# What Tools Are Best for Analysing Competitor Ads and Keywords? A Stack for 2026 > The best free and paid tools for analysing competitor ads and keywords, what each can and can't tell you, and a 90-minute monthly routine that turns rivals' ads into test ideas. - **Author**: Marxx - **Published**: 2026-10-24 - **Category**: Creative strategy - **URL**: https://marxx.ai/posts/competitor-ads-keywords-tools --- **What tools are best for analysing competitor ads and keywords?** For keywords, use Google Keyword Planner (free), or Semrush or Ahrefs for depth. For search ads, Google's Ads Transparency Center (free), or SpyFu and Semrush. For social ads, the Meta Ad Library (free). To see the pattern across all of them, with days live, copy counts and tags for named rivals, use a tracker like Marxx. ## Your competitor's customer doesn't split search and social. Your team does. In most marketing teams, competitor research is done by two groups that rarely compare notes. The search team looks at keywords and search ads. The social team looks at feed ads. Each produces its own report. But buyers don't move that way. Google's research on what it calls the **"messy middle"**, published in 2020, describes buyers looping repeatedly between two modes. They **explore**, widening their options: searching, browsing, seeing ads. Then they **evaluate**, narrowing them: comparing offers, reading reviews. They go round as many times as they need before they buy ([Google, 2020](https://business.google.com/in/think/consumer-insights/navigating-purchase-behavior-and-decision-making/?hl=en-US)). Buyers loop between exploring and evaluating, from trigger to purchase, and a competitor can win them at any turn, in search or in the feed A competitor can win your buyer at any turn of that loop: with a search ad, a feed ad, or an offer that makes evaluation easy. So the useful question isn't "which tool is best for keywords?" or "which is best for ads?" It's "**which tools let me see the whole loop, and what can I honestly conclude from them?**" This guide covers: 1. What each type of competitor data can and can't tell you. 2. One competitor followed from search to feed, with the tool that sees each step. 3. The stack, free and paid, step by step. 4. A way to turn the findings into ideas: mapping rivals' offers to the biases that drive decisions. 5. A 90-minute monthly routine, and the mistakes to avoid. Follow one competitor from search to feed: the search, the search ad, the feed ad, the pattern, with the tool that sees each step --- ## Part 1: Competitive analysis is inference You can never see a competitor's results: their sales, profit or return on ad spend. Everything you learn is **inference**, reasoning from what's visible to what isn't. Some signals support that reasoning much better than others. What each signal can tell you: keyword volume is rough; search ads and live feed ads are seen; days live plus copies is strong evidence; their results are never visible | Signal | What it tells you | How much to trust it | |---|---|---| | **Keyword volume** | Roughly how many people search for something | **Rough.** Estimates vary widely between tools | | **Search ads** | What a competitor bids on and claims | **Seen.** You know it ran; not whether it worked | | **Live feed ads** | What they're testing now | **Seen.** Many are short tests | | **Days live and copy counts** | What they keep paying for | **Strong.** Repeated spending reveals conviction | | **Their results** | Sales, profit, ROAS | **Never visible.** Don't pretend otherwise | ### Why keyword volume is rougher than it looks Keyword tools estimate search volume in different ways, and the estimates can disagree enormously. In a November 2024 comparison, the keyword "reddit marketing" showed **880,000** monthly US searches in Google Ads, **590,000** in Semrush and **100,000** in Ahrefs ([Practical Ecommerce, 2024](https://www.practicalecommerce.com/keyword-volume-google-vs-semrush-vs-ahrefs)). The same keyword, three answers: Google Ads 880,000, Semrush 590,000, Ahrefs 100,000 monthly US searches **Why it happens:** Each tool uses different data and methods. Google's Keyword Planner reports a 12-month average, so a newly trending term can look small. It has also shown only volume *ranges* (like 1K-10K) to accounts with low ad spend ([Search Engine Land](https://searchengineland.com/google-officially-throttling-keyword-planner-data-low-spending-adwords-accounts-255795)). Third-party tools model volume from a mix of sources. **How to use it well:** Treat volume as a way to **compare and sort** keywords, not to forecast clicks. "This term is about ten times bigger than that one" is a reasonable conclusion. "We'll get 8,800 clicks" isn't. > **The quotable version:** Keyword volume tells you which searches are bigger. It doesn't tell you how big. --- ## Part 2: Follow one competitor from search to feed To show how the steps connect, we followed Wonderla, an Indian amusement park chain, using Marxx data from October 2026. ### Step 1: The search **The question:** Which searches exist in this category, and which are biggest? **Tools:** Google Keyword Planner (free with a Google Ads account), Semrush, Ahrefs. **What you learn:** The search terms people use, and their size relative to each other. **What you miss:** Everything about the ads. ### Step 2: The search ad **The question:** Who bids on those searches, and what do they promise? **Tools:** Google Ads Transparency Center (free; search any advertiser's ads by region and format, with the date each was last shown), SpyFu and Semrush (paid; estimate which keywords competitors buy). **What you learn:** Which brands show up and the claims they make in search. **What you miss:** What they run on Meta, which is often where most creative testing happens. ### Step 3: The feed ad **The question:** What is the competitor running in social feeds right now? **Tools:** Meta Ad Library (free), TikTok Creative Center (category trends, not named competitors). **What you learn:** On 16 October 2026, Wonderla had **more than 500 ads** live on Meta. The free library shows each one with its start date. **What you miss:** The pattern. Five hundred ads is a list, not an insight. ### Step 4: The pattern **The question:** Of everything they run, what are they actually backing? **Tools:** A competitor tracker that groups near-identical ads and counts how long they've run. In Marxx, Wonderla's 500+ ads collapse into **four creatives**, all selling the same "Buy 3, Get 1 Free" offer, each with **35 to 44 near-identical copies**, all live since 9 July. **What you learn:** This isn't 500 ideas. It's one offer, backed hard for months. That's the strongest public evidence of what a competitor believes works. Marxx creative board grouping competitor ads by hook, angle and format **What you still can't conclude:** That the offer is profitable. A brand can back an offer for brand reasons, seasonal reasons, or simply inertia. Strong evidence isn't proof. --- ## Part 3: The stack, step by step Which tool sees which step: keywords, search ads, social ads, patterns | Step | Free | Paid | Watch out for | |---|---|---|---| | **Keywords** | Google Keyword Planner | Semrush, Ahrefs | Volumes are estimates; use them to compare, not forecast | | **Search ads** | Google Ads Transparency Center | SpyFu, Semrush | Paid tools estimate keywords bought; they don't see bids | | **Social ads** | Meta Ad Library, TikTok Creative Center | Ad spy tools | The free Meta library drops most stopped commercial ads outside the EU | | **The pattern across all of it** | Spreadsheets and time | Marxx (Meta, Google and keywords for named rivals, with days live, copies and tags) | Patterns are evidence of belief, not results | *Based on each tool's public features, October 2026.* **How to choose:** - **One or two competitors, checked monthly?** The free tools are enough if you take careful notes. - **Five or more competitors, or weekly checks?** The time cost of the free route adds up. A tracker that keeps history and groups copies pays for itself. - **Deep SEO work?** Pair any ad tracker with a dedicated SEO tool. They answer different questions. --- ## Part 4: From findings to ideas: Map their offers to the six biases Collecting competitor ads is the easy part. Turning them into a decision is harder. One useful lens comes from the same Google research: it identified **six cognitive biases** that shape choices in the messy middle. They are **category heuristics** (shortcuts like price or key specs), the **power of now**, **social proof**, **scarcity**, **authority**, and the **power of free** ([Google, 2020](https://business.google.com/in/think/consumer-insights/navigating-purchase-behavior-and-decision-making/?hl=en-US)). **The method:** For each competitor ad that's lasted, note which biases it uses. Then look for the columns nobody fills. Map rivals' offers to the six biases: Wonderla and SHOTT use the power of free; Smaaash's bowling at INR 9 uses the power of now and a price heuristic; Make Monday your Sunday uses the power of now; scarcity, social proof and authority are unused **Worked example (real ads, our reading), entertainment category in India, October 2026:** | Ad | Live | Biases used (our reading) | |---|---|---| | Wonderla: "Buy 3, Get 1 Free" | 99 days, 35-44 copies | Power of free | | SHOTT: "Hit a strike, grab a drink" (free drink with a game) | 48 days, 6 copies | Power of free | | Smaaash: "Bowling at INR 9" on a Thursday | 36 days | Power of now; a price shortcut | | Smaaash: "Make Monday your Sunday" | 37 days | Power of now | Real competitor ads with days live and copies: Smaaash 'Bowling at INR 9' 36 days, Smaaash 'Make Monday your Sunday' 37 days, SHOTT 'Hit a strike, grab a drink' 48 days with 6 copies, SHOTT 'Make it epic' 37 days **What the map suggests:** In this sample, rivals lean on **free** and **now**. **Social proof**, **scarcity** and **authority** go unused. A venue could test "Rated 4.8 by 2,000 families" (social proof) or "Only 20 weekend party slots left" (scarcity, if it's true). **Where this lens misleads:** It's a sample of four ads, and assigning biases is a judgement call. An empty column might be empty because it doesn't work in the category, not because nobody's tried it. Treat gaps as **hypotheses to test**, with a small budget, not as conclusions. --- ## Part 5: The 90-minute monthly routine The monthly competitor routine: 15 minutes keywords, 15 search ads, 30 feed ads, 15 bias map, 15 write one sentence per rival 1. **Keywords (15 minutes):** What's rising in your category? Sort, don't forecast. 2. **Search ads (15 minutes):** Who bids on your top terms, and what do they promise? 3. **Feed ads (30 minutes):** For each rival, find the longest-running ads, the most-copied ones, and anything that stopped. 4. **Bias map (15 minutes):** Which levers do they pull? Which does nobody pull? 5. **Synthesis (15 minutes):** One sentence per rival, and one test for you. ### The one-sentence test The one-sentence synthesis: they bid on a keyword, promise a claim in search, have backed a hook and offer on Meta for N days with N copies, and nobody in the category uses a given lever If you can complete this sentence for a competitor, you have analysis rather than a pile of screenshots: > "They bid on **[keyword]**, promise **[claim]** in search, and have backed **[hook and offer]** on Meta for **[N] days** with **[N] copies**. Nobody in the category uses **[lever]**." **Why it works:** The sentence forces you to connect search and social, to use the strongest signal (days and copies), and to end on a gap you can act on. Marxx trending tactics showing hooks and formats rising and declining ### How to measure whether your competitor research is working | Metric that matters | Why | |---|---| | Tests per month that came from a competitor insight | Research that never becomes a test is trivia | | Win rate of those tests vs your other tests | Shows whether the insights are any good | | Time from a rival's launch to your noticing it | Faster awareness means faster response | | Metric that misleads | Why | |---|---| | Number of competitor ads saved | Collection isn't analysis | | Estimated competitor spend | Usually a guess; Meta and Google don't publish it for most ads | | Keyword volume totals | Estimates that can differ several-fold between tools | --- ## Six mistakes, and how to spot them | Mistake | Symptom | Fix | |---|---|---| | Separate search and social reports | Nobody connects the two | One sentence per rival, across both | | Forecasting from keyword volume | Click estimates that never materialise | Use volume to rank, not to predict | | Treating every live ad as a winner | Copying short tests | Weight by days live and copies | | Concluding a stopped ad failed | Avoiding ideas that may have worked | "Stopped" is a question, not a verdict | | Copying the ad | Me-too creative | Copy the lever; find an unused one | | Never turning insight into a test | Beautiful decks, same results | End every review with one test | ## What to take into your next planning meeting - **Buyers loop between search and social; your research should too.** Connect them in one sentence per rival. - **Keyword volume ranks; it doesn't forecast.** One keyword showed 880,000, 590,000 and 100,000 searches in three tools. - **Days live and copies are the strongest public signal.** Wonderla's 500+ ads were really four creatives and one offer. - **Map offers to the six biases.** The empty columns are your test ideas. - **You never see their results.** Good analysis says what's evidence and what's a guess. ## FAQ **What's the best tool for competitor keyword research?** Semrush or Ahrefs for depth, and Google Keyword Planner for free data. Use volumes to compare keywords, not to forecast clicks. **What's the best tool for analysing competitor ads?** The Meta Ad Library and Google's Ads Transparency Center are free. For days live, copy counts, history and tags across Meta and Google, use a tracker like Marxx. **Can one tool do both ads and keywords?** Marxx covers Meta ads, Google ads and keywords for named competitors. For deep SEO work, pair it with a dedicated SEO tool. **How accurate is competitor keyword data?** Rough. Different tools can disagree several-fold on the same keyword. Use it for ranking and trends. **Can I see how much a competitor spends on ads?** Not reliably. Meta and Google don't publish spend for most commercial ads, so any figure is an estimate. --- ## See the whole loop for your top rivals Book a demo and we'll track your top five rivals across Meta, Google and keywords, then walk through what they're backing, live. **[Book a demo: Track your top 5 rivals ->](https://www.marxx.ai)** *Related: [PPC Competitive Analysis Template](/blog/ppc-competitive-analysis-template) - [What Features Should a Reliable Ad Spy Tool Have?](/blog/ad-spy-tool-checklist) - [Free Ad Spy Tools](/blog/free-ad-spy-tools)* ### Sources - Google (2020). [Decoding decisions: Navigating purchase behaviour and decision-making](https://business.google.com/in/think/consumer-insights/navigating-purchase-behavior-and-decision-making/?hl=en-US) (the messy middle and six biases). - Practical Ecommerce (2024). [Keyword volume: Google vs Semrush vs Ahrefs](https://www.practicalecommerce.com/keyword-volume-google-vs-semrush-vs-ahrefs). - Search Engine Land. [Google officially throttling Keyword Planner data for low-spending accounts](https://searchengineland.com/google-officially-throttling-keyword-planner-data-low-spending-adwords-accounts-255795). - Marxx ad library, 16-25 October 2026. Days live and copy counts were accurate at the time of writing. --- - [More Creative strategy articles](https://marxx.ai/posts/category/creative-strategy) - [All articles](https://marxx.ai/posts)