188: How To Run A 7 Figure Ecommerce Business As A Digital Nomad With Greg Mercer

Podcast: Download (Duration: 51:53 — 59.7MB)

You can run a seven figure Amazon business from anywhere in the world by never touching your inventory. Greg Mercer sells roughly 200 to 250 SKUs on Amazon and has not physically handled a product in years, because goods ship from the factory straight to Amazon’s distribution centers.

Greg is the founder of Jungle Scout, along with Fetcher, Splitly, and Jump Send. He sold his house, his cars, and his possessions about three years ago, and everything he owns fits in a backpack and a carry-on.

His physical products business takes three to four hours a week. A manager runs it, and the software company gets the rest of his attention.

This episode covers how he inspects samples without receiving them, the two-stage factory inspection process he uses, how he orders in container multiples, and which listing elements actually move conversions.

A note on timing: this is a 2017 conversation. Amazon has since restricted the aggressive coupon and giveaway tactics described in the launch section, so treat those specifics as historical.

Key takeaways

• Products ship from the factory directly to Amazon warehouses, so no inventory ever passes through his hands.
• A contact in Hong Kong records phone video of each sample for about $30, replacing physical sample review.
• Roughly 70% of inspection reports fail on the inspector’s standards, which is why during-production inspections matter more than sample checks.
• During-production inspections at 10% to 20% completion let the factory correct problems before the full run is built.
• Order quantities are set by what fits in a 20 or 40 foot container rather than by arbitrary unit counts.
• Deleting an unfavorable Amazon shipping plan and regenerating it days later can produce better warehouse assignments.
• Product criteria are 2,000 to 4,000 units of monthly demand across a niche with several top sellers under 50 reviews.
• Price and main image drive split test results, while bullet points and descriptions change almost nothing.

How do you inspect product samples without receiving them?

Have someone near the factory record video of the sample and describe what they find. Greg uses a contact in Hong Kong who can receive a sample the day after it ships from China.

The review is deliberately unscripted. They assemble the product, note anything that does not fit or line up, stress it physically, and comment on molding flaws or fabric weight.

No checklist gets handed over. Someone competent looking at a chair already knows to sit on it, stand on it, and drop it from a height.

The cost is about $30 per sample. It is a small operation, one person recording on a phone, found through a friend’s recommendation.

He no longer weights sample inspections heavily anyway. Anyone can produce one good sample, so what comes off the production line matters more.

How do you inspect factory production remotely?

Run two inspections through a third-party firm: one during production and one before the container loads. Greg uses Asia Inspection and finds them inexpensive and capable.

Roughly 70% of his inspection reports come back as failures against the inspector’s standards. The useful work is reading why, since grading distinguishes minor, major, and critical defects.

Timing is what makes the during-production inspection valuable. Scheduling it at 10% to 20% completion means a factory building 1,000 units gets corrected after about 100.

Early feedback also changes the negotiation. Showing a factory a failed report on 100 units makes fixing them straightforward, where a finished run leaves you arguing over discounts.

Remedies belong in the original agreement. Greg specifies in advance what happens on a failed inspection, whether a discount or a rework, and factories generally prefer fixing the units to granting a 20% discount.

Cost sets the threshold. A $200 inspection is obvious on a $30,000 order and hard to justify on a $2,000 one.

What is a poor man’s factory inspection?

Require your sales rep to photograph specific moments in production and write it into the contract. Greg did this exclusively for his first two years and had no serious problems.

The requirements are specific points in time. A photo on the assembly line, a photo of the first carton being loaded, and a photo before the container doors close.

It is biased and still useful. The rep chooses the shots, and obvious problems like the wrong color still jump out immediately.

How do you order inventory for oversized products?

Ask how many units fit in a 20 foot and a 40 foot container, then order in those multiples. Greg orders almost exclusively that way now.

Oversized products make this unavoidable. When only 300 or 400 units fit in a 40 foot container, a partial order makes little sense.

Shipping cost is the trap. Freight from China to the West Coast is well understood, and Amazon’s distribution assignments after that are where the surprises live.

Warehouse spread is what costs you. A shipment landing in Los Angeles and then splitting across three distant distribution centers costs considerably more than one going entirely to a nearby facility.

Shipping plans can be created far in advance, with no limit he has found. Creating them early gives you room to work.

An unfavorable plan can simply be deleted. Delete the entire plan, wait a couple of days, and generate a new one, and Greg has cycled through ten plans in a month to find a good assignment.

He uses Flexport for freight forwarding. The rates are not the cheapest, and the platform is convenient enough that the difference is not worth switching.

What product criteria should you use on Amazon?

Strong existing demand paired with the weakest competition you can find at that demand level. Greg wants at least 2,000 units of monthly demand across a niche, preferably 3,000 or 4,000.

That number validates the market. It confirms people are searching for the product category and buying it.

Review counts are his competition proxy. Reviews signal social proof and reflect the sales history that feeds ranking.

The specific bar is three or four sellers in the top results with fewer than 50 reviews. That combination suggests a niche you can enter.

His catalog has no brand coherence at all. A comb, a lamp, a chair, and a car bumper coexist because the data selected each one independently.

Oversized and higher-priced products are currently less competitive. Both carry a higher barrier to entry, which keeps the field thinner.

How did Amazon sellers launch products in 2017?

Deep discount coupons for two weeks, then aggressive PPC with a reduced price, then a slow climb back to target. This is the historical method, and Amazon has since restricted much of it.

The first phase set the list price at target while giving away coupons at roughly 75% off. Five to ten coupons a day for about two weeks produced early reviews and some organic sales.

Phase two swapped coupons for advertising. PPC bids went high enough to rank first for main terms, with the price cut from a $50 target to $35 or $40.

Sales velocity was the entire objective. Velocity was the dominant ranking input, so early advertising losses were accepted deliberately.

Phase three was patience. After two to four weeks of strong ranking, the price climbed gradually back toward target with 20 to 40 reviews accumulated.

The margin curve was planned as a loss. Months one and two lost money, month three ran 5% to 10% margins, and months four and five reached 25% to 30%.

What do you do when an Amazon product does not sell?

Push harder on price and advertising until it moves, or accept a loss and never reorder. Greg targets a minimum of 300 to 400 units a month and prefers 800 to 1,000.

Slow movers get aggressive treatment. More coupons, more PPC spend, and lower prices until velocity appears.

The outcome is binary by design. A product either sells at least 500 units a month or gets discontinued at break-even or a small loss.

He keeps nothing that trickles. A SKU selling 50 units a month does not stay in the catalog.

Which Amazon listing elements actually affect conversions?

Price and the main image, in that order. Split testing shows both moving results substantially.

Bullet points and descriptions are close to worthless in his testing. Titles help slightly.

The main image drives sessions rather than conversion directly. It is what pulls people onto the listing in the first place.

Price sensitivity is high in most niches. That is why price testing produces such clear results.

Secondary images are the next lever when sessions and price are already good. Lifestyle photography and a competitive matrix comparing your product against top sellers both help.

Feature callouts belong on the images themselves. Most shoppers never read the description, so anything they need to know should appear visually.

How do you diagnose a poorly converting Amazon listing?

Compare it against the competition as a neutral buyer would. Ask why someone would choose a competitor over your product.

Three questions cover most cases. Is the audience price sensitive, do competitors have materially better reviews, and are their images and listings simply better?

Poor keyword ranking is a separate diagnosis. If your main terms rank badly, listing optimization is not your problem.

The fix for ranking is velocity. More coupons and more aggressive PPC bids, accepting a worse ACOS in the short term.

Is enhanced brand content worth using?

Worth adding when the images already exist, and unlikely to be decisive. Greg puts it on every listing while acknowledging mixed results in his own testing.

Some listings showed a conversion improvement and others showed nothing. He has not seen it hurt.

Marginal cost is what justifies it. A photo shoot produces more images than the main listing needs, so populating it takes little extra work.

He also treats it as insurance. If Amazon ever charges for the feature, existing users may be grandfathered in.

How do you manage an Amazon business and a software company remotely?

Empower a manager and stay out of the details. Greg spends at most three or four hours a week on physical products and gives the rest to software.

Tooling is deliberately light. Slack for communication, Trello for project management and purchase orders, and Forecastly for inventory.

FBA removes most of the operational surface. With no merchant-fulfilled orders, there is no warehouse to supervise.

The software company was built remote from day one, now around 35 people. Communication and project management were designed for distributed work rather than retrofitted.

Hiring screens for it explicitly. They look for traits that predict success without an office, since some people genuinely work better with in-person accountability.

Coworking beats working from home in his view. The company reimburses coworking space, and almost nobody works from home.

Should you start a physical products business or a software company?

Physical products, unless you are a developer. Replacing a $5,000 to $8,000 monthly income is more reliably done through Amazon.

Development cost is what makes software hard to bootstrap. Hiring developers is expensive enough that most software companies raise money.

Greg is candid that his own path was unusual. Jungle Scout bootstrapped with good timing and product market fit, which is rare in that space.

Startup capital for private label is modest by comparison. Two to three thousand dollars is tight, $5,000 is workable, and $10,000 or more lets you grow reasonably fast.

How should a beginner start differently than an experienced seller?

Pick a less competitive niche and accept smaller numbers. A hundred units a month at $5 profit is $500, which is real progress toward quitting a job.

The economics of small niches favor beginners. Five reviews and 20 giveaway units can establish a listing without heavy advertising spend.

Overspending early is the risk to avoid. Losing a thousand dollars in month one on PPC is painful when your total budget is a few thousand.

The target profile is roughly 1,000 units of monthly niche demand. One seller with 100 reviews, most others at 20 to 40, and a few succeeding with very few reviews.

Shipping the first order to yourself is worth the extra step. Handling the product before it goes to Amazon is reassuring, and there is time to drop that habit later.

Do Amazon listing hijackers cause serious problems?

Less than seller communities suggest, especially with customized packaging. Generic products in plain bags are easy to duplicate, and distinctive boxes are not.

His products are largely protected by their own packaging. A hijacker cannot easily claim to have the same item when the box and product carry specific customizations.

Small incursions are not worth attention. Someone listing a single discounted unit costs one sale.

Serious cases are rare. People arriving with 300 units in stock have happened a handful of times and never became a real issue.

Frequently asked questions

Can you run an Amazon business without touching inventory?

Yes. Goods ship from the factory directly to Amazon’s warehouses, samples are inspected by video, and returns are disposed of without ever reaching you.

How much does a China factory inspection cost?

Around $200 through firms like Asia Inspection, which is easy to justify on a $30,000 order and hard to justify below about $2,000.

When should you schedule a factory inspection?

At 10% to 20% of production, so defects can be corrected on the remaining units, plus a final inspection before the container is loaded.

How many units should you order from a supplier?

Ask how many fit in a 20 foot and a 40 foot container and order in those multiples, particularly for oversized products where freight dominates the cost.

How do you get better Amazon warehouse assignments?

Delete an unfavorable shipping plan entirely and create a new one a few days later. Repeating this can produce assignments concentrated in fewer distribution centers.

What demand should a product niche have before you enter it?

At least 2,000 units of monthly demand across the niche, preferably 3,000 to 4,000, with several top-page sellers holding fewer than 50 reviews.

What matters most in an Amazon listing?

Price and the main image. Bullet points and descriptions showed almost no effect in split testing, so put key selling points directly on the images.

How much money do you need to start private label?

Two to three thousand dollars is difficult, $5,000 is manageable, and $10,000 or more allows meaningful growth early.