Amazon Advertising Manager

Lucky 21

Location: Remote, Philippines

Hours: 9:00pm – 5:00am Manila, Monday to Thursday nights ·9:00pm – 4:00am Friday. This matches our New Jersey hours (9:00am – 5:00pm ET). 39 hrs/week

Rate: $5.00 – $7.00 USD/hour on demonstrated experience (roughly $845 – $1,183/month)

Paid: Every two weeks

Start: Mid-August 2026

▶ LOOM VIDEO — 60 Second Introduction from our President

First, the hours

9:00pm to 5:00am Manila, Monday through Thursday nights, and 9:00pm to 4:00am on Friday. That is our New Jersey working day, which is the whole point of the seat — the buyers you work with are in NJ and they message you during their work day.

We are blunt about this up front because we would rather lose you now than in week three. The overlap is not partial and not optional. If you have worked overnight before, this is a good job. If you have not, think hard. We ask about your sleep plan, workspace, and backup internet in the application, and we read those answers.

What this job is

Lucky 21 runs the marketplace business for apparel brands on Amazon and Walmart.com — Tommy Hilfiger, Reebok, Puma, U.S. Polo Assn. and others.We are a full service agency: buying, inventory, pricing, shipping, advertising, and reporting. Advertising is one piece of a much larger operation.

That is what makes this role different from an ad agency job. You can see what a product cost, what is left in stock, what is on order, and what it actually earns — and you are expected to use all of it to grow the business profitably.

You will own the advertising for several real storefronts, decide where the spend goes, defend those decisions to the buyers who own the merchandise, and be measured on profit rather than clicks. Most applications we get are from people who have watched PPC videos and can talk about ACoS. This process is built to find the few who have made hard calls with real money.

The daily work

Every morning a data file lands for each storefront — one row per product, with a parent-level pivot carrying inventory on hand, units on order, weeks of supply, price, wholesale cost, Amazon fees, trailing sales, ad spend, ACoS, and current budgets.

Our tooling flags products where something looks off and proposes a change. It is good at arithmetic and knows nothing about context — whether a reorder just landed, whether a product is being discontinued, whether last week was a fluke.You are the judgment layer. You work down the flagged rows, approve, reject, or override with your reasoning, and the approved changes go out as Feedvisor bulk files.

That review sets the daily rhythm. The rest of the job matters just as much:

Budget allocation and pacing — one of the most important parts of this role.How much should a storefront spend in total. How should that be split across products, so the money sits behind the ones that deserve it. And is it actually happening: staying inside the number, while also spending FULLY across the day rather than running dry by noon or leaving budget unused. Getting this wrong is expensive in both directions, and it needs watching constantly rather than once a month.
Launching new products.New arrivals land constantly. You build the campaign structure, research and seed the keywords, set the opening budget and target, and nurse it through its first several weeks.
Search term harvesting.Find the terms converting inside broad and auto campaigns and promote the winners into the right campaign as exact match. Not every good term belongs in the campaign that found it.
Cut the terms burning money with no conversions, at the right match type — and know the difference between a term worth negating and a broad root that is feeding the rest of the campaign.
Keyword and rank research.Find the terms a product should be ranking for, track where it actually ranks, and close the gap.
Brand leakage,dividing each storefront’s daily spend pool across products, and answering “why are sales down on this” with evidence.

What we need you to get right

Profitability, not ACoS.A 30% ACoS is fine on a 45% margin and a disaster on a 22% one. You need to compute what a product actually earns after wholesale cost, commission, and FBA fees, and judge spend against that.
Traffic and rank on new products — where the profit rule does not apply.A new item has no reviews, no rank, no history. Early spend is buying velocity and organic position, and it will run at an ACoS that would be unacceptable on a mature product. The judgment is knowing when that is investment and when it is waste: is it converting, is it gaining rank week over week, and at what point do you conclude the problem is the price or the listing rather than the ads. Getting it backwards either way is expensive.
Inventory reality.Never scale spend into a product with three weeks of stock and nothing on order — going out of stock costs the rank we paid to build. It cuts both ways: a product sitting on a year of inventory needs a different answer than the profit math alone suggests. This is the most common blind spot in ad people who have never carried inventory.
Budget caps versus actual spend.Campaigns rarely spend their cap. If you do not understand why that means your caps must sum above your target spend, you will chronically underspend the account.

Working with the buyers

Our buyers own the merchandise and the purchasing decisions. They are not your clients and not your bosses — they are counterparts, and you will disagree with them regularly. One will want spend pushed on a product they over-bought. Another will want ads pulled off something quietly profitable. A third will ask why their launch is not selling when the answer is the price, not the ads.

Hold your position when the data supports it, change it when it does not, and say either one in language a busy person reads in ten seconds:

“Cutting this from $40 to $18. Its ACoS is 62% and it earns 24%, so every extra dollar is losing money. Happy to revisit if you want the visibility for the promo, but wanted you to see the tradeoff.”

You also work overnight and largely unsupervised. That means daily reviews land on the day they are due, you flag a slipbefore the deadline rather than after, and when you make a mistake you raise it immediately with the number attached, before you have finished fixing it. We are far more forgiving of errors reported fast than of small problems found late.

The tools

Feedvisoris where our campaigns live and the tool you will spend the most time in. Most campaigns are managed to atarget ACoS rather than by hand-setting bids — you choose the target and the budget, and Feedvisor handles the minutiae underneath, including individual keyword bid prices. Your skill is picking the right target and knowing how long to leave it alone before judging it. Prior Feedvisor experience is a significant advantage; we will train the right person who does not have it.

Helium10for keyword and rank research — finding the terms a product should be ranking for, sizing the opportunity, checking what competitors are doing, and tracking rank movement on a launch.

Amazon Ads console and Seller Centralfor search term reports, Search Query Performance, Brand Analytics, and campaign diagnostics.

Requirements

2+ years on Amazon Sponsored Products and Sponsored Brands, on accounts you can describe in specifics — SKU count, monthly spend, what you changed and what happened.
An Amazon Ads certification.Amazon Ads Academy offers these free (Sponsored Ads Advanced, Campaign Optimization, or Advanced Retail). Send it with your application. They take an afternoon.
The profit math, unassisted. Given price, cost, commission, and FBA fee, you can tell us what a product earns and where ACoS breaks even.
Serious Excel.Not “familiar with Excel.” You will work in large macro-enabled workbooks every day and you need to be fast in them: pivot tables, the lookup and conditional-aggregation formulas, filtering tens of thousands of rows, and — most importantly —pulling a meaningful answer out of a large messy dataset rather than just navigating it. If someone hands you 40,000 rows and asks which products are the problem, you should know how to find out.
You use AI as a working tool.We run a lot of our analysis through Claude, including Claude connected directly to our database, and we write our own Claude skills. You need to already be using AI seriously for research and analysis rather than just for rewriting emails, and to be interested in building tooling rather than doing everything by hand. Someone who wants to automate the boring half of this job will do well here.
A workspace that supports overnight work — reliable power, backup internet, somewhere quiet.

What helps

Experience launching products from zero and taking them to a rank you were proud of
Prior Feedvisor experience, or another platform where you managed to a target rather than hand-bidding
Apparel or softlines, where size and color variations make everything messier
Time on the inventory, buying, or merchandising side rather than only the ad side

Application code: ADS-2026 — the application form will ask you for this.

Terms

Independent contractor, paid every two weeks.The offer includes a two-week paid onboarding period at your full rate — real work on real storefronts with your decisions reviewed before they go live. It is how we get you productive and how we both confirm the fit, and you are paid in full for it either way. Our processes are documented properly; you will not be dropped into an account and told to figure it out.

How to apply

The application takes about ten minutes. Have ready: your résumé, a link to or copy of your Amazon Ads certification, and the application code above. It also asks you to work through a short profit calculation and show your numbers, so do it somewhere you can think rather than on your phone.

Optional, and it helps:the form has a slot for a Loom or short video of yourself, up to two minutes — who you are and why this role. You saw ours above. A webcam is fine. We watch all of them.

Skip the cover letter. We would rather read your answer to the calculation.

What happens next:plausible candidates get a written assessment with real numbers plus a short Excel task. Strong assessments go to a group interview with a few other candidates and two or three of us. Then an offer.

We have openings for 2 to 3 people and want them started by mid-August.

Lucky 21 is an equal opportunity employer and considers all qualified applicants without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, veteran status, or disability status.

Lucky 21 remote jobs

Amazon Advertising Manager

Design Coordinator – eCommerce (Amazon & AI Image Generation)