Profitability improvement year on year on one major brand account I managed in Spain: +444 basis points of NetPPM, Amazon's net product margin. One account, not the whole company.
Newest first. These are the ones I'd tell you about over a coffee.
01 · Freelance · Hospitality growth
Low season, no budget
Building a growth strategy for a premium villa rental business
The starting point
The founder had built a business that did well when there was little competition. By the time I met them, that wasn't enough. The team was small and mostly reactive, the brand had no clear positioning, the website was outdated, and the OTA set-up left little room to customise.
I mapped what needed to change: a rebrand, a new website, a more flexible OTA platform and a real acquisition plan. The investment wasn't approved. So instead of waiting for the ideal set-up, I worked with what we had.
What I did
01Content with a job to do. Recurring series that were aspirational (the destinations and the lifestyle), educational (planning and travel tips) or commercial (villas, availability, offers), plus new TikTok and Pinterest profiles. When the photographer and the agency were cut, I kept it going with free stock libraries and AI-generated content, at no extra cost.
02Email that sells, not a catalogue. A monthly property showcase for past guests became a monthly proposition: early access to new villas, exclusive codes, themed issues for families and groups. Sent to a wider audience, not only past guests.
03Agency outreach on autopilot, with a human at the end. A Google Apps Script found European travel agencies; AI scored each one on destination and customer fit, found its public business email and drafted an introduction to our affiliate model. I read and approved every email before the automation sent it.
The results · September 2025 – February 2026, low season, vs the year before
+34%direct bookings year on year, the channel where my work had the most direct impact
8% → 25%email open rate, 3.1x what it was before
500+travel agencies found and approached in the first month
+110%bookings year on year for the whole business, across direct and OTA channels
Organic traffic and coupon-code use went up too, and we secured an influencer partnership. All of it pushed towards the bigger goal: a business that depends less on OTAs and can win its own customers.
The +110% is the whole business's booking growth over those six months, not revenue, and not the result of my work alone.
02 · Commercial growth · Camper
The ballerina bet
Turning market intelligence into commercial growth
In 2024, the winter season was unusually warm, and winter styles weren't selling as planned. Then a colleague shared market intelligence from Camper's largest digital sales channel: demand was growing for classic, high-quality ballerinas. That channel wasn't part of my portfolio. My luxury-tier accounts had very little ballerina stock, and production took about two months.
I took the insight to the Product team and made the case to double production for the next delivery, focusing on carryover styles that matched the trend better than the bolder seasonal designs.
Stock was still limited, so I didn't give every marketplace the same thing. Selected niche accounts got exclusive variations: one the matte black, another the shiny finish. Each had a reason to push Camper, and the demand supported a price increase of about 10%. With 24S, the LVMH-owned luxury marketplace, I used that exclusivity to negotiate extra banner visibility, at no advertising cost.
Ballerinas finished at 170% of target, and every account I managed hit its €350K monthly sales target, in a season that was hurting the rest of the range.
Ballerina performance vs target170%
2xproduction requested for the next delivery
10%price increase on the ballerina offer
24Sfree banner visibility, in exchange for exclusivity
€350Kmonthly sales target hit on every account I managed
2021–2023 · PAN-EUROPEAN BRAND ACCOUNTS ON AMAZON VENDOR MANAGER
Amazon case studies
Two and a half years, seven accounts (eight brands), a lot of deal events.
03 · Profitability+444bps NetPPM, year on year, one account
Promotions that paid for themselves
One of the brands in my portfolio was a global consumer brand I'll keep anonymous. Promotions move volume, but they can quietly eat margin if nobody is watching what each deal actually earns.
My answer was an always-on promotion approach, negotiated with the brand, so customers kept getting good deals and the account kept earning from them. In August 2021, the category's vendor manager wrote to say it was helping the category "boost both growth and profitability, apart from delighting customers with nice deals".
By May 2022, profitability on the brand's Spanish business was up 444 basis points of NetPPM year on year, Amazon's net product margin. The same negotiations also softened a tough comparison on sales against the year before.
It's a single-account result, not a company-wide one. It's also the clearest example of how I like to work: more deals aren't always better deals.
On file
04 · Peak season
Five Prime Days, two Black Fridays, eight brands
Peak events are where a marketplace account shows what it's made of. Deals have to be planned weeks ahead, stock has to be there on the day, and everything happens at once across every brand you manage.
Across five Prime Days and two Black Fridays, I ran eight Amazon Spain brands through it. Six of them beat their targets.
Prime Day 2023+7%
My vendor set closed Prime Day 2023 seven per cent above its Deal OPS target, the ordered product sales generated through deals.
05 · Change management
Getting 150+ people to actually use a new tool
DOTv2, the second version of Amazon's Deal Optimizing Tool, was rolling out across the whole Europe offshore brand-specialist organisation, more than 150 people. When I became the main point of contact for its adoption, 65% of deals went through it, or 68% weighted by deal value. The goal was 90% on both.
Announcing a tool is easy. Getting a large team to change how it works, week after week, is not. So I treated it as a change-management job rather than a training slot, and used the stretch between two major deal events to train the whole organisation, more than 150 people.
I wrote the training content and ran the European sessions, owned the users' feedback log, turned the questions from the trainings into an FAQ wiki, and worked with the onshore senior PM to fix access issues that were blocking a negotiations dashboard. The tool's engineers also invited me to beta-test it.
In October 2022, offshore adoption reached 84.9%: 4.9 points above that month's goal, and 30.5 points ahead of the onshore teams. An internal newsletter named me First Among Equals for it.
On file
How
01Weekly checks on adoption by brand, deal and specialist, with targeted follow-up where it lagged.
02Monthly newsletters and action plans, so the tool stayed visible after launch day.
03Small sessions of 8–10 people, three organisation-wide sessions and a recorded session for European teams.
04Weekly office hours, plus rotating ones with European teams. I didn't miss one.
05Escalation to leadership when a gap needed more than a reminder.
06 · Financial recovery · Office Products brand
The three-year shortage claim
One of my brands, in Office Products, had goods lost in Amazon's fulfilment centres. The claim had been open for more than three years, and the brand was still waiting for its money.
I stopped treating it as historical noise. I took it to Finance and General Ledger leadership, set out the money involved, and made it clear the relationship with the brand was at risk if it stayed open any longer. Then I pushed for a confirmed resolution instead of another follow-up.
The disputed cases were closed with a 100% refund to the brand. It got its money back in full: €24,803. The vendor experience improved noticeably afterwards.
On file
The best part
My EU manager celebrated. So did I.
07 · My own idea · Amazon Vendor Services
Kamala Project
Turning proactive customer support into a scalable business practice
The challenge
Product issues were creating customer contacts and concessions that didn't need to happen. Manufacturers already had troubleshooting content; customers just weren't seeing it before things went wrong.
The approach
It was my idea, and I owned it within Amazon Vendor Services (AVS), together with a colleague from the Concessions team. We used Amazon's data to find the brands and products where help would matter most, then worked with manufacturers to put the right information in front of customers.
The results · within the first three months
6brands onboarded
375pieces of troubleshooting content implemented
1,703reported customer support contacts avoided
€41,000in combined reported interim savings from support contacts and concessions
11.2%reduction in Net Concessions
+450 bpsimprovement in UPR
From my idea to their routine
Three months in, it was working, and it was growing beyond what two people could run. So I did the less glamorous part of owning an idea: I handed it to the Concessions team, the people who should run it every day.
The takeaway
Owning a project also means knowing when it's ready to belong to someone else.
Results reflect the initial reporting period and should not be interpreted as final program outcomes.
The paper trail
Letters, notes and the occasional "well done".
Accolades, emails and a newsletter from my Amazon years, as they were sent. Pick one to read it. Names, emails and brands are blacked out.
Seven pages, 2021–2022. All from Amazon. Swipe to see them all.
From the other side of the table · December 2022
“Since she has been supporting us, we have noticed improvements … We are satisfied with the work developed by Eva.”
A key account manager at one of my vendor groups wrote to Amazon a few months into working with me: a cleaned-up catalogue, weekly updates on out-of-stock items to raise their confirmation rate, and clear next steps to improve their product data quality score.
On file
Before all this: fashion product management in Madrid and an ecommerce team of eight at Ekseption. The full story, chapter by chapter, is on the About page.
