368: What Successful Amazon Sellers Do Differently To Grow With Jeff Cohen

Podcast: Download (Duration: 59:30 — 68.4MB)

Managing an Amazon seller account well comes down to segmenting your catalog into three tiers, setting specific goals at the SKU level rather than the account level, and building an audit schedule so each part of the business gets checked at the right frequency. Without that structure you spend mornings reacting to problems on products that do not matter.

In this episode I brought Jeff Cohen back on the show. He is VP of Marketing at Seller Labs and one of my go-to people for keeping up with what is actually changing on Amazon.

Below is the full system: how to set goals that are actionable, how to split your catalog, what to monitor and how often, the unintended consequences trap, and which Amazon features are worth your time.

Key takeaways

• “Grow from $1M to $2M” is not a goal. It tells you nothing about what to do tomorrow.
• Split your catalog into three tiers and give each tier a different strategy.
• Think of yourself as an underperforming version of your target size, which reveals gaps.
• Audit at different frequencies. Ads need several times a week, listings need quarterly.
• Two things usually cause a sudden sales drop: a negative review or a competitor price change.
• Watch for unintended consequences. Lowering ACoS can lower sales and profit.
• Sometimes you lower ACoS by spending more. Better placement improves every downstream metric.
• Images sell, not bullet points. Nobody reads bullets on a simple product.

Why most Amazon sales goals are useless

A goal like growing from $1 million to $2 million tells you nothing about what to do, because it does not specify where the growth comes from.

The useful version breaks it down. Growth comes from increasing traffic, improving conversion, or adding SKUs, so the goal has to name which.

Jeff’s worked example: five SKUs generating $100,000 a month, which you believe you can improve 20% to reach $120,000. That leaves $80,000 unaccounted for, which means new SKUs or rehabilitating a weak product in your existing catalog.

The categories worth setting goals around: monthly sales, profit level, ad spend, target ratio of paid to organic sales, and review count.

His framing on why this matters: without a goal, a plan is just a wish.

The mindset shift that reveals your gaps

Think of yourself as an underperforming version of the business you want to be rather than a smaller business trying to grow.

The difference is real. A half-million dollar brand trying to double looks at growth tactics. An underperforming million-dollar brand looks for what is broken.

The second framing produces a gap analysis, which is what actually tells you where the products, profit, or targets will come from.

How to segment your Amazon catalog

Split your catalog into three sections using the Pareto principle, then apply a different strategy to each.

The head. Roughly 20% of SKUs generating about 80% of sales. These are the ones you cannot afford to run out of, because a top SKU that falls is genuinely hard to restore and hurts the most.

The middle. Around 70% of your catalog. Every item here gets either a green arrow, meaning you are working to move it into the head, or a red arrow, meaning it is on its way out and deserves less of your time.

The tail. Products you should stop selling.

Applying one strategy to an entire catalog fails. Grouping products that behave similarly and giving each group its own plan is what works.

The middle tier deserves nuance. Some products are highly profitable at low velocity, which is fine. Others sit there because of stock problems, which means fixing the supply chain might promote them.

Why catalog segmentation saves your mornings

Knowing which SKUs matter tells you what to ignore when something breaks.

The failure mode: you wake up, see a SKU suspended or sales dropped, and spend half the morning fixing something with no material impact on the business.

If you already decided that SKU is in the red-arrow group, the drop is expected and you move on with your day intact.

Jeff frames this through outputs versus outcomes. The number of posts made, ads adjusted, or keywords optimized are outputs. If they do not drive the outcome you defined, working on them is not progress.

How to build an Amazon audit schedule

Different parts of the business need checking at different frequencies, so assign each one a cadence rather than treating everything as daily.

Advertising: several times a week. You cannot set it and forget it. It is an investment that only returns if you keep adjusting, whether you do it yourself, use a tool, or hire a service.

Listings: roughly quarterly. Check your category listing report, refresh A+ content, and update images. Automation should watch it in between so you get notified if a title or bullet changes unexpectedly.

Inventory: ongoing. IPI score, long-term storage fees, lost inventory and reimbursements Amazon owes you, and whether a promotion is needed to move stock.

Fees: on change. Amazon re-measuring your product can change its fee tier and destroy profitability without warning.

Reviews and feedback: ongoing, plus a process for acting on what they tell you about the listing or your operations.

Traffic and placements: Amazon Posts follower counts, Stores page traffic, and the newer ad placements including video, sponsored brands, and display.

Alerts versus indicators

Alerts by email become noise, which is why Seller Labs pairs them with visual indicators inside the dashboard.

Their customer survey found something counterintuitive: sellers said alerts were not very important to them, while recommendations and suggestions for improving the business were.

The system Jeff describes has three parts, and most tools only handle the first. There is the alert that something happened, the information explaining what happened, and the action you should take.

The indicator approach uses red and green comparisons across a chosen window, showing seven, thirty, or sixty day performance against the prior equivalent period.

The seasonality trap is worth knowing. Looking at a sixty day window in February compares you to December, so every arrow will be red regardless of how the business is doing.

My own preference is year-over-year comparison on the same day of the week, which handles seasonality cleanly. It made March 2021 look excellent against a March 2020 that was cut in half.

What causes a sudden drop in Amazon sales

Two things account for most immediate sales drops: a new negative review, or a competitor changing their price.

Jeff’s own recent example was the second. A competitor cut their price by nearly 50%, possibly to liquidate inventory before long-term storage fees, which immediately hit his sessions and advertising performance.

His response was to take no action, which is a decision rather than a failure to decide. He did not want to compete on price, and three days later the competitor restored theirs.

The cost was real. His BSR dropped with the sales, so the listing had to rebuild.

The reason he caught it immediately is that he tracks monthly pacing against a monthly goal. Sessions dropping told him traffic fell, and investigating told him why.

The unintended consequences trap

Optimizing hard for one metric frequently damages another, and ACoS is the clearest example.

If your stated goal is lowering ACoS, Jeff’s question is whether that is really the goal or whether you actually want to reduce advertising’s drag on profitable sales. He can lower your ACoS and reduce your sales volume and overall profit doing it.

The genuinely counterintuitive case is that lowering ACoS sometimes requires spending more. Reaching a better position on the page or a higher conversion rate can flip every downstream metric, so a product unprofitable at 8% ACoS becomes more profitable at 10% because organic position improved and total sales rose.

The protection is the pivot-or-persevere discipline from The Lean Startup: build, measure, learn, then either change direction or double down.

Where people fail is choosing a direction without defining the metric that will tell them whether it worked, which leaves no way to close the loop.

Placement reports are the concrete tool here. Comparing conversion rate by placement can show mathematically that one placement deserves a higher bid than another because it converts twice as well.

What to automate and what to keep manual

Goals stay manual and monitoring gets automated, because goals are the one thing no tool can supply.

You are responsible for setting goals and building plans for your products. Software handles the monitoring, dashboards, and gap analysis that tell you where to spend time.

Jeff is direct about the limits of automation in advertising. A tool can build a fairly sophisticated campaign, and Amazon’s ad platform now has so many options and variables that no tool can capture every input.

The dividing line by intent: if you want to maintain a certain sales level and ACoS, AI can run that. If you want to use advertising to actively grow the business, you need human judgment for decisions like when to start testing video ads or bid harder on a specific placement.

His honest sales filter reflects this. When someone asks whether the tool can help them grow sales, the answer is no. When someone says they want to reduce ACoS on a specific set of SKUs from 10% to 8%, that is a goal a tool can work toward.

Why video is the highest-value Amazon feature right now

Video ads are producing strong results and still carry a first-mover advantage because most sellers have not started.

Video ads require brand registry and appear roughly a third of the way down search results, plus mobile placements, with Amazon adding more.

Production value is not the barrier. Jeff’s t-shirt client simply films someone wearing the shirt turning in a circle so buyers can see the fit.

Video on the product detail page matters separately from ads, and is no longer restricted to large sellers. It uploads through Seller Central under inventory, which is an odd location, and also appears in video shorts.

Prioritize by traffic. Sort your products by sessions, either in a Seller Central report or a tool, and build video for your highest-session products first.

How Amazon Posts work

Posts appear mostly on mobile and are worth doing once you have spare capacity rather than when starting out.

The competitive argument is direct: if your competitors are not posting, your products appear on their posts.

What performs is lifestyle imagery and genuine product-in-use photography, the kind of image you would want to see in a customer review, rather than infographics.

The constraint to plan around is that images cannot be repeated, so Amazon wants new content every time. Sit down and plan a batch monthly.

Expectations vary by category. One seller may be pleased with a thousand views and another may see ten thousand, so experimentation is required to know what it is worth to you.

Jeff’s priority ranking is clear. If you are just starting, skip Posts entirely because they will not move the needle. Video ranks higher because it affects engagement and conversion on the detail page directly.

Why images matter more than bullet points

Listing copy mainly feeds the organic search algorithm, and images are what actually sell the product.

Unless the product is technically complex, people are not reading your bullets. They are looking at every image.

The behavior splits by purchase size. Buying a ping pong table means reading everything and multiple reviews. Buying a coaster or a silicone ice tray means looking at pictures.

The practical conclusion is that your main selling points have to appear inside the images, frequently as text on the image itself, or buyers will not understand your product or your value.

Jeff’s checkpoint: look at your own detail pages and your competitors’ pages on desktop, mobile, and tablet. A surprising number of sellers have never viewed their own listing on a phone.

The state of counterfeits and brand protection

Counterfeiting is a cat and mouse game Amazon will never fully win, because the money involved is too large.

Brand registry has genuinely improved, with better data intake around design patents, patents, and trademarks to identify approved sellers.

One metric worth watching is Amazon’s price competitiveness score, which compares your Amazon price to your price elsewhere. Brands frequently have trouble here because knockoffs undercut them on other platforms.

Transparency, which puts a unique per-unit barcode on your product for a small fee, works well for sellers using it. The implementation catch is timing: it is easy on a new SKU and difficult on existing inventory, since you effectively need to reach zero stock before restocking with transparency units.

Why inventory management became the hardest part

Just-in-time inventory is now mandatory on Amazon, which is a complexity sellers did not face two years ago.

The conditions driving it: backed-up docks, longer shipping times, rising shipping costs, IPI score limits, and inventory restrictions on new products.

Receiving times illustrate the shift. Amazon took about a week to receive and stock products before the pandemic, and three to five weeks after.

The new-SKU trap is severe. Jeff describes a seller planning to sell 200 units a day on a new SKU who was capped at 200 units of inventory total, which makes the plan impossible.

The workaround for a known Q4 product is getting inventory in early and building the allowance up over time.

The structural conclusion: you can launch on FBA alone, and growing a real business now requires third-party logistics. That is a good problem, since an owned fulfillment path is worth having regardless.

FBA remains essential despite all of this. Jeff saw a product’s sales more than double moving into FBA with no other change, purely from the Prime badge.

Seller Fulfilled Prime is effectively closed. It has been shut to new applicants for over two years, and Amazon has not made third-party logistics providers eligible, so historically only sellers with their own warehouses qualified.

Frequently asked questions

How should you manage an Amazon seller account?

Segment your catalog into three tiers by sales contribution, set specific goals at the SKU level rather than the account level, and build an audit schedule assigning each area a checking frequency. Advertising needs attention several times a week, listings roughly quarterly.

What makes a good Amazon sales goal?

One that specifies where growth comes from. “Grow from $1 million to $2 million” is not actionable. Naming which SKUs will improve by how much, how many new SKUs you will add, and what ad spend and profit level you are targeting is.

How do you segment an Amazon catalog?

Into three tiers. The head is roughly 20% of SKUs producing 80% of sales, which you must never let run out of stock. The middle is about 70%, each item marked for promotion or removal. The tail is products to stop selling.

Why did my Amazon sales suddenly drop?

Usually a new negative review or a competitor changing their price. Check reviews first, then competitor pricing. A competitor cutting price by 50% will hit your sessions and advertising performance immediately, even if they restore it days later.

Should you always try to lower your ACoS?

No. Lowering ACoS can reduce sales volume and total profit. Sometimes lowering it requires spending more, because better page placement improves conversion and organic position enough that a product unprofitable at 8% ACoS becomes profitable at 10%.

Are Amazon video ads worth it?

Yes, and they still carry a first-mover advantage since most sellers have not started. Brand registry is required. Production value is not the barrier, since simply filming a product in use works. Prioritize video for your highest-session products first.

Do Amazon bullet points matter?

Mainly for the search algorithm. Buyers rarely read bullets on simple products and look at every image instead. Put your main selling points into the images themselves, and check how your listing appears on mobile as well as desktop.