At a glance
Cross-engine gauge estimating how much of your branded paid spend on Microsoft Advertising is buying clicks you would likely have won for free. Branded keywords on Bing that also rank top organically, often on Google as well, mean you are paying for visitors who were already coming to you. A high reading is the signal to pause Bing branded or restrict it to genuinely incremental clicks, for example when a competitor is bidding on your brand. This card joins Microsoft branded search performance against organic ranking signals, a cross-channel view neither the ad account nor a rank tracker shows alone.
| What it estimates | The share of branded paid clicks on Microsoft Advertising that would probably have arrived through organic search at no cost. |
| How the join works | Branded paid clicks from Microsoft are lined up against organic ranking strength for the same brand terms, including strong organic positions on Google where most search volume lives. |
| Why cross-engine matters | A brand can rank first organically on Google while still paying Bing for branded clicks. The cannibalisation is invisible if you only look inside one engine. |
| What “branded” means | Search terms containing your brand or close variants. Non-brand terms are out of scope for this card. |
| Unit | A percentage. A higher percentage means more of your branded paid spend is likely redundant. |
| What it does not claim | It is an estimate of incrementality, not a measured holdout. The only exact answer is a brand-bidding pause test; see the FAQs. |
| Currency | The card is a percentage; the spend it puts at risk is in account currency. |
| Time window | 30-day rolling. Branded behaviour is stable, so a monthly view is the right granularity. |
| Alert trigger | A high share of branded paid clicks judged likely to have been free organic. The threshold is tuned so genuinely defensive brand bidding does not over-trigger. |
| Roles | owner, marketing |
Calculation
Calculated automatically from your Microsoft Ads (Bing) branded search data joined to organic ranking signals. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A US DTC skincare brand bids on its own name across Google and Microsoft. The card reads on 30 May 26, trailing 30 days.| Signal | Reading |
|---|---|
| Bing branded paid clicks (30D) | 4,200 |
| Bing branded paid spend (30D) | $1,050 |
| Brand organic position, Google | #1 |
| Brand organic position, Bing | #1 |
| Competitor bidding on brand? | No |
| Estimated incremental share | ~12% |
| Estimated cannibalised share | ~88% |
- You rank first organically on both engines and no competitor is bidding on your brand. That is the cleanest case for cannibalisation. The vast majority of those 4,200 paid clicks would have clicked your free organic listing anyway.
- Roughly 1,050 monthly branded spend is at risk of being redundant. It is not certain waste, paid does add a little real incremental traffic, but most of it is buying clicks you owned for free.
- The defensive-bidding excuse does not apply here. Brand bidding is genuinely valuable when a competitor is bidding on your name and pushing your organic listing down the page. With no competitor present, that justification is absent.
- The right test is a holdout, not a guess. Pause Bing branded for two weeks and watch whether total branded sessions and revenue hold. If organic absorbs the traffic, the spend was redundant; if total branded revenue falls, some of it was incremental after all.
Sibling cards merchants should reference together
| Card | Why pair it with Branded Paid Clicks Cannibalising Organic |
|---|---|
| Microsoft Ads ROAS by Campaign | Branded campaigns carry sky-high ROAS that flatters the account. This card explains why that ROAS is partly illusory. |
| Microsoft Ads Top Keywords by Spend | Surfaces the branded terms eating the most spend, the first candidates to test pausing. |
| Microsoft Ads Spend by Campaign | Quantifies the branded spend at risk so you can size the holdout test. |
| Microsoft Ads Impression Share | If a competitor is taking branded impression share, defensive bidding is justified and the cannibalisation read should be discounted. |
| Google Ads ROAS | The same cannibalisation usually runs on Google branded too; test both engines, not just Bing. |
| Shopify Total Revenue | The truth side of a holdout test. Watch total branded revenue, not just paid, when you pause. |
Reconciling against Microsoft Advertising
Where to look in Microsoft Advertising: Microsoft Advertising → Campaigns → branded campaign → Keywords, with Clicks, Spend, and Impression Share columns. Microsoft will show strong branded performance, but it cannot tell you what share of those clicks you would have won organically, because it has no view of your organic rankings. That comparison only exists when paid performance is joined to organic ranking data. Why Microsoft cannot show this on its own:- The ad account sees paid clicks but not the free organic listing competing on the same page.
- Branded paid ROAS in the Microsoft UI looks excellent precisely because it counts conversions that organic would often have captured for free.
| Reason | Effect |
|---|---|
| Incrementality is inherently probabilistic without a live holdout test. | The percentage is a well-grounded estimate, not a measured truth |
| Organic rank fluctuates day to day and by query variant. | The 30-day view smooths this but cannot be perfect |
| Competitor brand bidding changes the incremental value of your paid clicks. | The estimate accounts for it but a sudden competitor entry can shift it |
| Card | Expected relationship | What causes legitimate divergence |
|---|---|---|
google_ads.gads_roas | If Google branded also shows high cannibalisation, the redundancy is cross-engine and the holdout test should cover both | Organic strength can differ by engine, so the per-engine estimate may differ. |
shopify.total_revenue | During a branded-pause holdout, total brand-driven commerce revenue is the figure that proves or disproves cannibalisation | UTM gaps and assisted paths can blur the holdout read; run it for a full two weeks. |