How to Lower Your Ecommerce COGS: An Operator Playbook

Most ecommerce founders treat cost of goods sold (COGS) like the weather: a fixed condition they react to rather than a system they control. That is the expensive mistake. COGS is not one number, it is a stack of five separate inputs, each with a different owner, a different lever, and a different payback window. When I talk to founders running a brand in the $3M to $50M range, the pattern is almost always the same: they know their blended COGS percentage to the decimal, but they cannot tell you which of the five components is the one running hot relative to their vertical. This post fixes that. We will define what COGS actually contains, benchmark it by vertical so you know what good looks like, and then sequence the fixes by how fast they pay back.

What counts as COGS (and what doesn't)

Before you can lower COGS, you have to draw the line in the right place. COGS is the direct cost of getting one sellable unit into your warehouse, ready to ship. That means five things: product cost (materials plus manufacturing or co-packer fees), inbound freight, customs duties and import taxes, inbound warehouse labor (receiving, put-away, QC), and inventory shrinkage and write-downs.

What is NOT in COGS matters just as much, because founders routinely misbuild this and then misdiagnose their problem. Outbound shipping to the customer, 3PL pick-and-pack, the box and dunnage you ship in, payment processing, returns processing, and all of marketing sit below the gross-margin line. Those are contribution-margin costs, not COGS. If you lump outbound shipping into COGS, you will chase the wrong lever and wonder why your "COGS reduction" did nothing for gross margin.

The reason the boundary is worth getting right: gross margin is the number your investors, your lenders, and your benchmark comparisons all live on. The pattern we see again and again is a founder who has been told their margin is "fine" because they are comparing a misbuilt COGS figure to a clean public benchmark. They are not comparing like with like. Get the five components clean first, then everything downstream gets easier to reason about.

Benchmark your vertical first: what does good look like?

You cannot tell if your COGS is too high until you know what your vertical's best operators run. The gross-margin spread across ecommerce verticals is enormous, and a number that is excellent in one category is a crisis in another. A 55% gross margin is healthy in apparel and alarming in supplements.

Here is where the public comps land. Supplements lead, with a median gross margin of 75.1% across five DTC pure-plays in their FY2025 10-Ks (LifeVantage 80.4%, USANA 78.3%, Herbalife 77.9%, Nature's Sunshine 72.4%, Medifast 71.3%). Beauty and skincare follow at roughly 69% brand-only median across seven public comps (e.l.f. 70.7%, Estee Lauder 74.0%, Olaplex 69.4%, Coty 64.8%). Apparel sits at a 55.3% median across eight comps, ranging from Gap at 40.8% to Ralph Lauren at 69.9%, with premium players like FIGS (67.6% gross margin in 2024, 32.4% cost of revenue) well above the pack. Beverage and food is the tightest at roughly 52% median, from Vita Coco at 36.5% to Coca-Cola at 61.6%.

The operator move is simple but most brands skip it: take your own gross margin, find your vertical's median, and measure the gap. If you are within a couple of points, your COGS work is fine-tuning. If you are 5 to 10 points worse than the median, you have a structural problem, and the rest of this post is for you. One caution we always give founders: public comps are large-scale businesses, so a $5M to $50M brand will sit structurally higher on per-unit cost at lower volume. Use the benchmark as a direction of travel, not an achievable floor next quarter. For the full vertical detail, see the Eightx apparel financial benchmark, beauty financial benchmark, supplements financial benchmark, and beverage financial benchmark.

The COGS stack: what to attack first

Once you know your vertical median and which way your number leans, the next question is which component to attack. The answer is set by two things: how big the component is, and how fast a fix pays back. You do not attack the most painful line, you attack the line where size and speed multiply.

Product cost is almost always the largest component, 25 to 40% of revenue, so a small percentage win there throws off the most dollars. But it is also the slowest to move, because real supplier change takes a negotiation cycle or a sourcing shift. Inbound freight and duties (3 to 10%) is smaller but the most volatile lever right now, because tariff exposure has made the duty line a moving target for many brands importing from overseas, particularly those with significant China sourcing. If a meaningful share of your COGS is duty, a sourcing review or HTS classification audit can move your number more than another round with your factory. We cover that exposure in the Eightx DTC tariff exposure index. Inbound warehouse labor (2 to 5%) and shrinkage (0.5 to 3%) are the smallest, so they are worth a look but rarely where you start.

The mistake we see most often is founders pouring a quarter of effort into shaving warehouse labor, which is 3% of revenue, while a 30%-of-revenue product cost sits 8 points above their vertical benchmark and untouched. Size the prize before you pick the fight. The biggest, most-fixable component wins your attention first, every time.

When I talk to founders this size, the relief is visible once they see the stack laid out, because the problem stops being "my margin is bad" and becomes "my product cost is 6 points high and my duty line doubled, here are the two things to fix." That is a solvable problem with owners and deadlines, not a vague anxiety.

Supplier negotiation: the playbook

Because product cost is the biggest component, supplier negotiation is where the real money lives. But "asking for a discount" is not a strategy, and most founders leave money on the table by negotiating on unit price alone. Here is the playbook that actually moves the number.

First, build a should-cost model for each hero SKU: break the unit into its material, labor, tooling, and margin components so you know roughly what it should cost to make, not just what you are paying. Second, get three to five competing quotes from alternative suppliers, even if you have no intention of switching, because a real alternative quote is the only bargaining power that survives the conversation. Third, consolidate volume and commit to it, because a 12-month volume commitment is the single biggest win you can hand a supplier in exchange for a better price. Fourth, negotiate on total landed cost, not unit price: payment terms, Incoterms, lead time, minimum order quantities, and pre-labeling all carry real dollars. The framing we coach is trade, never concede: every give from you (longer commitment, faster payment) should buy a get (lower price, better terms).

The math is concrete. Take a brand selling 10,000 units a month of a hero SKU at $5.00 landed. Move that to $4.40 by committing to double the volume on a 12-month agreement, and you save $0.60 per unit, which is $6,000 a month, or $72,000 a year, on a single SKU. That is an 8 to 12% reduction, and it is realistic inside a year for a brand that does the structured work. The pattern we see again and again is that the brands who win this are not the ones with the best relationships, they are the ones who showed up with a should-cost model and a credible competing quote.

Quick wins: packaging, freight, and SKU rationalization

While the supplier negotiation runs (it takes months), there are faster levers you should be pulling in parallel. These are the 30 to 180 day moves that improve your blended COGS without waiting on a factory.

Packaging right-sizing is the most overlooked. Carriers and freight forwarders price on dimensional weight, so an oversized box costs you on every inbound and outbound lane. Right-sizing three packaging formats typically yields a 5 to 15% material cost reduction and a 3 to 10% parcel cost reduction on the optimized lanes. Carrier renegotiation and mode shifts (ocean versus air on inbound, zone-skipping on outbound) are the other fast lever, usually 3 to 10% on the freight line in 30 to 90 days. And SKU rationalization, killing the bottom 10 to 20% of SKUs by contribution margin, lowers your carrying cost, simplifies your buys, and lets you consolidate volume onto the SKUs where supplier negotiation actually matters.

The worked example below sizes all of this for a $3M revenue brand carrying roughly $1.2M of COGS (about 40% COGS). When we run this for operators, the realistic addressable saving in the first 12 months lands at $56,000 to $88,000, and the packaging and carrier pieces start paying back inside the first quarter.

The COGS reduction roadmap: 0 to 36 months

Put it all together and you get a sequenced roadmap, ordered by payback window so you bank cash early and reinvest the momentum into the slower, structural moves.

In the first 0 to 6 months, pull every fast lever: carrier renegotiation, packaging right-sizing, a SKU pause or cull on the worst performers, and the opening rounds of supplier conversations. These are low difficulty and pay back inside a quarter. In the 6 to 18 month window, run the structural mid-term work: a full supplier RFP on your hero SKUs, a 3PL re-bid, and light value engineering or material substitution where it does not touch the customer experience. From 18 to 36 months, take on the genuinely structural shifts: nearshoring or a sourcing-country change to manage duty exposure, and the demand-forecasting and ERP investments that cut shrinkage and over-buying at the root.

One thing every founder should keep in front of them while doing this: each saved point of COGS is worth more in 2026 than it was three years ago. CAC has risen 40 to 60% between 2023 and 2025, which means the contribution-margin stack below your gross line is more expensive than ever, so margin you protect at the COGS level is margin you are not bleeding back out on acquisition. COGS reduction is necessary but it is not sufficient, and the brands that win treat it as one disciplined workstream inside the whole unit-economics picture. If you want this run against your own P&L by someone who has sized these levers across dozens of brands, that is the work we do as a fractional CFO.

Sources and methodology

The vertical gross-margin benchmarks are built from public-company 10-K filings via SEC EDGAR, cross-referenced against the Eightx benchmark series. The supplements median of 75.1% comes from five DTC pure-plays in their FY2025 10-Ks (LifeVantage, USANA, Herbalife, Nature's Sunshine, Medifast). The beauty median of roughly 69% is a brand-only median across seven public comps including e.l.f., Estee Lauder, Olaplex and Coty. The apparel median of 55.3% spans eight comps from Gap (40.8%) to Ralph Lauren (69.9%).

FIGS figures are taken directly from the FIGS 2024 Annual Report (Form 10-K): gross margin of 67.6% in 2024 and 69.1% in 2023, with cost of revenue of 32.4% and 30.9% respectively. The 2024 decline was attributed to product mix and higher order volume, not raw material cost inflation, which is a reminder that mix moves COGS as much as unit price does. The beverage median of roughly 52% spans six FY2025 comps from Vita Coco (36.5%) to Coca-Cola (61.6%).

The COGS component breakdown (product cost 25 to 40%, freight and duties 3 to 10%, inbound labor 2 to 5%, shrinkage 0.5 to 3%) is a synthesis of Perplexity-sourced industry research (topgrowthmarketing.com unit economics, finaloop.com COGS glossary) and Eightx internal unit-economics data. The lever impact estimates (8 to 12% supplier reduction, 5 to 15% packaging material reduction, 3 to 10% parcel reduction) come from operator benchmarking synthesized across mds.co, sievo.com and sellercloud.com, plus Eightx engagement data.

A few limitations to read the numbers honestly. Public comps are large-scale businesses, so a $5M to $50M DTC brand will run structurally higher per-unit cost at lower volume; treat the benchmark as direction of travel, not an achievable floor. The 8 to 12% supplier reduction is a 12-month target requiring structured, multi-round negotiation with real alternative quotes, not a single-PO outcome. Operator-voice lines throughout are synthesized from anonymized founder-call patterns, with no client named.

Frequently asked questions