Your competitors' Google ads show you which searches they pay to appear on, which products they push, which offers they lead with and where they send the click. You can't see their budgets or their results, but you can see what they keep running.
That last part matters most. Advertisers rarely keep paying for an ad that loses money month after month, so an ad that stays live for a long time is a useful signal. A short-lived ad is usually a test that didn't make it.
Below is an audit you can run in an afternoon: what to look at, the steps in order, a scorecard to compare brands side by side, and a clear line between what to borrow and what to leave alone.
What a competitor's ads can and can't tell you
Start with honest expectations. From the outside you can see the ad creative, the offer, the page behind the ad, the network it runs on, how long it has been live and how many versions exist.
You can't see what they pay per click, their conversion rate, their daily budget or their margin. An ad that runs for months might be profitable, or it might be a brand-protection ad that nobody has looked at in a while. Treat everything you find as a hypothesis to test, not a result to copy.
The five things to look at
Search ads versus Shopping ads
Search ads are the text ads above the results for a query. They tell you the intent a brand pays for: its own brand name, a competitor's name, a product category, or a problem the product solves.
Shopping ads are the product listings with an image, a price and a store name. They tell you which products the brand pushes hardest and at what price points. A competitor that runs Shopping ads on only a handful of products is showing you its best sellers.
Note where else the brand appears too. Some run video on YouTube or image ads across the display network, which usually means a bigger budget and a longer buying cycle.
Headlines and offers
For each ad, write down three things in plain words: the main promise, the proof, and the offer. The promise is the benefit ("keeps coffee hot for 12 hours").
The proof is what makes it believable (reviews, a guarantee, a number of customers). The offer is the reason to act now (free shipping, a bundle price, a discount code).
After ten or fifteen ads you will see patterns. Some brands always lead with price, others with quality or speed. The pattern tells you how they see their own position.
Landing pages
Click through and look at the landing page on your phone. Is it a product page, a collection, a dedicated campaign page or the homepage?
Does the headline on the page repeat the promise of the ad? Is the offer visible without scrolling?
A brand that builds dedicated pages for its ads is usually testing seriously. A brand that sends every ad to the homepage is often leaving money on the table, which is an opening for you.
How long ads run
Run time is the closest thing you have to a performance signal. Group each brand's ads into rough buckets: under two weeks, two weeks to two months, and longer than two months.
Ads in the last bucket deserve the most attention. The exception is ads on the brand's own name, which tend to run all year whatever they return, so don't read too much into those.
How many variants
Count how many versions of the same message a brand runs. Many close variants of one idea usually means active A/B testing: the brand believes in the angle and is searching for the best wording or image. A single ad that has run unchanged for months is either a stable winner or an account nobody touches.
A step-by-step audit
- Pick three to five competitors. Choose brands that sell to the same customer at a similar price, not just the biggest names in your category.
- Collect their ads. Google's Ads Transparency Center lets anyone look up an advertiser's ads. You can also search your main keywords yourself, keeping in mind that results change with location and history.
- Sort by run time. Put the longest-running ads at the top of your list, and set aside ads on the brand's own name.
- Record each ad in one sheet. Use the same columns for every ad: network, main promise, proof, offer, landing page type, days running and number of variants.
- Open every landing page on a phone. Note load speed, whether the offer is repeated and how many taps it takes to reach checkout.
- Score each competitor. Use the scorecard below so you compare brands on the same terms.
- Write three hypotheses. Turn what you found into tests for your own account, for example "a bundle offer in the headline will lift click-through rate on our category keywords".
A simple scorecard
Score each competitor from 0 to 2 on six criteria, where 0 means weak or missing, 1 means present and 2 means clearly strong. The maximum is 12.
| Criterion | What to check | Score |
|---|---|---|
| Offer clarity | Can you tell what you get and why to act now from the ad alone? | 0 to 2 |
| Proof | Reviews, guarantees, numbers or awards in the ad or right below it | 0 to 2 |
| Message match | The landing page repeats the promise and offer of the ad | 0 to 2 |
| Longevity | Several non-brand ads live for more than two months | 0 to 2 |
| Testing activity | Close variants of the same angle running at the same time | 0 to 2 |
| Network coverage | Present on Search and Shopping, and possibly YouTube | 0 to 2 |
As an example, say competitor A scores 2, 1, 2, 2, 1 and 1: that is 9 out of 12. Competitor B scores 1, 0, 1, 2, 0 and 2: 6 out of 12.
A has done more of the work, so its ads are the better source of ideas. B's low scores on proof and testing show where a sharper offer could beat it on the same searches.
What to copy and what not to
Borrow the thinking, not the words. These are worth adapting:
- Angles that last. If three competitors keep running ads on the same benefit, shoppers respond to it. Find your own way to say it.
- Offer structures. Bundles, free shipping thresholds and guarantees can be adapted to your products and margins.
- Proof types. If reviews or guarantees show up in every long-running ad, add the proof you have.
- Landing page patterns. A dedicated page per ad group, the offer above the fold, shipping details near the button.
- Product focus. The products a competitor pushes in Shopping ads point to where demand is.
And these are worth leaving alone:
- Exact wording. Copied headlines make you look like a second choice, and they rarely fit your brand.
- Their brand name in your ad text. It can run into Google's trademark rules and it confuses shoppers.
- Their prices and discounts. Their costs are not yours. A discount that works for them may push you below break-even.
- Short-lived ads. An ad that ran for a week is a test that most likely failed.
Before you bid on the same keywords, work out what you can afford to pay for a sale. The break-even ROAS calculator gives the most you can spend per order. Compare it with what the keyword is likely to cost: if a typical cost per click means you need an unrealistic conversion rate to stay above break-even, compete with a different offer or on different searches.
How to do this with SellWinners
The Ads Library includes a Google ads library next to the Meta and TikTok ones, so you can run the whole audit in one place.
- Open the Google tab. Go to Ads in the app and choose Google. Search a product, a niche or a competitor's name.
- Keep the ads that matter. Set Status to active and a minimum for Days running, for example 60, to keep the long runners. With a text search, Google results are ranked by relevance, so the days running filter is the way to keep long-running ads.
- Split by network. Use the Networks filter to look at Search, Shopping and YouTube ads separately, and Countries to stay in the markets you sell to.
- Browse without a search. Leave the search empty, pick a category, and sort by longest running or by reach to see which ads carry a whole category.
- Read each ad. Every result shows the creative, the advertiser and its domain, the network, the days running, the reach and the countries.
- Look behind the ad. Open the advertiser's domain in Shop Analytics to see its traffic, best sellers and estimated sales, and use Similar Shops to find competitors you missed.
- Keep your findings. Save the ads you want to discuss to a board in Folders (Boards in the app) and share it with your team or a client with one link.
The same search works from your code through the REST API (see the API docs) and from an AI agent over MCP, so you can ask an agent to list active Shopping ads for a product that have run for two months. To run the same audit on Facebook and Instagram, read how to research competitors' Meta ads. Google ads are on paid plans, and the Free plan includes the web app and the MCP server (150 credits a month), without the API; compare them on the pricing page.
Key takeaways
- Look at five things: network, headline and offer, landing page, run time and variants.
- Long-running non-brand ads are your best signal. Ads on a brand's own name are not.
- Score competitors on the same six criteria and look for their gaps, not just their totals.
- Borrow angles, offer structures and page patterns. Never copy wording, prices or trademarks.
- Check your break-even numbers before you bid on the same keywords.