How to Spy on Competitors’ Google Ads (Legally)

Yes, It’s Legal — and Your Competitors Are Already Doing It to You

Let’s clear the air on the word “spy.” Everything in this article uses data that Google publishes deliberately, tools that aggregate public search results, and searches anyone can run from a phone. There is no scraping of private dashboards, no fake-account trickery, no grey area. Google built the Auction Insights report specifically so advertisers could see who they compete against. It built the Ads Transparency Center specifically so anyone could look up any advertiser’s ads. Using these is not espionage — it’s reading the instruments the platform hands you.

The reason to bother is structural: Google Ads is an auction, and in an auction your results are never just about you. Your cost per click, your impression share, and your conversion volume are all functions of who else is bidding, how hard, and with what message. You can run a technically flawless account and watch performance erode for months because a funded competitor entered your market — and if you’re not looking at competitive data, the decline looks like a mystery instead of a cause with a name.

The inverse is also true. Competitors leak information constantly: which keywords they think are worth money, which offers they lead with, which landing pages they trust enough to send paid traffic to, when their budgets run thin. None of it is hidden. Most advertisers simply never assemble it.

What follows is the assembly process: the three native Google sources worth reading, where third-party tools genuinely add value, and how to turn the resulting picture into bid, copy, and budget decisions instead of a folder of screenshots nobody opens twice.

Auction Insights: The Only Report Built From Real Auction Data

Start with Auction Insights, because it’s the one source built from actual auctions you participated in — not estimates, not panels, not crawls. You’ll find it inside Google Ads at the campaign, ad group, or keyword level, and it shows every advertiser who entered the same auctions you did, whether or not you ever noticed them.

Five metrics matter. Impression share is the percentage of eligible auctions where an advertiser’s ad actually showed — your own and each competitor’s. Overlap rate tells you how often a competitor’s ad appeared in the same auctions as yours; it’s your proximity measure, and the advertisers at the top of that column are your real rivals regardless of who you think your competitors are. Outranking share shows how often your ad ranked above theirs or showed when theirs didn’t. Position above rate is the reverse angle: when you both showed, how often they beat you. Top of page rate tells you who is consistently buying the premium real estate.

The single most useful way to read this report is over time, not as a snapshot. Segment by month and watch the movement. A competitor whose impression share climbs steadily quarter over quarter is investing; one who appears suddenly at 40% overlap is a new entrant worth a full workup; one who fades in December and reappears in March is telling you their budget calendar. Rising CPCs almost always have a face in this report.

Two honest limitations. First, you only see advertisers in auctions you entered — Auction Insights cannot show you keywords you’re not bidding on. Second, it shows presence, not performance: a competitor with 80% impression share might be printing money or burning it. For everything beyond presence, you need the next two sources.

The Ads Transparency Center: Read Every Ad They’re Running

Auction Insights tells you who; the Google Ads Transparency Center tells you what. It’s a public site — no Google Ads account required — where you can search any advertiser by name or website and browse the ads they’ve run across Search, YouTube, Display, and Shopping, with date ranges and format filters. Google launched it for accountability reasons, but for a practitioner it’s a free, always-current swipe file of your competitive set.

Look up each of your top Auction Insights overlaps and read their search ads as a body of work, not one at a time. Patterns are the prize. What claim leads every headline — price, speed, credentials, guarantee? Which offers repeat across many ads (those are the ones surviving their testing), and which appeared briefly and vanished (those likely lost)? Are they running seasonal promotions on a calendar you can anticipate? Do their video and display ads tell the same story as their search ads, or is search an afterthought?

The date filtering is quietly the most strategic feature. An advertiser whose ad library shows constant fresh variants is actively testing and probably actively managed; a library where the same three ads have run unchanged for a year suggests an account on autopilot — which is competitive information in itself. A sudden burst of new creative often precedes a push into new keywords or markets.

What the Transparency Center won’t give you: spend, impressions for most ad types, keyword targeting, or landing pages beyond what the ad itself reveals. It’s the message layer only. But message is where most differentiation lives, and reading three competitors’ libraries side by side takes under an hour and reliably changes how you write your own ads — usually by revealing that everyone in your market is saying the same thing, which is an invitation.

Third-Party Tools: What Semrush and SpyFu Actually Know

Third-party competitive tools — Semrush, SpyFu, iSpionage, and similar — fill the gap the native sources leave: estimated keyword lists, estimated budgets, and historical ad copy for advertisers you don’t yet compete with. They work by crawling search results at scale and modeling spend from observed ad positions and clickstream data. That method is genuinely useful and genuinely imprecise, and you need to hold both ideas at once.

What they’re good at: breadth and history. Plug in a competitor’s domain and you get the paid keywords the tool has seen their ads on, the copy variations they’ve run going back years, and which keywords they’ve bid on persistently versus tested and dropped. Persistence is the most trustworthy signal in these tools — a keyword a competitor has paid for every month for three years is almost certainly profitable for them, and that inference survives all the noise in the underlying data. The keyword-gap views, which list terms your competitors buy and you don’t, are the fastest way to find expansion candidates you’d never have brainstormed.

What they’re bad at: precision. Spend and traffic estimates are routinely off by a wide margin — treating them as order-of-magnitude indicators is fair, quoting them as facts is not. Coverage skews toward desktop and high-volume terms, so local and long-tail activity is underrepresented, which matters a lot for service businesses in markets like Toronto or Vancouver where the money terms are geo-modified. And the data lags; a campaign launched last week may not surface for a while.

The sensible posture: use third-party tools for direction — which competitors are serious, which keyword territories they occupy, what they’ve stuck with — and use Auction Insights and your own account data for truth. Direction from estimates, decisions from real data.

Manual SERP Recon: The Free Layer Most Advertisers Skip

The least glamorous source is still indispensable: actually looking at the search results page. Tools summarize; the live SERP shows you exactly what a customer sees at the moment of choice — ad order, extensions, prices in the ad, the map pack crowding the paid results, and how your headline reads next to four others making similar claims.

Do it properly, though, because casual self-searching causes two problems. First, repeatedly searching your own keywords and never clicking teaches Google’s personalization a skewed picture and pollutes your own impression data. Second, your logged-in, location-specific results are not what a customer across the city sees. The clean method is Google’s Ad Preview and Diagnosis tool, inside the Google Ads interface — it shows the live SERP for any keyword, location, device, and language without registering an impression or a click. For local advertisers this is the killer feature: you can stand, virtually, in any postal code in your service area and see which competitors show up there, which is how you discover that your toughest rival in the east end is someone who never appears in your own neighbourhood’s results.

What to record on each recon pass: who holds the top position for your five to ten money keywords, what their headline promises, which extensions they’re running that you aren’t — sitelinks, calls, prices, promotions — and what their landing page does in the first screenful. Click through from the preview tool’s URL listing and study that page: the offer, the form length, the proof elements, the load speed on your phone. Ad copy gets the click, but the landing page is where competitors actually win or lose, and most competitive analyses never look at it.

Fifteen minutes, twice a month, per core market. It’s the cheapest intelligence in this entire article.

Turning Intelligence Into Bids, Copy, and Budget Decisions

Competitive data that doesn’t change a decision is trivia. Every finding from the sources above should map to one of four moves.

Move one: reposition your message. If the Transparency Center review shows every competitor leading with price, leading with speed or guarantee makes your ad the one that doesn’t blur into the others. Differentiation in ad copy isn’t about being better-written; it’s about claiming the angle the SERP leaves open. This is the highest-leverage, lowest-cost action on the list, and it comes straight from reading ads side by side.

Move two: expand or defend keywords. Keyword-gap analysis surfaces terms competitors buy persistently that you don’t — vet them against your own offer and test the plausible ones. The reverse matters too: if Auction Insights shows a new entrant taking impression share on your proven converters, that’s the moment to check your impression share lost to rank and decide whether to defend with budget or bids before they accumulate Quality Score history.

Move three: time your aggression. Competitors who reliably fade at month-end or in the off-season are handing you cheaper auctions on a schedule. If your data shows rivals’ presence dropping in certain windows, those are the weeks to lean in — same budget, redistributed toward the gaps.

Move four: fix the funnel, not the ad. If a competitor consistently outranks you while spending in the same band, the likely explanation is better Quality Score or better conversion economics letting them bid more per click. Their landing page from your SERP recon is the clue. Matching a rival who converts at twice your rate by raising bids is a losing race; closing the conversion gap is the only durable answer.

One caution: never let competitive data overrule your own conversion data. Competitors make mistakes with conviction. Copy their persistence, test their ideas, and trust only your own numbers.

Bidding on Competitor Brand Names: Legal, but Read the Rules

No competitor analysis discussion survives long without this question: can you bid on a competitor’s brand name? In Canada and most markets, yes — bidding on a competitor’s name as a keyword is permitted by Google and has generally been treated as lawful competition. The hard line is trademark use in the ad text itself: Google’s trademark policy restricts using someone else’s registered trademark in your ad copy in most cases, and doing so invites both disapproved ads and legal letters. Bid on the name; don’t put their name in your headline. When in doubt about a specific trademark situation, that’s a question for a lawyer, not a blog post.

Whether you should is a different calculation. The economics are characteristically lopsided: clicks on someone else’s brand terms are usually cheap in absolute terms but convert poorly, because most people searching a brand name want that brand. Your Quality Score on their name will be weak, which inflates your CPC relative to theirs. As typical ranges go, expect conversion rates well below your non-brand campaigns and treat any competitor-brand campaign as guilty until its own conversion tracking proves otherwise.

Where it can work: markets where buyers comparison-shop by default, offers with a genuinely sharp differentiator, and conquesting ads honestly framed as an alternative — “Comparing providers? See how we’re different” outperforms pretending to be them, and it’s the only sustainable posture anyway.

Expect retaliation as a cost line. Bid on a competitor’s name and they will often bid on yours, which raises your own brand CPCs from nearly nothing to something. Sometimes the smartest outcome of testing competitor bidding is the quiet, mutual de-escalation that follows. Either way, run your own brand campaign as defense — it’s pennies, and it makes conquesting against you expensive and unrewarding.

Build the System: A Monitoring Cadence That Survives Busy Months

One heroic competitive audit changes nothing if nobody looks again for a year. The value compounds only when monitoring becomes a calendar item, so make the cadence boring and small.

Monthly, 30 minutes: pull Auction Insights for your top campaigns, segmented by month, and note movement — new names above 10% overlap, anyone’s impression share trending up two periods in a row, shifts in outranking share on your money keywords. Log it in a simple spreadsheet: date, competitor, metric, change. The log is the entire point; trends are invisible without it.

Quarterly, two hours: full workup on your top three to five overlap competitors. Read their current ads in the Transparency Center and note what changed since last quarter. Run the keyword-gap report in whichever third-party tool you use and shortlist new terms worth testing. Do the SERP recon pass through the Ad Preview tool for each core geography, and click through to landing pages to see what they’ve rebuilt. End every quarterly review by writing down, in one or two sentences each, the decisions it produced — copy angles to test, keywords to add, defense moves. No decisions, no value.

Event-driven, as needed: a sudden CPC spike, an impression-share drop, or a lead-volume dip should trigger an immediate Auction Insights check before any internal soul-searching. Competitive cause is the first hypothesis to eliminate, because it’s the fastest to confirm.

A solo operator can run this entire system on native tools and a spreadsheet for free; a third-party subscription adds breadth when budget allows. This is also exactly the kind of standing ritual a good agency should be running for you — at SearchPod it’s baked into monthly reporting, because auction dynamics change whether or not anyone is watching. The advertisers who watch simply find out first, and in an auction, finding out first is the advantage.