SEO or Google Ads: only pay where you are absent

Paying for a visit that would have come anyway

In many companies, organic search and advertising live in two worlds, one looking at rankings, the other at cost per click. Nobody asks the question that links them: on this query, do we need to pay?

If your site ranks first organically for a search and your ad shows right above it, some of the ad’s visitors would have clicked your organic result: those clicks, you pay for nothing. Conversely, on queries where you are absent from the first page, the ad is often your only presence: that is where every euro works.

The principle: pay where you are absent, and don’t pay blindly where you are already present. The whole difficulty lies in “blindly”.

What Google’s studies measure

Google studied the question by observing advertisers who had paused their campaigns. Its first publication, in July 2011, covered more than 400 of these pause studies and concluded that on average, 89% of ad clicks were not replaced by organic clicks when the ads stopped. A March 2012 update, on more than 5,300 cases, brought that figure to 85% for paused campaigns.

The same year, a second analysis looked at what changes depending on the advertiser’s organic position for the same search. That is the one that answers our question:

Two honest readings. First, cannibalisation exists, but it isn’t total. The arithmetic: out of 1,000 paid ad clicks on a query where you rank first, the study’s average says about 500 would have arrived without the ad. Second, these are averages: the authors warn that an advertiser’s results may differ, that the study measures clicks and not sales, and recommend controlled experiments. They also date from 2011 and 2012, before AI Overviews.

The weekly review: crossing the two lists

The crossing starts from two lists you already have: your organic rankings and the queries that trigger your ads. Google Ads makes it easier with its paid and organic report, which puts side by side, per query, the statistics of your text ads and of your organic results, once Search Console is linked to the account. Two limits to know: it only counts text ads, not Shopping, and your ads’ location settings don’t apply to organic results. To know where you really rank first, city by city, you need rank tracking by area, not a national average.

Every week, each important query falls into one of these four boxes:

1. 01Absent organically, the query sells.

   Buy it, and polish the ad: this is where advertising pays most.

2. 02First organically, ad running.

   A candidate for a cut test, except on your brand (see below).

3. 03Present organically, no ad.

   Leave it alone, and watch the ranking.

4. 04The ad brings in customers, no page answers the query.

   That is an SEO page to write.

The fourth box turns advertising into market research: the queries that bring qualified requests, measured all the way to the sale as detailed in our guide to sending CRM sales back to Google Ads, are the ones where an organic page will be worth most.

Testing query by query before cutting

Cutting every ad on queries where you rank first, all at once, means betting on an average. The test runs query by query, over a fixed period:

If the total holds, the budget goes to a query from the first box. If it drops, you turn the ad back on. To rule out the effect of a season or the news, compare with the same period in an area where the ad stays live: that is the principle of the geo experiments the study authors recommend.

Your brand: the case where the ad is often worth keeping

Your brand name is the query where you most often rank first organically, so the first one you want to cut. It is also the one where a competitor can place itself above you: Google’s trademark policy doesn’t restrict buying a brand name as a keyword. Cutting your brand ad sometimes means handing the top line of the page to someone else.

The question becomes “who takes the spot if I leave?”. So the weekly review checks which competitor ads appear on your name: if nobody bids, a cut test is defensible; if a competitor is there, the brand ad protects traffic that is yours.

Ads pick the SEO pages to write

SEO takes months to respond, a campaign a few days. Use that gap: when a query converts in ads and none of your pages answers it, write the page, then scale the ad down as the page climbs, with the same test as above.

When those queries look alike (one service declined by city), the answer can be a series of pages generated from your data, provided each one brings real information: see when programmatic SEO works, and when it hurts.

We apply this reading discipline to our own site: our SEO engine reads its Search Console and files each target query as invisible, seen, within reach or active. That is the first half of the crossing: the list of paid queries lands on top of it.

Where to start

Export the paid and organic report, sort by spend, and look at the ten most expensive queries: how many are queries where you already rank first? If you’re still choosing a platform, start with Google Ads or Meta in B2B.

To run this crossing every week, our programmatic Google Ads offer adjusts searches and budgets continuously, on an account that stays in your name. And if you want advertising, SEO and CRM to work as one, that is what digital marketing in one piece is for.

Frequently asked questions

Sources

1. Google Research, Studies Show Search Ads Drive 89% Incremental Traffic, 21 July 2011 (new tab)research.google
2. Google Research, Search Ads Pause Studies Update, 12 March 2012 (new tab)research.google
3. Paid and organic report (new tab)support.google.com
4. Google’s trademark policy (new tab)support.google.com