Most online stores pay for clicks they would have received for free. A shopper types your brand name into Google, clicks your ad at the top, and buys. Your ad report calls that a win. But your organic listing sat right below it, so the shopper may have bought anyway.
This is the question of incrementality: how many sales happened because of the ad, and would not have happened without it? This article shows how to answer that for branded terms and for queries where you already rank first organically.
Why paid and organic clicks overlap
Google shows ads above organic results. If you bid on a term and also rank first for it, your store appears twice on the same page. The shopper picks one, and often picks the first thing they see.
That click is charged to you, and your ad platform credits it with the sale. But the shopper was already looking for you. The free link was one scroll away. The ad did not create the sale. It turned a free click into a paid one.
Overlap is rarely total. Some shoppers click the ad only because it sits higher. You cannot know the split from the ad report, because platform reporting counts every sale that follows an ad click. It cannot show what would have happened without the ad.
Branded versus non-branded
Branded queries include your store name, your product names and common misspellings. The shopper already knows you. Non-branded queries are generic, like "linen bedsheets queen". The shopper is still choosing a store.
Non-branded bidding usually reaches new people, so it has a better chance of being incremental. Branded bidding reaches people who already chose you. Branded campaigns often show a very high return because they are cheap and the buyer is already convinced. That high return tells you little about whether the campaign is needed.
Treat non-branded queries where you rank first organically the same way as brand terms.
When brand bidding is worth it, and when it is waste
When it earns its place
Keep it when one of these holds:
- Competitors bid on your name. If a rival's ad appears above your organic listing, you may lose clicks you would have won.
- You run a promotion. An ad can say "20% off this week" while your organic listing cannot change that fast.
- You want more space. Sitelinks (extra links under the ad) and Shopping listings fill more of the page, which can matter in a crowded category.
- Your organic listing is weak. A vague snippet, or a reseller outranking you, may mean the ad protects sales.
Each of these is a hypothesis, not a fact. A competitor appearing in search does not prove they take your sales. Test it.
When it is waste
Brand bidding is likely waste when no competitor bids on your name, your organic listing is first and clear, and there is no promotion.
A common mistake is judging by return on ad spend (ROAS, revenue divided by ad cost). For example, a campaign spends 1,000 and reports 12,000 in sales, a ROAS of 12. If only 10% of those sales are incremental, the real gain is 1,200 for a cost of 1,000. That is close to break-even before product costs.
How to run a defensible test
You need a comparison. There are two practical designs.
Geo split
Pick regions with similar sales patterns. Pause branded ads in some and keep them running in others. Compare the change in each group over the same weeks. This controls for events that hit everyone, like a holiday.
Time-based pause
Turn the campaign off for a set period and compare against a similar earlier period. This is easier but weaker, because other things change over time.
Either way, measure total results, not ad-attributed results:
- Total clicks from search, paid plus organic.
- Total orders, ideally from all channels.
- Total revenue.
If the ad was incremental, total orders fall when you pause. If organic clicks rise to fill the gap and total orders hold, the ad was mostly moving traffic.
Duration and contamination
Run the test across your normal buying cycle. For many stores that means two to four weeks at least, and longer if orders are few. Extend it until daily numbers stop swinging.
Watch for these:
- Seasonality. Comparing a pause week to a holiday week proves nothing.
- Promotions. A sale lifts branded searches and distorts both sides.
- Other channels. A new email, influencer post or paid social push changes how many people search your name.
- Competitor changes. A rival starting to bid on your name mid-test can look like a failed pause.
Keep everything else steady and write down anything that changes.
Reading the result and reinvesting
Compare total orders and revenue between the test and control groups. If they barely differ, the ad was doing little. If orders dropped meaningfully, divide the spend by the orders lost to get the true cost per extra order. Results vary by store, so do not borrow another store's number.
Then set a rule. For example: "Bid on brand terms only when a competitor bids on our name, during promotions, or when a test shows a clear drop in orders." Review it each quarter.
Send saved budget toward shoppers who have not yet chosen a store: non-branded search, Shopping for products you do not rank for, or product page improvements. Measure that spend by total results too.
What to do next
- Pull your branded campaign report for the last 90 days and note spend, clicks and reported sales.
- Check your organic ranking for your brand name and top product terms, and see whether competitors show ads on them.
- Choose a geo split or time-based pause, and write down the start date, length and what you will measure.
- Run it with promotions and big channel changes kept off the calendar, then compare total orders and revenue.
- Write your bidding rule and decide where saved budget goes.