Stop Guessing Your Catalog: A Simple Test‑Keep‑Retire Framework for Independent Ecommerce Brands

Most independent ecommerce founders don’t have a product problem. They have a visibility problem.

You’re surrounded by dashboards, ad reports, and marketplace metrics, but on any given Tuesday it’s still hard to answer three basic questions:

• Which products are quietly carrying the business?
• Which ones are stuck but fixable?
• Which ones are slowly draining cash, attention, and warehouse space?

When you can’t see those answers clearly, you end up doing what everyone does under pressure: launch another product, chase another trend, or discount whatever feels slow this week. The catalog gets noisier, ad spend gets harder to manage, and your team spends more time reacting than deciding.

This article offers a different path: a simple, operator-level framework that turns your catalog into three clear lanes—test, keep, retire—so you can make calmer, more honest decisions about what stays, what gets a second chance, and what needs to go.

Why assortment decisions feel harder than they should

Independent ecommerce brands live in a strange middle ground. You’re not a marketplace-only seller with one product, and you’re not a big-box retailer with a full-time merchandising team. You’re running a real business with dozens or hundreds of SKUs, but the people making assortment calls are the same people handling customer emails, ad tweaks, and supplier issues.

That creates three predictable problems:

• Everything feels urgent, nothing feels comparable. You remember the last angry email or the last big order, not the full pattern.
• Metrics are scattered. Ads live in one system, returns in another, reviews somewhere else, and margin in a spreadsheet that only one person really understands.
• Emotional attachment clouds judgment. The product you love, the one your designer fought for, or the one that “should” work often gets more chances than the numbers justify.

Without a simple, shared way to classify products, every assortment conversation becomes a debate. The goal of the test‑keep‑retire framework is to replace those debates with a calm, repeatable decision rhythm.

The test‑keep‑retire framework in plain language

At its core, the framework is just a disciplined way to put every product into one of three lanes:

• Test: Products that are still in learning mode. You’re deliberately running experiments on price, positioning, creative, or channel to see if they deserve a permanent spot.
• Keep: Products that have earned their place. They pull their weight on margin, volume, or strategic value and don’t need constant debate.
• Retire: Products that no longer justify their space, attention, or working capital—and need a clear, respectful exit plan.

The power of this framework isn’t in the labels. It’s in the rules you attach to each lane and the weekly rhythm you use to move products between them.

Step 1: Build a simple, visible board

Start with something you can see in one glance. That might be a physical board in your studio or a simple digital board in your project tool. The key is that it’s not buried in a spreadsheet only one person opens.

Create three columns: Test, Keep, and Retire. For each product, add a card with:

• Product name and SKU
• Primary job (e.g., hero, entry offer, bundle builder, margin stabilizer)
• Current lane (test, keep, retire)
• One or two key metrics you’ll use to judge it (for example: contribution margin, 60‑day repeat rate, return rate, or attach rate in carts)

Don’t overbuild this. The goal is to make the state of the catalog obvious to anyone who walks by, not to recreate your BI stack on the wall.

Step 2: Define clear rules for each lane

Next, decide what it actually means for a product to live in each lane. Without rules, the board becomes decoration.

For Test products, set expectations like:

• Time box: “New products live in Test for 60–90 days.”
• Experiment count: “Each Test product gets at least two deliberate experiments—such as a new hero image, a different bundle, or a targeted email feature—before we judge it.”
• Minimum data: “We don’t make a final call until we’ve seen at least X sessions or Y orders.”

For Keep products, define what “earning their spot” means:

• “Keeps must hit a minimum contribution margin over the last 90 days.”
• “Keeps either drive new customer acquisition or support high‑value repeat behavior.”
• “Keeps don’t create outsized operational pain—no constant returns, fragile packaging, or support tickets that drain the team.”

For Retire products, be explicit about triggers:

• “If a Test product misses its targets after two experiments, it moves to Retire.”
• “If a Keep product’s margin or repeat behavior drops below our line for two review cycles, it moves to Retire or back to Test with a specific fix.”
• “If a product creates more operational friction than it’s worth—damages, returns, support load—it’s a retire candidate even if revenue looks okay.”

These rules don’t have to be perfect on day one. They just have to be clear enough that your team can apply them consistently.

Step 3: Attach honest numbers to each card

Once the lanes and rules are defined, attach a small set of numbers to each product card. The point is not to track everything—it’s to track the few things that actually change your decision.

For most independent ecommerce brands, a practical starting set looks like:

• Contribution margin: After discounts, returns, and typical shipping costs.
• 60‑ or 90‑day repeat rate: How often buyers of this product come back for anything, not just the same SKU.
• Return or complaint rate: Any signal that the product is quietly eroding trust.
• Ad efficiency (if applicable): Roughly how much you spend to get a new customer through this product.

If you already have dashboards, pull these numbers into a simple weekly export. If you don’t, start with rough but consistent estimates. The goal is comparability, not perfection.

Step 4: Run a short weekly assortment review

The framework only works if it lives inside your week. That means a short, protected review—30 to 45 minutes—where you and one or two key people walk the board and make decisions.

In that review, ask:

• “Which Test products have enough data to judge?”
• “Which Keeps still clearly earn their spot?”
• “Which products are on the edge and need one more deliberate experiment?”
• “Which products are quietly draining cash or attention and should move to Retire?”

Move cards deliberately. When you move a product from Test to Keep, write down why. When you move something to Retire, note the trigger—margin, repeat behavior, operational pain, or a combination.

Over time, this review becomes less about arguing and more about calmly applying the rules you already agreed on.

Step 5: Design respectful exit paths for Retire products

Retiring a product doesn’t mean dumping it overnight or pretending it never existed. It means giving it a clear, respectful exit that protects cash, brand, and customer trust.

For each Retire product, decide:

• Whether to run a quiet close‑out offer to existing customers who already like the product.
• Whether to bundle remaining units with stronger products as a value add.
• Whether to keep a small quantity for specific use cases (for example, VIP gifts or limited‑time drops) while removing it from the main catalog.

The key is that you’re not letting the product linger in a gray zone where it still consumes attention, ad spend, or warehouse space without a clear job.

Step 6: Let AI support the framework, not replace it

AI can help here—but only if it serves the framework instead of trying to be the framework.

Practical uses that fit a small brand’s week include:

• Summarizing reviews for a single product into a few clear themes you can add to its card.
• Suggesting alternative positioning or bundle ideas for Test products that are close to working.
• Highlighting products whose metrics have quietly shifted since the last review.

What you don’t need is a black‑box model making assortment calls you don’t understand. The board and the weekly review keep humans in charge; AI just helps you see patterns faster.

Step 7: Protect the board from constant exceptions

As the framework takes hold, you’ll feel pressure to make exceptions—especially for products you’re emotionally attached to or that had one big win in the past.

Protect the board by agreeing on a few guardrails:

• “We don’t move a product back from Retire to Test without a specific, testable hypothesis.”
• “We don’t keep a product in Test indefinitely; it either earns its way into Keep or exits.”
• “We don’t override the rules in the middle of a bad week; we wait for the next scheduled review.”

These guardrails keep the framework from turning into another set of suggestions you ignore when things get noisy.

What changes when you run this way

When independent ecommerce brands adopt a simple test‑keep‑retire framework and actually run it every week, a few things tend to happen:

• The catalog feels lighter. You carry fewer products that quietly drain energy.
• Marketing gets clearer. You know which products deserve attention and which are experiments, so campaigns stop feeling random.
• Supplier conversations improve. You can point to concrete data when you ask for better terms, exclusives, or support.
• The team argues less. Decisions move from “I feel” to “We agreed on these rules, and this product is on the wrong side of them.”

Most importantly, you get your attention back. Instead of reacting to every slow day or hot trend, you’re running a calm, visible system that helps you see which products truly support the business you’re trying to build.

You don’t need a giant merchandising department to make better assortment decisions. You need a simple board, a few honest numbers, and a weekly habit of moving products between test, keep, and retire with your eyes open.