Attribution vs Incrementality: The Most Important PPC Question

Imagine you spend $100,000 on Amazon advertising this month. By every traditional measure, the month looks like a success.

This month — as your dashboard reports it

Advertising spend$100,000
Attributed sales$620,000
ROAS6.2
ACOS16.1%

The question no dashboard asks

How many of those sales would have happened if you had spent nothing at all?

Most sellers can’t answer. Not because they aren’t experienced — because almost no dashboard asks the question. Yet it may be the single most important one in Amazon advertising.

Attribution feels like success

Advertising platforms are built around attribution. Someone clicks an ad, later they buy, the sale is attributed to that campaign. Simple, necessary, useful — and limited in one major way.

Attribution assumes that because advertising was involved, advertising caused the sale. Real life is rarely that simple.

Imagine someone searches your exact brand name, clicks your sponsored listing, purchases immediately. Advertising receives full credit. Now ask: if that advertisement hadn’t existed, would the customer still have purchased? For many branded searches, the answer is almost certainly yes.

The distinction

The sale was attributed. It may not have been incremental.

Imagine two stores

Guided an existing customer

Store A

Filled with loyal customers. People walk in asking specifically for your product. A salesperson simply points them to the correct shelf.

Created a new customer

Store B

Has never heard of your brand. A salesperson introduces the product, explains why it’s different, answers questions, and convinces someone to buy.

Both sales happen. Both employees receive credit. Did they create the same value? Amazon advertising often treats those two situations as identical. They aren’t.

The illusion of great performance

Campaign Alpha

ROAS 15

Branded keywords, existing customers, extremely high conversion rate. Captures demand that largely already exists.

Campaign Beta

ROAS 4

Generic keywords, new-to-brand shoppers, lower conversion rate. Creates demand that otherwise belongs to competitors.

The trap

Most dashboards celebrate Alpha. Most budget meetings increase Alpha’s spend. One looks better. The other may be far more valuable.

Sales are easy to measure. Influence is not.

Businesses naturally gravitate toward metrics that are easy to calculate — clicks, spend, sales, ROAS, conversions. Incrementality isn’t nearly as convenient, because it requires asking uncomfortable questions:

Would the customer have bought anyway?
Did advertising accelerate the purchase?
Did it merely redirect the click?
Did another campaign deserve the credit?
Did organic ranking do most of the work?

The trade

Those questions don’t produce clean dashboards. They produce much better decisions.

The campaign you’re most proud of might be your least important

Branded campaign disappears tomorrow

A large share of those customers simply click the organic listing instead.

Many sellers fear disaster. In practice, much of the demand survives without you paying for it.

The campaign was recording the business, not creating it.

Generic category campaign disappears

Would those customers still discover your brand?

Probably not. The discovery channel closes and nothing replaces it.

The lower-ROAS campaign may be the one creating the business.

Great businesses buy customers, not clicks

It’s easy to become obsessed with advertising efficiency. But efficiency isn’t the objective — growth is.

Imagine paying more to acquire a completely new customer today, and that customer returns six more times over the next three years. Was the first purchase profitable? Maybe not. Was the investment worthwhile? Almost certainly.

Incrementality forces businesses to think beyond today’s transaction. It asks: what did this campaign create that wouldn’t otherwise exist?

Organic growth complicates everything

One reason incrementality is hard to measure is that advertising changes organic performance. Increase advertising on a new product and sales increase, organic ranking improves, reviews accumulate, brand awareness grows. Six months later, organic sales surge.

The unanswered question

How much of those organic sales were actually created by advertising?

The consequence

Some campaigns look expensive today because they’re quietly building tomorrow’s organic business.

Not every sale has equal value

Customer one

1 sale Searches your exact brand name. Purchases immediately.

Customer two

1 sale Never heard of you. Generic search, compares ten products, reads reviews, then buys.

Both generate one sale. The effort required to create them is completely different, and so is the strategic value. Incrementality recognises that difference. Attribution does not.

Why experienced sellers often overspend

Here’s an uncomfortable reality. Many businesses increase investment in campaigns with excellent historical performance — especially branded ones — because the numbers look incredible. High ROAS, low ACOS, high conversion.

The problem

If those campaigns mainly capture customers already intending to buy, additional investment produces diminishing returns. You’re paying more to secure sales you were already likely to receive. It feels safe. It isn’t necessarily productive.

The most valuable campaigns usually look imperfect

Campaigns that create new demand often have characteristics management dislikes:

Higher ACOS
Lower ROAS
Lower conversion rates
More experimentation
Greater uncertainty

Those imperfections are exactly what make them valuable. If every campaign looks perfectly efficient, your business may not be investing enough in future growth.

Imagine running your business without attribution

Suppose attribution disappeared tomorrow. How would you decide where to invest?

You’d ask these instead

Which campaigns introduce new customers?
Which improve organic ranking?
Which increase lifetime value?
Which strengthen the brand?
Which would hurt the business most if they disappeared?

Those are fundamentally different questions — and much closer to how CEOs think.

Incrementality is about opportunity cost

Every advertising dollar has an alternative use: launching products, increasing inventory, hiring people, improving operations, expanding internationally. If a campaign isn’t generating truly incremental value, that capital could probably create greater returns elsewhere.

Why it matters

It’s not just measuring advertising. It’s measuring whether advertising deserves the capital it’s consuming.

The dashboard we actually need

Imagine opening your dashboard tomorrow and seeing, alongside the usual numbers:

Estimated incremental sales
Estimated new customers
Organic ranking contribution
Brand discovery impact
Repeat purchase influence
Long-term profit contribution

The question today

“Which campaign performed best?”

The question that matters

“Which campaign actually grew the business?”

Final thoughts

Amazon advertising has become incredibly sophisticated — bid automation, AI optimisation, audience targeting, keyword management. Yet one of the most important questions remains surprisingly difficult to answer: would this sale have happened without advertising?

Until businesses begin thinking about that question, they’ll continue making investment decisions on incomplete information.

Attribution tells you where a sale was recorded. Incrementality tells you where value was created. Those are not the same thing.

The companies that learn to separate them stop chasing the appearance of advertising success and start investing in the campaigns that genuinely grow the business. That’s where the next generation of competitive advantage will come from — not better bids, but better decisions about what advertising actually creates.