Good product research filters reject bad buys quickly, so you only spend time on real candidates. This article lists the screening criteria and thresholds that experienced online arbitrage, retail arbitrage and wholesale sellers use, explains where each number comes from, and covers the mistakes that slowly eat margins.
Disclosure: I build WallScout, a Walmart research tool. The benchmarks below come from common seller practice, not from any one product.
Why written thresholds work better than gut feel
New sellers get stuck: millions of products, no rules, and hours gone. The fix is to write your criteria down once and apply them mechanically, so a product either clears the bar or doesn't. A filter-based product finder automates exactly this. You enter your thresholds and it only shows what qualifies.
Here are the criteria that matter, roughly in the order to apply them.
The core screening criteria
| Criterion | Common threshold | Why it matters |
|---|---|---|
| Minimum net profit / unit | $3 (many go higher) | Covers returns, price drops and estimate error |
| Minimum ROI | 30% floor; 50%+ for RA finds | Below 20% rarely worth the work |
| Maximum seller count | 15 (sweet spot 3 to 15) | Above this the Buy Box gets split and prices collapse |
| Minimum monthly velocity | ~10 units (slow) to 100s (deep stock) | No margin saves a product nobody buys |
| Price range | Often $15 to $50 | Staying above $10 avoids the WFS sub-$10 surcharge and limits cash at risk |
| Review count | Low reviews + real demand = opportunity | Lots of reviews can mean entrenched competition |
| Rating | Watch listings under ~3.5 stars | Poor ratings often mean returns and complaints |
| Weight / size | Prefer light, small | Drives WFS fulfillment cost |
Net profit and ROI
The pair of benchmarks sellers cite most is 30% ROI and $3 net profit per unit as a minimum. You need both: 40% ROI on a $2 profit is still fragile, and a $6 profit at 12% ROI ties up too much cash for what it returns. Retail arbitrage sellers often want 50% ROI or more, because they buy in small, opportunistic lots.
Calculate both after real 2026 Walmart fees, never off the raw spread. That means the category referral fee (6% to 20%, most often 15%), WFS fulfillment (from $3.45 per unit, plus a surcharge under $10) and monthly storage. The full tables are in Walmart seller fees 2026.
Seller count
Everyone on a listing shares its demand. A listing doing 200 units a month across 25 sellers may get you fewer sales than one doing 50 units with three sellers. The common range is 3 to 15 sellers: enough to show the product sells, few enough that you can realistically hold the Buy Box some of the time. Above 15, most experienced sellers move on. (See how to win the Walmart Buy Box.)
Velocity
Set a minimum estimated monthly units figure based on how much you plan to stock. For a cautious buy of a few units, 10 a month can be enough; for a product you want to reorder aggressively, ask for 100+. Be wary of velocity numbers with no data behind them (more on that below). Our guide on Walmart units sold per month explains how to read velocity.
Set your thresholds once and let the database apply them. WallScout's Product Finder only shows Walmart products that pass your ROI, profit, seller-count and velocity filters. Try it at wallscout.io.
Adjust the filters to your business model
The same product can be a great wholesale buy and a terrible arbitrage buy.
For online arbitrage, put repeatable sourcing and 30%+ ROI first, and expect thinner net margins after all fees (often 10 to 20%) than the headline numbers suggest. Retail arbitrage needs higher ROI, 50% or more, to pay for the time spent sourcing in stores, and light, easy-to-ship items help. In wholesale, volume and the ability to reorder matter more than the ROI on any single unit, and seller count and Buy Box rotation become critical. For private label, filter for products with solid sales and some weakness you can beat, like low review counts or poor listings.
If you haven't settled on a fulfillment method yet, read FBA vs WFS and selling on Amazon vs Walmart first.
Screening mistakes that cost money
Even with good thresholds, sellers get burned by process errors:
- Calculating ROI off the spread. The gap between cost and price ignores fees. A $20 buy that "sells for $35" can net closer to $7 after referral, fulfillment and shipping.
- Trusting velocity estimates with no data behind them. On a brand-new listing, estimated units are a guess. That's why WallScout shows no estimate with fewer than two data points, badges two to seven as "limited data" and only gives a full estimate at eight or more. It also calculates monthly units from the four most recent weeks times 4.3, so the number reflects current velocity instead of a stale average.
- Forgetting the sub-$10 WFS surcharge. The extra fee on items under $10 can turn a cheap "profitable" product into a loss.
- Over-tightening filters. Set 12 criteria at their strictest and you'll get zero results, or a few outliers that only got through because their data was thin.
- Filtering for demand without filtering for competition. High units plus a high seller count usually means a price war.
- Skipping restriction and sourcing checks. A product that passes every numeric filter is useless if you can't buy it again or aren't approved to sell it.
Save your criteria as a filter
Once your thresholds are written down, you can reuse them. Most serious research tools, including Jungle Scout's Product Database, Helium 10's Black Box and WallScout's Product Finder, let you save filter presets and rerun them every week in seconds. Make one preset per business model and one per category you specialize in.
Frequently asked questions
What ROI should I look for when sourcing products for Walmart? Most experienced sellers want at least 30% ROI after all Walmart fees, and retail arbitrage sellers often aim for 50% or more on individual finds. Below 20%, there's rarely enough margin to absorb returns, price swings and estimate error. Always calculate ROI on net profit after fees, not on the raw price difference.
How many sellers on a Walmart listing is too many? About 3 to 15 sellers is generally considered healthy. Past 15, the Buy Box is split among too many competitors and prices erode, so many sellers cap their seller-count filter at 15.
What filters should I use to find products on Walmart? The most useful are category, price range (often $15 to $50), estimated monthly units (your demand floor), seller count (your competition ceiling, often capped at 15), review count and rating, and weight or size (which drives WFS fulfillment cost). Start loose, look at the results, then tighten.
What's the most common product research mistake on Walmart? Calculating profit from the raw price spread instead of net profit after fees. Walmart's referral fee (6% to 20%, usually 15%), WFS fulfillment fees (from $3.45 per unit, plus a surcharge under $10) and storage can turn a product that looks profitable into a loss. Trusting sales estimates built on too little data comes a close second.
Save your screening criteria as a filter and let the database apply them. Start free at wallscout.io.