The context
The brand is a historic Italian pasta maker, with a mill next to the pasta factory. It makes dried pasta from Italian durum wheat and a few organic lines, and has an online store where anyone can buy directly.
The pasta maker works with the agency I consult for, and I follow the online store's numbers and the campaigns. From 2023 to 2025 online revenue had stayed almost the same, year after year.
The pasta market is saturated, and competitors have bigger budgets. The people buying online already knew the brand. To grow, it had to reach people who didn't know it yet.
The problem was the store, not the advertising
The company's plan for 2026 said it plainly. Visits to the online store had grown by 116%, but the share of visitors who bought stayed at 2%. Bringing more people to those pages would only have increased the investment, without increasing sales by the same amount.
The product pages were all the same. The title was just the name of the pasta shape, the description was short and often identical across products in the same line. FAQs, cooking tips, reviews and trust signals at the top of the page were missing.
Product pages, rewritten for buyers
I wrote a proposal for the site manager that applies to all product pages, to be applied once to the store template and in phases. Each title states the shape, the grain, the line and the weight, which are the words people search for on Google. The descriptions become longer and different from product to product.
Every page gets answers to the most common questions, a section on how to cook the pasta and what to pair it with, and reviews. This is the information buyers need, and it's the same information Google uses to judge the quality of the page an ad lands on.
The budget is measured against margin
Before deciding how much to invest, I worked out what a customer is worth. The store's contribution margin is 48%, and a customer spends on average about €80 in the first twelve months. That means for each new customer you can invest at most €38.40 to break even.
The stop rules came from that number. If a campaign passes a certain investment threshold without bringing in a customer, it is stopped. Twice a week an automatic check applies these rules to the Google campaigns, and keeps safe the ones that bring the most orders.
The same calculation made us switch off the Google campaign on generic searches in August 2026. It was reporting €0.51 of sales for every euro invested. The budget went where it paid off.
Twelve videos, tested one by one
On Meta the job is to find people who don't know the brand. Twelve documentary-style videos arrived, shot with an actor, telling the story of the wheat, the supply chain and the drying. If twelve videos sit in the same ad group, the budget is split into pieces too small and none of them gets out of the platform's learning phase.
That's why we proposed a test campaign with a separate budget for each group, to understand which videos bring new customers. The months when people buy less are the right ones for these tests, so that you reach the peaks with the creatives already chosen.
Counting real customers, not attributions
In August Meta attributed 124 purchases to itself. The orders whose first visit came from Facebook, read in the online store, were 71. Platforms tend to take credit for sales that would have come anyway.
That's why we measure the investment for each new customer on the new customers registered in the online store, above the average of the previous months. It's more laborious, but it's the number that decides whether the advertising makes money.
The results
The Google and Meta campaigns brought in over 4,000 online orders. In 2026, after three almost flat years, the online store's revenue from January to September grew 31% on the year before.
The plan for 2027 starts from here: Meta to find new customers, Google for people already searching for the brand, and a cap on investment set month by month from the break-even per customer.
What this case teaches
1. If the store doesn't convert, advertising isn't enough
Traffic had more than doubled and sales hadn't. The first intervention was on the product pages.
2. The right budget is calculated from margin
Margin, customer value over the first twelve months and break-even per new customer tell you how much you can invest. The rest is opinion.
3. A campaign that doesn't pay off gets switched off
Keeping it on out of habit takes budget away from the ones that work.
4. You count new customers in the management system
Platform attributions are worth reading, but the decision is made on real orders.
Frequently asked questions about marketing for pasta makers
How much should you invest in advertising to sell pasta online?
You start from margin. Here the contribution margin is 48% and a customer spends about €80 in the first twelve months, so break-even is at €38.40 of investment for each new customer. Above that figure the advertising makes sales, but doesn't make money.
Why do visits grow and sales don't?
Usually the problem is in the store. Product pages that are all the same, short descriptions, no reviews and no answers to questions hold back people who are about to buy. First you fix the pages, then you raise the budget.
Can you trust the purchases Meta attributes?
You read them, but they aren't enough. In this case Meta attributed to itself 124 purchases in a month, while the orders whose first visit came from Facebook were 71. Decisions are made on the new customers registered in the online store.


