$100 Million in Ad Spend — How Many Sales Did It Actually Buy?

A while back, a friend who works in marketing at a consumer goods company vented to me.

He said the year-end budget meeting is the most nerve-racking moment of his year. The boss asks: that $100 million in marketing spend — how many sales did it actually drive? He can't answer. All he can do is grit his teeth and list every channel they poured money into over the year: TV, elevator screens, in-feed ads, livestream selling, field marketing...

There's the list. Then what?

Then there is no "then what." Because nobody knows which half of the money worked.

That line probably sounds familiar. The advertising industry has an old saying: I know half my ad spend is wasted — I just don't know which half. A question from a hundred years ago, and today plenty of companies still can't answer it.

Why?

Because nobody can do the math

Think about it: a consumer goods company might run a dozen or more brands, selling into dozens of markets worldwide, and each market has several channels — e-commerce, supermarkets, convenience stores, livestreams. Each channel has its own data, in different formats, with different definitions — they can't even agree on how to count "sales from this campaign."

The data lies scattered everywhere, and nobody can assemble the full picture.

The result: budgets get allocated by gut feel. Spent heavily on TV last year? Put more into TV this year. Competitors are all doing livestreams? Then let's set up a livestream studio too. If the gut call is right, everyone's happy; if it's wrong, tens of millions go down the drain.

In 2023, a Fortune 500 consumer goods (CPG) company handed this exact problem to WNS Analytics — the team within WNS dedicated to data, analytics and AI. Their playbook is worth taking apart.

What is Market Mix Modelling?

First, a term: Market Mix Modelling, or MMM.

Put simply, it comes down to one thing: feed everything into a machine learning model — how much was spent on each channel historically, and how much sales rose in the same periods — and let it calculate the true contribution of every channel and every campaign.

How much did TV ads bring in? In-feed ads? Price promotions? And how much came from factors that have nothing to do with marketing spend, like weather and seasonality? Line by line, worked out clearly.

It's a bit like giving the company a full physical exam. Before, you only knew the total number on the scale; now you've done the CT scan and the blood work — which organ is healthy, which one is weak, all laid bare.

Of course, before the physical, the shattered data has to be pieced back together. The first thing WNS Analytics did was clean, align, and consolidate all of this company's scattered data — traditional media, digital media, consumer research, attitudinal studies — onto a single platform. This is the dirtiest, most exhausting step, and there's no way around it.

Then, the modelling. AI models quantified the effect and efficiency of every brand, every channel, every marketing lever — and the models get regular checkups, to guard against results drifting over time.

Finally, they equipped the marketing team with a simulator and optimization tools. What does that mean? From now on, budget allocation no longer relies on gut calls. Feed next year's plan in, and the model tells you directly: allocate the money this way, and the return is roughly this much; shift it another way, and here's how much more you earn. Marketers can run it themselves, without waiting for an analyst to free up a slot on the calendar.

Once the math is done, the money speaks for itself

So, the results?

Every dollar of this company's $100 million marketing budget is now tracked in real time. What's working, what's burning cash — crystal clear at a glance, with campaigns adjusted on the fly.

The two numbers that follow are even more interesting: total sales up by more than 3%, and profit up by more than 5%.

And note — the budget didn't rise by a single dollar.

The same hundred million used to buy only so much sales. Once the math was done and the money was allocated right, that same hundred million squeezed out 3 extra points of sales and 5 extra points of profit. Those 3 and 5 weren't blasted out by throwing in more money — they came from reclaiming the wasted half and moving it to where it truly works.

That's the value of doing the math. Same budget, different allocation, different result.

Final words

After hearing this case, my friend was quiet for a moment, then said: so the problem was never that we didn't have enough money — it's that we never knew where the money went.

Exactly. Marketing waste usually isn't because the team isn't working hard; it's because the math was never done. When the numbers are murky, no amount of dashboards and post-mortem meetings will help — you're just circling around a fuzzy number.

Drucker said you can't manage what you can't measure. In marketing, that line cuts especially deep: if you can't measure, you don't even know where the waste is.

May every cent of your marketing budget be spent with total clarity.