Best for
- B2B teams competing in expensive Google Ads categories
- Marketers launching search campaigns against larger incumbents
- Product marketers trying to sharpen competitive positioning
Analyze competitor paid search keywords, landing pages, and messaging patterns, then build counter-campaigns around the gaps they are leaving open.
A competitive paid search brief with keyword gaps, counter-positioning angles, and ready-to-launch Google Ads copy.
Quick view
Check the audience, core tools, and access before you start the detailed steps.
Best for
Core tools
5 tools used across the workflow.
Access
The complete workflow is available from the first step through the final result.
Why this works
Most paid search teams look only at keyword volume and CPC, then write ads that sound like everyone else. This workflow looks at the whole competitive surface: keywords, ad claims, landing page promises, proof points, and missing objections. The counter-campaign works because it does not try to outspend competitors. It finds the specific claims they overuse and the buying fears they ignore.
Expected results
Competitors analyzed
3-5 competitors
This is enough to identify repeated messaging patterns without turning the workflow into a full market research project.
Ad variations produced
40-80 ad copy assets
With 3-5 ad groups and 12 headlines plus 6 descriptions each, this range is realistic from one structured Claude prompt.
Strategy time
2-3 hours
The workflow compresses manual ad review, keyword export, and copywriting into a structured research and generation process.
Campaign quality
Higher message differentiation
The output is explicitly based on avoiding overused competitor claims, so the ads should be more distinct than category-default copy.
Step-by-step workflow
This workflow is fully available. Follow the steps below to build the system from start to finish.
60-90 min
60-90 min
Start with three to five competitors that repeatedly appear in deals or paid search results. Use Google Ads Transparency Center and AdScan AI to capture active ads, screenshots, headlines, descriptions, CTAs, destination URLs, dates observed, and visible formats. On each landing page, record the hero claim, offer, proof, form or CTA, pricing language, implementation promise, and major objections addressed. Keep dated evidence because campaigns and landing pages can change during the analysis. For each input, keep the source, collection date, owner, and confidence level so later AI summaries can be checked instead of trusted blindly.
A dated competitor swipe file linking each ad to its landing-page promise and proof.
Save the destination page with the ad. Copy alone can look differentiated until you see that every competitor sends traffic to the same generic demo page.
60-90 min
60-90 min
Use Semrush to export competitor paid keywords, estimated positions, CPC, traffic share, ad history, and destination URLs. Then use Perplexity to research a small set of high-intent queries and summarize the current SERP, including ads, organic framing, comparison pages, and buyer questions. Separate category, problem, competitor, alternative, and implementation intent in the working data. Verify important observations manually because paid results vary by geography, device, and time. Use explicit fields and reason codes rather than gut feel, and keep the evidence for each score next to the recommendation. Use the prompt template in this step only after pasting the relevant source material; require the answer to return assumptions, missing information, and a clear next action.
A keyword and SERP map showing where competitors concentrate spend and how each query is framed.
A high-CPC keyword is not automatically valuable. Check whether the searcher is close to a decision your offer can actually support.
30-45 min
30-45 min
Use Claude to combine the ad swipe file, keyword export, and SERP notes into one comparable table. Map each record to competitor, keyword intent, claim, offer, proof type, CTA, destination page, and observed weakness or inconsistency. Preserve the original copy and source URL beside the classification. A paid media owner should correct any misread intent or product terminology before gap analysis. Ask for a working draft with named sections, assumptions, gaps, and review flags rather than treating the first AI answer as final. Use the prompt template in this step only after pasting the relevant source material; require the answer to return assumptions, missing information, and a clear next action.
A normalized competitor evidence table connecting keyword intent, ad promise, landing page, and proof.
The important gap may be between the ad and landing page, not between two headlines.
45-60 min
45-60 min
Use Claude to identify repeated claims, missing buyer questions, proof gaps, weak landing-page handoffs, and intent segments that competitors treat the same. Require evidence across more than one record before labeling a claim crowded. Compare each opportunity with your approved capabilities and proof so the team does not build a counter-position around something it cannot substantiate. Review the findings with product marketing and paid media together. Sample real examples, mark pass and fail reasons, and send ambiguous cases back to the owner before anything is published, synced, or used in front of a buyer. Use the prompt template in this step only after pasting the relevant source material; require the answer to return assumptions, missing information, and a clear next action.
A ranked gap map with evidence, buyer relevance, proof requirements, and risks.
Being different is not enough. The gap must matter to the searcher and be supported by proof you can show on the page.
45-60 min
45-60 min
Use Claude to turn the approved opportunities into three to five ad groups. For each group, define the search intent, keyword theme, exclusions, counter-positioning angle, proof requirement, CTA, landing-page section, and disqualifying traffic. Keep competitor comparison terms separate from broad category terms because the buyer needs different evidence. Confirm that each proposed claim can appear legally and accurately in both the ad and landing page. Give the AI concrete source material, required fields, and a review checklist so the output can be judged by a human operator. Use the prompt template in this step only after pasting the relevant source material; require the answer to return assumptions, missing information, and a clear next action.
A build-ready campaign plan with intent, keywords, negatives, message, proof, and page requirements.
Do not force every keyword into the same landing page. A comparison searcher and an early problem-aware searcher need different answers.
45-60 min
45-60 min
Use Claude to generate responsive search ad assets for each approved ad group. Keep keyword relevance visible, avoid crowded competitor phrases, and include proof only where it fits the character limits without distortion. Produce enough variation for testing, but label headlines that should not be pinned together because they repeat or create a false sentence. The paid media manager must verify character counts, trademark rules, claims, and final combinations in the ad platform. Ask for a working draft with named sections, assumptions, gaps, and review flags rather than treating the first AI answer as final. Use the prompt template in this step only after pasting the relevant source material; require the answer to return assumptions, missing information, and a clear next action.
Reviewed headline, description, sitelink, and callout options mapped to each ad group.
Responsive ads create combinations you may never write manually. Review assets as a system, not as isolated lines.
30-45 min
30-45 min
Use Claude to compare every proposed ad group with its destination page. Check whether the page repeats the core promise, shows the required proof, answers the intent-specific objection, and presents the same CTA. List missing sections, conflicting claims, weak evidence, and mobile-first content that must be visible before launch. The landing-page owner should resolve the gaps instead of expecting ad copy to compensate for a generic page. Ask for a working draft with named sections, assumptions, gaps, and review flags rather than treating the first AI answer as final. Use the prompt template in this step only after pasting the relevant source material; require the answer to return assumptions, missing information, and a clear next action.
A page-alignment checklist for each ad group with exact changes required before launch.
Quality problems often start after the click. The best counter-positioned ad still fails if the page retreats to category-default messaging.
30-45 min
30-45 min
Use Claude to turn the approved campaign into a QA and measurement plan. Include build checks, conversion tracking, negative-keyword review, search-term review, ad-to-page consistency, brand safety, budget guardrails, and the decision thresholds for pausing or expanding. Separate early quality signals from outcomes that need more volume. After launch, compare performance by intent and message rather than declaring the whole counter-positioning idea a success or failure. Ask for a working draft with named sections, assumptions, gaps, and review flags rather than treating the first AI answer as final. Use the prompt template in this step only after pasting the relevant source material; require the answer to return assumptions, missing information, and a clear next action.
A launch checklist and learning plan with owners, thresholds, and review dates.
Review search terms early. A counter-campaign can attract irrelevant competitor traffic quickly if matching is too broad.
Research the current search landscape for these paid-search queries.
Market and geography:
{{market_and_geography}}
Queries:
{{queries}}
Known competitors:
{{competitors}}
For each query, return:
1. Likely search intent
2. Competitors or categories appearing repeatedly
3. Common ad claims
4. Common organic-result framing
5. Comparison or alternative content present
6. Buyer questions or objections visible in the results
7. Important geographic or timing limitation
8. Source URL and access date for every observation
9. Confidence level
10. What a marketer should verify manually in an incognito search
Do not claim a stable ad position or impression share from a single observation. Do not invent ad copy. Mark unavailable or variable results clearly. A paid media owner will verify the priority SERPs before strategy decisions.Normalize this competitor paid-search evidence.
Ad and landing-page records:
{{ad_and_page_records}}
Semrush keyword export:
{{keyword_export}}
SERP research:
{{serp_research}}
Return one table with:
1. Competitor
2. Query or keyword
3. Intent category
4. Ad headline and description
5. Primary claim
6. Offer and CTA
7. Destination page
8. Proof shown
9. Objection addressed
10. Message-to-page consistency
11. Possible weakness
12. Source and date
13. Confidence
Do not infer performance, spend, or conversion from visible copy. Preserve exact source text where supplied and mark missing fields. A paid media owner will verify intent and landing-page classifications.Identify credible counter-positioning opportunities from this paid-search evidence.
Normalized competitor table:
{{normalized_competitor_table}}
Our verified capabilities:
{{our_capabilities}}
Our approved proof:
{{our_proof}}
Our limitations:
{{our_limitations}}
Return:
1. Crowded claims with evidence count
2. Missing buyer questions or objections
3. Ad-to-page proof gaps
4. Intent segments competitors treat too generically
5. Credible counter-positioning opportunities
6. Proof required for each opportunity
7. Risks or limitations
8. Opportunities to reject because we cannot substantiate them
9. Recommended priority and confidence
Do not invent competitor performance or our proof. Separate observation from interpretation. A product marketing and paid media reviewer will approve the opportunities used in campaign planning.Create a Google Ads counter-campaign plan from these approved gaps.
Approved opportunities:
{{approved_opportunities}}
Keyword data:
{{keyword_data}}
Our proof:
{{our_proof}}
Landing-page options:
{{landing_page_options}}
Return three to five ad groups with:
1. Search intent
2. Keyword themes
3. Match-type guidance
4. Negative-keyword themes
5. Counter-positioning angle
6. Claim and proof
7. CTA
8. Required landing-page section
9. Disqualifying traffic
10. Compliance or review risk
11. Launch priority
Do not recommend a claim without approved proof. Do not mix competitor, alternative, category, and problem intent in one group. Mark decisions that require paid media, legal, or product approval.Write responsive search ad assets for these approved ad groups.
Ad-group plans:
{{ad_group_plans}}
Approved claims and proof:
{{approved_claims_and_proof}}
Crowded phrases to avoid:
{{phrases_to_avoid}}
Brand and legal rules:
{{brand_and_legal_rules}}
For each ad group, create:
1. Twelve headlines, target 30 characters
2. Six descriptions, target 90 characters
3. Four sitelink ideas
4. Six callout ideas
5. Asset purpose: relevance, proof, pain, comparison, or CTA
6. Pinning caution
7. Claim requiring human verification
Do not invent metrics, guarantees, pricing, or competitor facts. Keep the search intent clear. Flag any asset that may exceed limits or create a misleading combination. A paid media manager will verify counts and platform policy before upload.Audit ad-to-landing-page alignment.
Ad-group plans and assets:
{{ad_groups_and_assets}}
Landing-page copy:
{{landing_page_copy}}
Approved proof:
{{approved_proof}}
For each ad group, return:
1. Search intent
2. Ad promise
3. Matching page headline or section
4. Proof present or missing
5. Objection answered or missed
6. CTA alignment
7. Contradictory wording
8. Mobile above-the-fold requirement
9. Exact page change needed
10. Launch status: ready, revise, or blocked
Do not assume an unshown page element exists. Do not suggest proof we do not have. A landing-page owner and paid media manager will approve all fixes before launch.Create a launch QA and learning plan for this counter-campaign.
Approved campaign plan:
{{campaign_plan}}
Final ad assets:
{{ad_assets}}
Landing-page audit:
{{landing_page_audit}}
Budget and conversion context:
{{budget_and_conversion_context}}
Return:
1. Pre-launch QA checklist
2. Tracking and conversion checks
3. Budget and brand-safety guardrails
4. First search-term review trigger
5. Negative-keyword review process
6. Early quality signals
7. Outcome metrics that need more volume
8. Pause, revise, and expand thresholds
9. Owner and review date for each check
10. Competitor-monitoring cadence
Do not invent benchmark thresholds. Use placeholders where the business must set a value. A paid media owner will approve the final controls before launch.Related workflows
Continue with workflows that share a similar GTM motion, category, or tool stack.
Create 50 ready-to-test ad concepts across LinkedIn, Google, and Meta from one campaign brief.
Build a competitor messaging map with repeated claims, proof gaps, overused language, and sharper positioning opportunities for your brand.
A keyword-to-ad-to-landing-page audit system that identifies message gaps and generates fixes for Google Ads quality score improvement.