How to do a marketing competitor analysis with ChatGPT or Claude (free template)

Collect each competitor's ads, funnel, content, reviews and search traffic from public sources, then have ChatGPT or Claude turn your notes into a teardown for each one. You finish with a list of gaps you can win and 5 first moves, ranked by impact and effort.

Christian MartinFrom the playbook Deconstruct my competitors' growth by Christian Martin ✓ · Updated Nov 19, 2026

What you need

  • Your own offer in two or three sentences: what you sell, to whom, at what price, and how customers find you now
  • Names and website addresses for 3 to 5 direct competitors, plus 1 or 2 bigger players your buyers also look at
  • A free Facebook login to use the Meta Ad Library (facebook.com/ads/library)
  • Access to the Google Ads Transparency Center (adstransparency.google.com), which shows Google and YouTube ads
  • A throwaway email address to sign up for competitors' lead magnets and email lists
  • Optional: a free account on an SEO tool such as Semrush, Ahrefs Webmaster Tools or Ubersuggest to see competitors' top pages and keywords
  • A doc or spreadsheet with one tab per competitor where you paste what you collect
  • ChatGPT or Claude (paid plans handle long pasted notes and screenshots better)

The steps

  1. List the competitors

    Write down 3 to 5 direct competitors (same buyer, same problem, similar price) and 1 or 2 bigger players your buyer also compares you with. If you aren't sure, give ChatGPT or Claude your offer and ask: "Who would my buyer compare me with before buying? Split them into direct competitors and bigger alternatives, including non-obvious ones like DIY tools or agencies." Check every name it suggests by visiting the site. AI often lists companies that have closed or that sell to a different buyer. Start one tab per competitor in your doc.

  2. Read their ads

    Search each competitor in the Meta Ad Library (set the country to yours and the category to All ads) and in the Google Ads Transparency Center for YouTube and search ads. For each ad, note the hook (first line or first 3 seconds), the offer, the call to action, the format and the "Started running on" date. Ads that have run for 60 days or more are probably profitable, since nobody keeps paying for a losing ad. Take screenshots of the 5 longest-running ads. Paste the notes or screenshots into the AI and ask: "Group these ads by angle and offer. Which angles have run longest, and what promise do they all repeat?" Compare its answer with the dates you saw yourself.

  3. Walk their funnel

    Go through each funnel the way a buyer would. Click an ad, read the landing page, take the lead magnet with your throwaway email, book or start a trial if it's free, and go as far into checkout as you can without paying. Note every step, the main promise, the price, any upsells or order bumps, and the guarantee. Forward their first 5 to 7 emails into your doc. Ask the AI: "Lay out this funnel step by step: what each step asks for, what it promises and what happens next. Where does it lose people or ask too much too early?"

  4. Read their content

    Open their main social channels and YouTube. Sort by popular, or scroll back 90 days, and note the top 10 posts or videos: topic, format (short video, carousel, long video, newsletter), and how often they post. Paste the list in and ask: "Which topics and formats get the most engagement for this competitor, and how often do they post? What topics do all the competitors cover, and which do none of them cover?" The topics nobody covers are worth keeping.

  5. Read their reviews

    Collect 20 to 50 reviews per competitor from wherever their buyers leave them: Google, Trustpilot, G2, Capterra, Amazon, app stores, Reddit threads. Make sure you include the 2 and 3-star ones, because those are the most specific. Paste them in and ask: "Sort these into what customers praise and what they complain about. Give the top 3 of each with counts and one exact quote each." Then check that the quotes really appear in the text you pasted. Each complaint you find is an opening for you.

  6. Check their search

    Enter each competitor's domain in a free SEO tool and export or copy their top pages and top keywords by estimated traffic. Without a tool, search Google for site:theirdomain.com plus your main topic and look at what ranks. Ask the AI: "From these pages and keywords, what buyer questions does this competitor win in search, and which offers do those pages push?" Treat traffic numbers as rough estimates. They're only useful for comparing competitors with each other.

  7. Map the engine

    Paste all your notes for one competitor and ask for exactly one line in this form: Attention (where buyers first see them) → Lead (how they capture contact or intent) → Close (how the sale happens, at what price). Repeat for each competitor. Check every part of the line against something you actually saw. If the AI writes "webinars" and you never found one, take it out.

  8. Find the gaps

    Put all the engine lines, complaint lists and content notes into one message and ask: "Across these competitors, list (1) buyers they ignore, (2) promises they make that reviews say they don't keep, (3) channels or formats none of them use well, (4) price points or offers nobody covers. For each gap, cite the evidence from my notes." Delete any gap with no evidence behind it.

  9. Make the plan

    Give the AI your own offer, budget, team size and current channels along with the gap list. Ask: "Pick 5 moves to make first. Score each 1 to 5 on likely impact and 1 to 5 on effort for a business like mine, rank them by impact divided by effort, and give the first concrete action for each this week." Change any effort score that doesn't fit your real capacity. Save the whole document and run it again next quarter or before a launch, so you can see what competitors started and stopped.

A prompt to start with

You are a growth strategist doing a marketing competitor analysis for me. Use only the evidence I paste. If something isn't in my notes, write "no data" instead of guessing.

My business: [what you sell, to whom, price, how customers find you now]
My constraints: [monthly marketing budget, team size, channels you already use]

Competitors:
1. [Name, website, direct or bigger player]
2. [Name, website, direct or bigger player]
3. [Name, website, direct or bigger player]

For each competitor, here are my notes:
- Ads (hook, offer, format, start date): [paste]
- Funnel steps (landing page, lead magnet, emails, checkout, upsells, price): [paste]
- Top content, last 90 days (topic, format, posting rate): [paste]
- Reviews (raw text, including 2 and 3 stars): [paste]
- Search (top pages and keywords): [paste]

Produce:
1. A growth teardown for each competitor: their active ads and longest-running angles, their funnel step by step with the main promise, content patterns, top 3 praises and top 3 complaints with one real quote each, and their search strengths.
2. One engine line per competitor: Attention → Lead → Close.
3. A gap list: ignored buyers, broken promises, open channels, uncovered offers or prices. Cite evidence for each.
4. 5 moves for me to make first, scored for impact (1-5) and effort (1-5), ranked, each with a first action I can take this week.

Ask me questions first if any section is missing evidence.

Paste it into ChatGPT or Claude and put your own facts in the [brackets].

An example of the finished work

Competitor teardown for Ledgerly (made-up business): monthly bookkeeping for Shopify stores doing $20k to $200k a month, $299/month. Run in Q3, before a price change.

COMPETITORS
Direct: ShelfBooks, CartLedger, Tallyroom. Bigger player: Northbook (general small-business bookkeeping, used by many store owners).

ENGINE LINES
- ShelfBooks: Meta video ads with founder-to-camera "Your Shopify payouts don't match your bank" (running 140 days) → free "Payout Reconciliation Checklist" + 6-email sequence → 15-min sales call, $349/month, 3-month minimum.
- CartLedger: SEO ("shopify sales tax by state" pages bring most of their traffic) → free sales-tax calculator, email required → self-serve checkout, $199/month, setup-fee upsell at $450.
- Tallyroom: Partnerships with two Shopify agencies + YouTube tutorials (2 per week) → webinar "Close your books in 5 days" → call, $399/month.
- Northbook: Broad YouTube and search ads to all small businesses → free trial → software plus an add-on bookkeeper, $250+/month. Nothing specific to e-commerce.

WHAT CUSTOMERS SAY
- ShelfBooks: praised for a fast first cleanup (14 reviews); complained about the 3-month lock-in (9) and "different bookkeeper every month" (7).
- CartLedger: praised for price (21); complained that "inventory/COGS is always wrong" (12), "no one to talk to" (8).
- Tallyroom: praised for its tutorials; complained "books arrive on the 20th" (6) despite the 5-day promise.
- Northbook: "they don't understand Shopify payouts or refunds" (15).

GAPS
1. Inventory and COGS accuracy. CartLedger buyers complain about it and nobody advertises it.
2. Same named bookkeeper every month. This is ShelfBooks' second-biggest complaint.
3. A real books-closed-by-day-5 guarantee. Tallyroom promises it and doesn't deliver.
4. TikTok and short-form video. None of the four post there.
5. Stores just over $200k/month. Every competitor's offer stops at that size.

FIRST 5 MOVES (impact / effort)
1. Run a Meta ad copying ShelfBooks' proven payout angle, adding "same bookkeeper, no lock-in" (5/2). This week: film 3 founder-to-camera hooks.
2. Publish a "Books closed by day 5 or that month is free" guarantee (5/2). This week: check last quarter's close dates, then add the guarantee to the pricing page.
3. Build a free "Shopify COGS checker" lead magnet (4/3). This week: write the 10-question spreadsheet.
4. Post 3 short videos a week on payout and COGS mistakes (3/2). This week: script 6 from the review complaints.
5. Test a $599 tier for $200k+ stores (4/4). This week: email 10 current clients near that size.

Next review: start of Q4. Check whether ShelfBooks' 140-day ad is still running.

The business in this example is made up.

Or hand the whole job to the playbook

Christian Martin wrote these steps as a playbook on Iggy. It runs in your own ChatGPT or Claude, and Iggy gives it the facts of your business, so you do not explain them each time. In the free webinar, we set Iggy up together, in real time.

Mistakes to avoid

  • Copying every ad a competitor is running right now.Look at the "Started running on" date in the Meta Ad Library and study the ads that have run 60 days or more. New ads are often just tests that may be turned off next week.
  • Asking ChatGPT or Claude to "analyze my competitors" with only their names.Collect the ads, funnel steps, reviews and pages yourself and paste them in. With nothing but a name, the AI fills the gaps with plausible guesses that read like facts.
  • Reading only the homepage and pricing page.Walk the whole funnel the way a buyer would. Take the lead magnet with a throwaway email, read their first 5 to 7 emails and go as far into checkout as you can. Upsells, guarantees and the real offer usually show up there.
  • Skimming 5-star and 1-star reviews only.Read the 2 and 3-star reviews. They name the specific promise that wasn't kept, and that's the gap you can claim.
  • Ending with a long list of observations and no plan.Make every analysis end in 5 ranked moves, each with a first action for this week, scored against your real budget and team.
  • Doing it once and never again.Save the document and redo it each quarter or before a launch. The useful signal is the change: which ads stopped, which offers changed, which prices moved.

Questions and answers

What should a marketing competitor analysis template include?

One section per competitor covering their active and longest-running ads, their funnel steps and price, top content from the last 90 days, what reviews praise and complain about, and the pages and keywords that bring them search traffic. Then a one-line engine for each competitor (Attention → Lead → Close), a gap list, and 5 ranked moves. The example on this page uses that layout. You can copy it into a doc, Excel sheet or PowerPoint deck.

How many competitors should I analyze?

3 to 5 direct competitors plus 1 or 2 bigger players your buyers also look at. More than that and you skim, and skimming won't turn up the specific gaps.

Can ChatGPT or Claude look up my competitors' ads by itself?

Not reliably. The Meta Ad Library and similar tools need searching, scrolling and dates the AI often can't get to or read correctly. Collect the ads yourself with screenshots or notes, then let the AI group and compare them.

How do I know which competitor ads are actually working?

Check how long they've run. An ad that's still active after 60 days or more is very likely making money, because businesses turn off ads that lose money. Also watch for the same hook used across several versions. That usually means it won a test.

Do I need paid SEO tools for this?

No. Free versions of Semrush, Ubersuggest or Ahrefs Webmaster Tools show a competitor's top pages and keywords well enough to compare them. A Google search of site:theirdomain.com plus your topic fills in the rest. Treat traffic numbers as estimates.

How long does this take?

Plan on 1 to 2 hours per competitor to collect ads, the funnel, content, reviews and search data. The AI part, turning your notes into teardowns, gaps and a plan, takes about 30 minutes.

How often should I redo it?

Once a quarter, and before any launch or price change. Compare it with the last run to see what competitors started, stopped or changed.

More how-to guides