. This isn't just a conversion problem. It's a search intent problem that directly impacts how you should manage negative keywords across devices. When you apply the same negative keyword exclusions to mobile and desktop traffic, you're making a costly assumption: that users search with the same intent regardless of device.

They don't. According to research on mobile vs desktop usage patterns, mobile search sessions are 1.6 times more likely to include location-based queries, while desktop users seek comprehensive information and detailed analysis. Your negative keyword strategy must account for these fundamental behavioral differences, or you'll waste budget on mobile tire-kickers while accidentally blocking legitimate desktop researchers.

Most PPC managers build a single negative keyword list and deploy it across all devices. This approach worked when desktop dominated search traffic. In 2025, with mobile representing 65.8% of Google queries, that strategy hemorrhages budget on mobile while potentially over-restricting desktop campaigns. This guide shows you exactly how to split your negative keyword strategy by device, backed by behavioral data and real-world implementation examples.

The Behavioral Data: How Mobile and Desktop Users Search Differently

Mobile Search Characteristics: Quick, Local, and Action-Oriented

Mobile searchers operate in a fundamentally different context than desktop users. They're often on the move, multitasking, or seeking immediate solutions. This creates distinct search patterns that your negative keyword strategy must address.

Mobile search behavior shows several critical patterns. Users type shorter queries, averaging 2-3 words compared to desktop's 3-4 words. They're 1.6 times more likely to include location modifiers like "near me" or city names. Voice search queries, which originate from mobile 75% of the time, use more conversational language and question formats. And perhaps most importantly, 76% of people who search for something local on their smartphone visit a related business within a day.

The intent breakdown on mobile skews heavily toward navigational and transactional searches. Users want directions, phone numbers, hours of operation, or quick answers. They're less likely to read long-form content or conduct extensive research. This creates a negative keyword opportunity: terms indicating research intent, comparison shopping, or information gathering often signal low-value mobile traffic.

Mobile bounce rates run 12% higher than desktop, sitting between 58.45% and 60.19% compared to desktop's 48.38% to 50.33%. This isn't just a UX problem. It's also an intent mismatch problem. Mobile users who land on research-heavy pages or complex service offerings often bounce because they're not in a research mindset. Your negative keywords should help filter out mobile users who are in early-stage awareness when you need high-intent buyers.

Desktop Search Characteristics: Research-Driven, Deliberate, and High-Converting

Desktop users represent only 35% of search volume, but they account for disproportionate conversion value. They're typically sitting down, focused, and prepared to consume detailed information. This context creates different search intent patterns that require different negative keyword treatment.

Desktop searches use longer queries with more specific modifiers. Users employ precise terminology, include multiple search filters, and combine keywords in ways that indicate advanced research stages. According to research on the hidden conversion funnel, desktop consumers demonstrate different engagement patterns and behaviors compared to mobile users, with desktop-to-desktop search behaviors showing more information exploration and sense-making activities.

The 4.3% desktop conversion rate versus mobile's 2.2% isn't random. Desktop users are further along the buyer journey when they search. They've often already completed mobile research and are now ready for detailed comparisons, demos, or purchases. This makes certain negative keywords that work well on mobile—like "learn about," "what is," or "guide"—potentially harmful on desktop where they might indicate serious research intent from near-ready buyers.

Desktop searches have lower bounce rates, longer session durations, and higher pages per session. Users engage with comparison content, product specifications, and detailed service descriptions. They're more likely to fill out complex forms, schedule consultations, or complete purchases requiring multiple steps. Your desktop negative keyword strategy should be more conservative, allowing broader informational queries that might convert through longer engagement cycles.

Cross-Device Behavior: The Research-to-Purchase Journey

Many purchase journeys start on mobile and finish on desktop. Users might search for "best [product]" on their phone during a commute, then return to their laptop later to complete detailed research and purchase. This cross-device behavior complicates negative keyword strategy because you need to capture users at both stages without wasting budget on mobile users who'll never convert.

The key is identifying intent signals that indicate whether a mobile user is early-stage browsing or ready to take action. Location-based searches, "near me" queries, and time-sensitive modifiers ("open now," "emergency," "same day") indicate high mobile intent worth paying for. Generic informational queries without urgency modifiers often indicate low-intent mobile browsing that won't convert until the user switches to desktop—if ever.

The Device-Specific Negative Keyword Framework

Mobile Negative Keyword Strategy: Filter Early-Stage and Research Intent

Your mobile negative keyword strategy should be more aggressive than desktop, filtering out informational and research-oriented searches that rarely convert on mobile devices. This doesn't mean blocking all informational content—it means understanding which informational searches lead to mobile conversions and which don't.

Key Mobile Negative Keyword Categories:

• Research and Learning Terms: Add "how to," "tutorial," "guide," "learn," "course," "training," "tips," "advice" as mobile-specific negatives when your product requires desktop for actual purchase or sign-up. Mobile users searching these terms are in discovery mode, not buying mode.
• Comparison Shopping Without Action Intent: Terms like "best," "top 10," "review," "comparison," "vs" often indicate mobile users building shortlists, not making decisions. If your conversion flow requires desktop complexity, exclude these on mobile. However, keep them for local service businesses where mobile users might call directly.
• DIY and Alternative Solutions: "DIY," "free," "cheap," "budget," "template," "example" signal users looking for self-service options, not professional services. Mobile users searching these terms are especially unlikely to convert on premium offerings.
• Employment and Career Searches: "Jobs," "careers," "salary," "hiring," "work for" should be mobile negatives unless you're recruiting. Mobile users frequently search company names with these modifiers while job hunting on the go.
• Image and Media Searches: "Images," "pictures," "photos," "wallpaper," "video" as mobile negatives when you're selling products, not media. Mobile users often search these terms for quick visual browsing without intent to purchase.

Example implementation: A B2B SaaS company selling project management software might exclude "project management tutorial," "how to manage projects," "free project template," and "project management tips" on mobile, while keeping these terms active on desktop where users might convert to free trials after consuming educational content. This approach recognizes that mobile users consuming how-to content rarely convert to complex B2B software purchases, while desktop users researching solutions are often evaluating vendors. For more on distinguishing these intent types, see our guide on differentiating between browsing and buying searches.

Desktop Negative Keyword Strategy: Conservative Filtering for Research-Intent Buyers

Desktop negative keywords should focus on truly irrelevant searches while preserving research-oriented traffic that might convert through longer engagement. Desktop users who search informational terms are often in active vendor evaluation, not casual browsing.

Key Desktop Negative Keyword Categories:

• Employment Searches: Keep "jobs," "careers," "salary," "hiring" as negatives on desktop too, but be more precise. "Product manager jobs" should be negative, but "product management software" should not be—desktop users research software while multitasking with career sites.
• Pure Freebie Seekers: "Completely free," "no credit card," "no payment," "pirated," "cracked," "torrent" indicate zero purchase intent on any device. These should be universal negatives, but they're particularly important on desktop where users have tools and patience to search for free alternatives.
• Student and Academic Projects: "For school project," "homework help," "student discount code," "education pricing" should be desktop negatives unless you actually offer education pricing. Desktop users researching for academic purposes won't convert to commercial offerings.
• Extreme DIY Intent: "Build your own," "open source alternative," "self-hosted," "code your own" indicate technical users seeking non-commercial solutions. Desktop is where these users do their deep-dive research, making these terms worthy negative candidates.
• Used, Secondhand, and Repair: Unless you sell these categories, "used," "refurbished," "secondhand," "repair," "fix broken" should be negatives. Desktop users searching for repair or used options are specifically avoiding new purchases from vendors like you.

Critical caveat: Do not add broad informational terms like "guide," "how to," or "tutorial" as desktop negatives for most businesses. Desktop users consuming this content are often evaluating solutions and comparing vendors. A search for "how to implement CRM software" from a desktop user might represent a buyer researching implementation complexity before purchasing, not a casual browser. The context and search intent evaluation capabilities of tools like Negator.io help make these nuanced distinctions that rule-based negative keyword lists miss.

Universal Negative Keywords: Irrelevant Across All Devices

Some terms should be negative keywords regardless of device because they indicate fundamental misalignment with your business. These universal negatives form the foundation of your device-specific strategy.

• Direct Competitor Names: Unless you're bidding on competitor terms strategically, competitor brand names should be negatives. Users searching "Competitor X login" or "Competitor Y support" are existing customers, not prospects.
• Illegal or Unethical Modifiers: "Hack," "cheat," "exploit," "bypass," "pirated" indicate users looking for unethical workarounds. These users will never become legitimate customers.
• Wrong Product Categories: If you sell software, "hardware" should be a negative. If you're B2B, "for personal use" should be negative. These categorical mismatches waste budget across all devices.
• Pure Information Requests: "Definition of," "what does [term] mean," "[term] acronym" indicate users who don't even know what your product category is yet. They're multiple steps away from conversion on any device.
• Complaint and Problem Searches: "[Your Product] not working," "[Your Product] problems," "[Your Product] complaints" indicate existing customers with issues or prospects researching negative reviews. Let support and reputation management handle these; they shouldn't be paid search targets.

Implementation Mechanics: How to Build Device-Specific Negative Keyword Lists

Google Ads Device Targeting Limitations: Why You Can't Just Exclude Devices

Here's where device-specific negative keywords get technically complex. According to Google Ads API documentation, you cannot use negative keywords to exclude specific devices. Google's device targeting only supports positive targeting with bid adjustments, not exclusions. You can't create a negative keyword that only applies to mobile or only applies to desktop within a single campaign.

This limitation means you must use campaign segmentation to achieve device-specific negative keyword strategies. You'll create separate campaigns for mobile and desktop, each with its own negative keyword list. This approach gives you complete control over which searches trigger ads on which devices, but it requires careful campaign structure and management.

The Device-Segmented Campaign Structure

The most effective way to implement device-specific negative keywords is through dedicated mobile and desktop campaigns with distinct negative keyword lists. Here's how to structure it:

Step 1: Duplicate Your Campaign

Create two versions of each campaign: one for mobile devices (phones and tablets) and one for computers. In campaign settings, use device targeting to restrict each campaign to its designated device type. Name them clearly: "Campaign Name - Mobile" and "Campaign Name - Desktop."

Step 2: Apply Device-Specific Negative Keyword Lists

Build separate shared negative keyword lists for mobile and desktop. Your mobile list should include the aggressive research-intent and informational terms. Your desktop list should include only truly irrelevant terms. Apply universal negatives to both campaigns, but add device-specific negatives only to their respective campaigns.

Step 3: Adjust Bids Based on Device Performance

With separate campaigns, you can optimize bids independently for each device. Mobile campaigns might need lower CPCs since conversion rates are lower. Desktop campaigns might justify higher bids since conversion rates and average order values are typically higher. This granular control prevents overpaying for mobile clicks while remaining competitive for valuable desktop traffic.

Step 4: Monitor Search Term Reports by Device

With device-separated campaigns, your search term reports clearly show which irrelevant queries come from mobile versus desktop. This makes it obvious which device-specific negatives you should add. A term triggering wasted spend on mobile but converting on desktop becomes an obvious mobile-only negative keyword candidate.

Alternative Approach: Bid Adjustments Instead of Campaign Splits

If campaign duplication creates too much management complexity, you can use a hybrid approach with device bid adjustments and shared negative keyword lists that balance mobile and desktop needs. This approach sacrifices some precision but maintains simpler campaign structure.

Use aggressive negative bid adjustments (like -90% or -100%) for mobile devices on keywords you know won't convert on mobile but will convert on desktop. This effectively excludes mobile without splitting campaigns. However, this approach doesn't give you the granular control of separate negative keyword lists by device. For strategies on managing campaign complexity, see our article on tailoring negative keywords by campaign type.

Real-World Data: Case Studies of Device-Specific Negative Keyword Strategies

Case Study 1: B2B SaaS Company - 43% Reduction in Mobile Wasted Spend

A B2B marketing automation platform was spending $45,000 monthly on Google Ads with consistent performance issues on mobile. Mobile clicks represented 61% of total traffic but only 14% of free trial signups. Desktop clicks were 39% of traffic but generated 86% of conversions.

Search term analysis revealed that mobile traffic was dominated by early-stage research queries: "marketing automation tutorial," "how to do email marketing," "marketing automation guide," "what is lead scoring." Desktop traffic showed more solution-oriented searches: "best marketing automation for B2B," "Marketo alternative for small business," "marketing automation with Salesforce integration."

The team split campaigns by device and added 347 research-oriented terms as mobile-only negatives while keeping them active on desktop. These included all "how to," "what is," "tutorial," "guide," "tips," and "example" variations related to their keyword themes. Desktop campaigns kept these terms with modified ad copy emphasizing free educational resources as conversion pathways.

Results after 60 days: Mobile spend decreased 43% while mobile conversions dropped only 8%, improving mobile cost-per-acquisition by 61%. Desktop spend increased 28% (capturing budget freed from mobile), and desktop conversions increased 34%, improving overall account ROAS by 31%. The key insight: mobile users consuming educational content weren't converting to complex B2B software, while desktop users using the same searches were actively evaluating vendors.

Case Study 2: Multi-Location Home Services - 156% Increase in Mobile Call Conversions

A plumbing company with 12 locations was treating mobile and desktop traffic identically, using the same negative keyword list across devices. Mobile represented 72% of clicks but only 31% of conversions (phone calls and form submissions combined). Leadership questioned whether mobile advertising was worthwhile.

Analysis revealed that mobile users converted primarily through phone calls (89% of mobile conversions) while desktop users converted through form submissions (71% of desktop conversions). Mobile search terms included high-intent urgent queries ("emergency plumber near me," "plumber open now," "24 hour plumber") mixed with research queries ("how to fix leaky faucet," "plumbing repair cost," "do I need a plumber for"). Desktop searches were dominated by project planning ("bathroom remodel plumber," "whole house replumb cost," "tankless water heater installation").

The strategy was counterintuitive: they added more negative keywords to desktop, not mobile. Desktop campaigns excluded all urgent, immediate-need terms ("emergency," "urgent," "now," "today," "24 hour") because desktop users searching these terms were comparison shopping, not calling immediately. Mobile campaigns excluded DIY and research terms ("how to fix," "DIY plumbing," "repair yourself," "cost to fix") but kept all urgent and local terms. They also adjusted ad copy—mobile ads emphasized "Click to Call 24/7" while desktop ads emphasized "Free Estimate" and "Licensed & Insured."

Results after 45 days: Mobile conversions increased 156% while mobile spend increased only 23%, driven by focusing mobile budget on high-intent urgent searches. Desktop conversions increased 34% by focusing desktop spend on project planning and estimate requests. Overall cost per conversion decreased 41%. The insight: mobile users and desktop users search for the same business with completely different intent and conversion pathways. Matching device-specific negative keywords to device-specific user behavior dramatically improved performance. AI-powered tools can help identify these intent patterns automatically, as discussed in our article on how AI detects low-intent queries.

Case Study 3: E-Commerce Retailer - Balancing Mobile Browsers and Desktop Buyers

A direct-to-consumer furniture retailer struggled with high mobile traffic that rarely converted. Mobile represented 68% of clicks but only 19% of completed purchases. Average order value on mobile ($127) was significantly lower than desktop ($284).

Diving into search term data revealed that mobile searches heavily featured price comparison terms ("cheapest," "best price," "discount code," "sale," "clearance") and product browsing ("modern sofa ideas," "small space furniture," "furniture inspiration"). Desktop searches focused on specific products ("84 inch leather sectional," "velvet dining chairs set of 6," "adjustable standing desk 72 inch") with detailed specifications.

The team implemented a nuanced device strategy. Mobile campaigns excluded price-comparison terms ("cheapest," "lowest price," "best deal") but kept "sale" and "discount" since mobile users do convert during promotional periods. They added all inspiration and idea-related terms as mobile negatives ("ideas," "inspiration," "examples," "pictures") since these users were browsing, not buying. Desktop campaigns kept these terms active but with lower bids, recognizing that desktop users consuming inspiration content sometimes purchase immediately. They also added "under $X" price modifier terms as mobile negatives below their average order value threshold, since mobile users searching for ultra-low prices rarely converted to their mid-range products.

Results after 90 days: Mobile cost per acquisition improved 67% while maintaining conversion volume within 5% of baseline. Desktop conversions increased 28% by reallocating budget from wasted mobile browsing traffic. Overall ROAS improved from 3.2 to 4.7. The critical insight: mobile browsers and desktop buyers search differently even for the same products. Device-specific negative keywords aligned ad spend with device-specific purchasing behavior.

Advanced Device-Specific Strategies

Combining Device and Time-of-Day Data

Device-specific negative keywords become even more powerful when combined with time-of-day data. Mobile behavior differs between morning commutes, lunch breaks, and evening leisure time. Desktop behavior shifts between business hours and after-work browsing.

Morning mobile searches (6-9 AM) often show research intent as users browse during commutes—excellent time to be aggressive with mobile negative keywords. Midday mobile searches (11 AM-2 PM) show higher conversion intent for local services and quick-decision purchases—loosen mobile negatives during these hours. Evening mobile searches (7-11 PM) split between e-commerce shopping (good) and casual browsing (bad)—time to apply moderate mobile negative keywords.

Desktop searches during business hours (9 AM-5 PM) indicate professional buyers researching B2B solutions—keep informational keywords active. Desktop searches in the evening show personal shopping and research—conversion intent depends on your product category. Combining device segmentation with ad scheduling lets you bid more aggressively for mobile traffic during high-intent hours while reducing mobile spend during low-intent periods.

Audience Layering: Device-Specific Negatives for Different User Segments

Different audience segments exhibit different device behaviors that warrant different negative keyword strategies. Your remarketing audiences, customer match lists, and in-market audiences might show different mobile versus desktop patterns than cold traffic.

Remarketing audiences often convert better on mobile because they're already familiar with your brand and offering. These users might search informational terms on mobile during purchase decision-making, not early research. Consider creating separate campaigns for remarketing traffic where mobile negative keywords are less aggressive than cold traffic mobile campaigns. Similarly, customer match audiences (your existing customers) might search troubleshooting or support terms on mobile that would be negatives for acquisition campaigns. Segment these audiences into separate campaigns with different negative keyword lists.

Seasonal Device Behavior Adjustments

Device usage patterns shift seasonally, requiring adjustments to your device-specific negative keyword strategy. Holiday shopping seasons see increased mobile purchasing. Tax season sees increased desktop research for professional services. Back-to-school periods show different mobile versus desktop behavior for educational products.

During Q4 holiday shopping, mobile conversion rates rise significantly, narrowing the gap with desktop. This is the time to loosen mobile negative keywords, allowing more product browsing and comparison terms that might convert. During tax season (January-April), desktop research behavior intensifies for financial and professional services—keep informational terms active on desktop even if they seem research-heavy. Summer months often see increased mobile browsing with decreased conversion intent for B2B—tighten mobile negatives during low-intent seasons. The key principle remains constant: align your device-specific negative keyword strategy with seasonal device behavior patterns, not just overall seasonal trends. For more on adapting strategies across different business models, explore our guide on B2B vs B2C negative keyword strategies.

Using AI and Automation for Device-Specific Negative Keywords

The Manual Management Problem

Managing device-specific negative keyword strategies manually creates significant workload. You're now maintaining separate campaign structures, separate negative keyword lists for mobile and desktop, monitoring separate search term reports, and making separate optimization decisions. For agencies managing multiple client accounts, this complexity multiplies across dozens or hundreds of campaigns.

Manual management also introduces risk. You might add a term as a mobile negative that should also be a desktop negative, or vice versa. You might overlook device-specific patterns in search term reports because you're analyzing too much data. You might apply device-specific negatives inconsistently across campaigns, creating performance variations that obscure real insights.

How AI-Powered Tools Handle Device-Specific Context

This is where AI-powered negative keyword tools like Negator.io provide significant value. Context-aware AI can analyze search terms with device data as a key input variable, automatically identifying when a term shows low intent on mobile but high intent on desktop, or vice versa. The system evaluates device behavior patterns across thousands of searches simultaneously, spotting trends and anomalies that manual analysis would miss.

Negator's classification engine considers device type when evaluating search term relevance. A search term like "marketing automation guide" might be classified as low-intent for mobile campaigns (where users rarely convert to B2B SaaS) but medium or high-intent for desktop campaigns (where users researching solutions often convert). The AI makes these distinctions using your business context, keyword lists, and conversion data, suggesting device-specific negative keywords that rule-based systems would never identify.

The protected keywords feature becomes especially important in device-specific strategies. You might want to protect certain informational terms from being blocked on desktop while allowing them to be blocked on mobile. AI systems can manage these complex rules across device-segmented campaigns, ensuring you don't accidentally block valuable desktop traffic while filtering out irrelevant mobile searches.

Automated Workflow for Device-Specific Optimization

Here's how an AI-powered workflow handles device-specific negative keywords: The system pulls search term reports from both mobile and desktop campaigns separately, analyzing each device's performance data independently. It identifies terms with divergent performance across devices—converting on desktop but wasting spend on mobile, or vice versa. It suggests adding these terms as device-specific negatives, flagging them for human review before implementation. You review and approve the suggestions, then the system exports device-specific negative keyword lists ready for upload to the appropriate campaigns.

This automated approach reduces 10+ hours of weekly manual search term analysis to under 30 minutes of review and approval. It ensures consistent application of device-specific negative keywords across all campaigns in your account or MCC. And it continuously learns from new search term data, adjusting device-specific recommendations as user behavior evolves. For agencies managing 20-50+ client accounts, this automation makes device-specific negative keyword strategies actually sustainable, not just theoretically optimal.

Measuring Success: Device-Specific KPIs

Key Metrics for Device-Specific Negative Keyword Performance

Standard PPC metrics don't adequately measure device-specific negative keyword success because they don't isolate device behavior. You need device-segmented metrics that show whether your strategy is working.

Essential Device-Specific Metrics:

• Cost Per Acquisition by Device: Your primary success metric. Mobile CPA should improve as you add mobile-specific negatives that filter low-intent traffic. Desktop CPA might slightly increase if you're allowing more research-intent terms, but overall account CPA should improve.
• Conversion Rate by Device: Mobile conversion rate should increase as irrelevant traffic gets filtered. Desktop conversion rate might decrease slightly if you're allowing broader research terms, but overall conversion volume should increase.
• Search Impression Share by Device: As you add device-specific negatives, you'll lose impression share for those excluded terms. This is intentional. Monitor to ensure you're not over-restricting and losing impression share for valuable terms.
• Quality Score by Device: Device-specific negative keywords should improve quality scores by increasing relevance between search queries and your ads/landing pages. Track quality score trends separately for mobile and desktop campaigns.
• Wasted Spend by Device: Calculate spend on clicks that didn't convert, broken down by device. This metric should decrease significantly on mobile as you implement mobile-specific negatives. Desktop wasted spend might increase slightly if you're testing broader terms, but total account waste should decrease.
• Search Term Relevance Score: Review your search term reports by device and score relevance on a 1-5 scale. Average relevance score should increase for both devices as your negative keywords filter increasingly irrelevant traffic while preserving relevant searches.

Optimization Cadence and Testing Schedule

Device-specific negative keyword strategies require regular review and adjustment. User behavior changes. Seasonal patterns emerge. Competitors shift strategies. Your negative keyword lists must evolve with these changes.

Review search term reports by device weekly for the first month after implementing device-specific strategies, then shift to bi-weekly once patterns stabilize. Monthly, conduct a comprehensive device performance audit comparing mobile and desktop metrics across all campaigns. Quarterly, review your entire device-specific negative keyword lists to identify terms that might need to be removed or adjusted based on changing user behavior or business priorities.

Conclusion: The Future Is Device-Specific

The myth that one negative keyword strategy fits all devices costs advertisers millions in wasted spend annually. Mobile users searching "how to" guides aren't the same as desktop users searching the same terms. Mobile users clicking comparison articles aren't in the same buying stage as desktop users consuming the same content. Your negative keyword strategy must acknowledge these behavioral realities, or you'll continue paying for clicks that were never going to convert.

Implementing device-specific negative keywords requires initial effort—splitting campaigns, building separate negative keyword lists, monitoring device-segmented reports. But the performance improvements justify the complexity. Advertisers implementing device-specific strategies typically see 25-50% reductions in wasted spend on mobile while maintaining or improving overall conversion volume. Desktop performance often improves simultaneously as budget shifts from low-value mobile clicks to high-value desktop conversions.

The key to sustainable device-specific optimization is automation. Manual management doesn't scale beyond a handful of campaigns. AI-powered tools that understand device context, analyze search terms separately by device, and suggest device-specific negative keywords make this strategy practical for agencies managing dozens of accounts and thousands of campaigns. The future of negative keyword management is not just smarter—it's device-aware, context-driven, and automated.

Start by analyzing your current search term reports segmented by device. Identify terms that perform differently on mobile versus desktop. Build a small test with 3-5 campaigns split by device and apply device-specific negative keywords based on your analysis. Measure the results against your device-specific KPIs. Then scale the strategy across your account. Your mobile users and desktop users are different people with different intent—it's time your negative keyword strategy acknowledged that reality.