The Cross-Channel Cannibalization Matrix (Brand Search, Retargeting, and Email walk into a bar)
Published September 17, 2026
The 60-second version
- Three channels claim credit for one purchase whenever a buyer clicks a newsletter, searches your brand, and views an ad before buying.
- Summing channel ROAS manufactures phantom growth. Brand search and retargeting often have single-digit incrementality.
- The fix: build an interaction window in your warehouse, cap brand spend to competitor conquesting, and dark-test retargeting.
A customer opens your Thursday newsletter, searches your brand name on Google to get back to the tab they closed, and scrolls past an Instagram retargeting ad on their phone before completing a ₹3,500 purchase.
Google Ads reports a conversion with a 14x ROAS. Meta reports a conversion with an 8x ROAS. Klaviyo reports a conversion with a 22x return.
The disagreement is not an attribution bug. It is that three separate teams just took 100% credit for the exact same order. When you sum your channel dashboards, you count ₹10,500 of attributed revenue from a single ₹3,500 bank deposit. The gap between your marketing reports and your P&L is not fraud — it is cross-channel cannibalization, and it is entirely measurable once you build the matrix.
Three invoices for one customer
📈 Total claimed: ₹10,500
Every channel claims credit as if it acted alone. Brand search takes credit for the navigation query. Retargeting takes credit for the impression. Email takes credit for the link.
- Google: ₹3,500 (14x ROAS)
- Meta: ₹3,500 (8x ROAS)
- Email: ₹3,500 (22x ROAS)
💰 Cash received: ₹3,500
The order happened once. The customer had high intent. You paid three different toll booths to let someone through the front door who was already holding the door handle.
- One customer, one order
- Three ad fees paid for zero incremental lift
When channels operate in silos, bottom-of-funnel channels compete with each other to harvest customers who were already going to buy.
The core error is treating every touchpoint as a creation of demand. In reality, marketing channels split cleanly into two functions: demand generation (introducing someone to your product) and demand harvesting (providing the easiest hyperlink to someone already determined to buy).
Harvesting looks brilliant in dashboards. Because buyers already have intent, conversion rates are high, CPAs are low, and ROAS looks astronomical. But harvesting adds zero incremental units if the customer would have navigated to your site anyway.
The three channels that harvest rather than create
The most common cross-channel collision happens between three specific channels. None of them is inherently bad, but left unconstrained, they quietly consume budget by bidding on the same warm intent.
Brand Search (Google Ads)
Buying your own company name. You pay ₹15–₹45 per click for people who deliberately typed your brand to find you, often poaching clicks from your #1 organic link.
Warm Retargeting (Meta / TikTok)
Serving ads to people who visited your cart in the last 48 hours. 1-day view attribution means anyone who bought on desktop after seeing an ad in their feed counts as an ad win.
Automated Flows (Email / SMS)
Abandoned cart flows firing while retargeting ads run. Both claim credit if an email is opened within a 5-day window of an ad click.
When you turn up spend on brand search or aggressive retargeting, platform ROAS almost always improves. But total business revenue stays flat while marketing expense increases. You are simply paying rent on existing momentum. the toll booth trap
The cannibalization matrix: a worked case
Here is what happens inside a typical D2C brand spending ₹4,50,000 a month across bottom-of-funnel channels.
A ₹4.5L harvesting spend that bought ₹80k of real lift
The true lift clears your ad costs. Keep monitoring your organic baseline so harvesting doesn't creep up over time.
Building the overlap matrix in SQL
To see how severe this is in your own business, you do not need an attribution software tool. You need your warehouse orders table and touchpoint logs.
This query looks at all orders in the last 30 days and checks how many had multiple touchpoints inside a narrow 4-hour window before purchase.
Show SQL
When you run this, pay attention to the rows where two or three harvesting channels overlap. If 40% of your brand search buyers also opened an abandoned cart email in the same 4 hours, your brand search campaign is not acquiring customers — it is charging you a toll for an email customer to find their checkout cart.
Monday morning harvesting audit run these three checks →
- Check your Brand Search impression share — if absolute top impression share is 98% and competitors are not bidding on your name, test lowering max CPC bids by 40%.
- Exclude 30-day past purchasers from Retargeting — stop serving catalog ads to customers who bought yesterday.
- Run a 7-day geo holdout on Retargeting — turn off retargeting in 3 representative cities and verify if delivered revenue in those cities actually drops.
Quick gut-check
One question. Get it and the whole post clicks. 30 seconds, no maths!
You pause Google Brand Search for 2 weeks. Total store revenue drops by 2%, while Google Ads spend falls by 18%. What does this tell you?
Frequently asked questions
Should I turn off Google Brand Search completely?
Not necessarily. If competitors are actively bidding on your trademark name, turning off brand search lets them capture your intent at the top of the search results page. The right strategy is defensive bidding: bid only on your exact trademark, set strict target impression share (e.g. 70–80%), and cap max CPCs tightly. Never run broad match keywords in a brand campaign.
Doesn't retargeting remind people to finish their checkout?
Yes, but you only need to remind them once or twice. High frequency (e.g. frequency > 6 on a 3-day window) simply spams buyers who were already going to buy on payday. Cap your retargeting frequency to 2 impressions per day and exclude existing purchasers immediately.
How does Klaviyo/email attribution fit into this?
Email platforms typically use a 5-day click window. If a customer clicks an email on Monday and buys on Thursday via a Google Search ad, both Klaviyo and Google claim 100% of the sale. Compare both against your warehouse order timestamp to avoid double-crediting.
What metric should we report to leadership instead?
Report MER (Marketing Efficiency Ratio) — total warehouse revenue divided by total marketing spend. When you reduce wasteful harvesting spend, channel ROAS might appear lower, but overall company MER and contribution margin will rise.
The summary
- Summing channel dashboards creates fictional revenue. When brand search, retargeting, and email touch the same buyer, each claims 100% of the conversion.
- Demand harvesting channels (brand search, warm retargeting) capture existing intent; they rarely create new demand.
- Running high spend on brand search when no competitor is bidding on your name is paying Google a tax on your own brand awareness.
- Build an overlap matrix in your warehouse to see what percentage of orders are touched by multiple bottom-of-funnel channels within 4 hours.
- Verify true causal lift by running lightweight geo holdout tests before scaling retargeting budgets.
Takeaways for your next report
- Summing platform revenues double- and triple-counts sales whenever buyers touch multiple channels.
- Brand search and retargeting harvest existing demand rather than creating it.
- A high dashboard ROAS on harvesting campaigns frequently masks an incremental ROAS below 0.5x.
- Use a warehouse overlap query and geo holdouts to find your true incrementality baseline.
MER Calculator
Total revenue divided by total marketing spend — the attribution-agnostic efficiency number, plus its contribution-margin-adjusted variant.
Chinmay Raibagkar
About author →Founder of DataLens AI. He helps non-technical teams read their ad and database numbers with confidence — which number to trust, what to do next, and what to ignore.