By Diosh — Founder, AHAeCommerce | eCommerce decision intelligence for $50K–$5M GMV operators
This is a decision piece for operators sitting on $1M–$5M GMV who are about to walk into a first institutional raise believing GMV, growth rate, and gross margin are what investors care about. They are not — those three numbers get you the first meeting and lose you the second. Real diligence is built on eight numbers, and the one that kills most DTC raises is not on the headline tear sheet. The decision this article forces: do you build the 8-number investor dashboard before you pitch anyone, or do you find out in the data room that your cohort retention has been decaying for four straight quarters?
Why the Headline Three Are Not Enough
Every operator I have watched stumble through a first raise opened the same way. GMV. Year-over-year growth. Blended gross margin. Three numbers, one slide, big font. That deck gets a meeting because it has to — partners triage hundreds of decks a week and that is the standard format. But the partner is not deciding to invest off that slide. The partner is deciding whether to spend two hours pulling your real numbers.
The structural problem is that GMV is a count of transactions, not a measure of business quality. Andreessen Horowitz spelled this out a decade ago and the framing is still binding: GMV reflects the size of the activity on your platform, not the cash that ends up with you (a16z — 16 More Startup Metrics). A $4M GMV apparel brand returning 38% of units, paying 22% to Meta, and bleeding 9 days of free Shopify Capital float is not a $4M business. It is a $1.1M contribution business with a working capital problem. Investors learn this in week one of diligence. You should learn it before the first pitch.
The other failure mode is timing. Headline gross margin reads beautifully when you average the last 12 months. It hides the fact that your last two quarters dropped 380 basis points because you raised ad spend 60% and now you are buying customers who only convert with 15% discount codes. Diligence pulls the trailing 90 days, not the trailing 12. If your trailing 90 contradicts your headline, the conversation ends — not because the trailing 90 is bad, but because you did not flag it yourself.
The eight numbers below are the ones that get pulled. Build the dashboard before the meeting. If you cannot extract these eight in under two hours from your current stack, your data infrastructure costs more than your raise will be worth.
Number 1 — Contribution Margin (Not Gross Margin)
Gross margin is COGS-only. Contribution margin is what is left after every variable cost of selling one more unit: COGS, payment processing, fulfillment labor, packaging, outbound shipping, returns, marketplace fees, and the marginal ad spend tied to that unit. On a $48 ASP DTC apparel order, gross margin might read 62%. Contribution margin — after 2.9% Stripe, $6.40 pick-pack-ship, $1.20 packaging, a $4.80 return reserve, and an $11 blended CAC allocation — is usually 18–24%.
Investors pull contribution because it is the only margin that tells them whether scale helps you or hurts you. A business with 65% gross margin and 8% contribution margin gets worse with growth: every new order eats fixed marketing cost without earning enough variable spread to cover it. A business with 52% gross margin and 28% contribution margin gets better with growth. That second business raises capital. The first does not, even though the gross margin slide looks cleaner.
The trap operators fall into is presenting blended contribution. Blend hides the SKU-level disasters. Pull contribution by SKU, sort descending, and show your top 20% — the partner will ask anyway, and if the top 20% drives 95% of contribution, that is a concentration risk worth flagging before they find it. The framework for separating these two margins is covered in depth in Contribution Margin vs Gross Margin, which is the prerequisite reading before you build the dashboard.
Number 2 — CAC Payback in Months (Not the Absolute Number)
Absolute CAC is meaningless without payback context. A $42 CAC sounds high until you realize the customer pays back in 2.1 months on contribution. A $14 CAC sounds cheap until you realize first-purchase contribution is $9 and the customer never reorders — payback never happens.
The benchmark investors triangulate against is the Bessemer Cloud framework: SMB-focused businesses should target CAC payback under 12 months, mid-market under 18, enterprise under 24 (Bessemer Venture Partners — Cloud Metrics). DTC is its own animal — SaaS frameworks are reference points, not law — but the shape of investor scrutiny is identical. Top quartile DTC sits at 4–7 months payback on contribution. Median is 9–12. Above 14 months and you are selling a brand story, not a unit economic story.
The number to put in front of an investor is "CAC payback, contribution basis, trailing 90 days" — explicitly trailing 90, not trailing 12. The trailing 12 hides what happened to your CAC when you scaled the ad budget last quarter. The full mechanics of the calculation, including the common error of using revenue payback instead of contribution payback, are in Customer Acquisition Cost Truth.
Number 3 — Repeat Purchase Rate at 90, 180, 360 Days
A single repeat rate is the most-cited and least-useful retention metric on the operator side. Investors do not want "32% repeat purchase rate." They want three numbers: what percentage of a cohort places a second order within 90 days, 180 days, and 360 days.
The three-checkpoint structure exists because it diagnoses different problems. Weak 90-day repeat (under 12% for apparel, under 18% for consumables, under 25% for skincare) signals a product-fit problem — people bought, did not love it enough to come back, and you are paying CAC to fill a leaky bucket. Strong 90 but weak 180 signals a merchandising problem — you are not giving the second-purchase customer a reason to return. Strong 90 and 180 but weak 360 signals a category problem — you are in a low-frequency category and need to be modeled differently, with longer payback assumptions and higher AOV expansion.
The a16z retention framing applies cleanly: investors plot cohorts and look for the curve to flatten (a16z — 16 Startup Metrics). A flat curve at month 6+ means you have a real customer base. A curve that keeps declining through month 12 means you do not have customers, you have transactions.
Number 4 — Cohort Revenue Retention Curves (the Number That Kills Most Raises)
This is the metric headline GMV cannot mask. Plot every monthly cohort from the last 18 months. Y-axis: percentage of cohort revenue retained (not customer count — revenue). X-axis: months since acquisition.
Healthy DTC cohorts compress in months 1–3 and then flatten. The 2021 Allbirds S-1 disclosed that average repeat-customer spend in year two ran more than 25% above year one, which is the canonical shape investors want to see — cohorts that expand on a revenue basis even as customer count declines (Allbirds S-1, SEC filing). Unhealthy cohorts decline monotonically. The killer pattern is decay across cohorts: the January 2024 cohort retained 18% of its revenue at day 90, the April cohort 14%, July 11%, October 7%.
That pattern is the single most common reason a $3M GMV DTC raise dies in diligence. The headline says growth. The cohort chart says ad spend is buying lower-quality customers each quarter — fitness-deal hunters, discount-stacking arbitrageurs, one-and-done bargain shoppers who would never have entered the funnel before you scaled paid social. Growth is real. Quality is collapsing. The investor sees both in one chart.
There is no recovery in the meeting. Build the chart. Run it monthly. If you see decay starting, pause scaling and fix the acquisition channel mix before the raise — not during. The strategic implications of cohort behavior over time, including how to model lifetime value off real cohorts instead of blended averages, are covered in Customer LTV Reality.
Number 5 — Inventory Days Outstanding
Inventory days = (average inventory at cost ÷ COGS) × 365. A $2M GMV DTC apparel brand running 95 inventory days is sitting on roughly 26% of annual COGS in dead capital. A brand running 145 days is sitting on 40%.
Investors pull this number to ask one question: is the working capital problem getting better or worse as you scale? A brand that ran 110 days last year, 125 this year, 140 next quarter is funding the raise to plug an inventory hole, not to grow. That is a refinance, not a growth investment, and growth investors price it accordingly — or pass.
The benchmark range for DTC apparel is 90–120 days. Beauty runs 60–90. Consumables 30–60. Hardgoods 120–180. If you are above the high end of your category, the partner will ask why before they ask anything else. Common answers — overbought a hero SKU, weather-dependent seasonal, slow-moving long tail — are all defensible if you are explicit about them. Surprises in the inventory aging report kill trust faster than the absolute number does.
Number 6 — Working Capital Cycle (Cash Conversion in Days)
Working capital cycle = inventory days + accounts receivable days − accounts payable days. For most DTC brands AR is near zero (Shopify Payments settles in 2 business days), so the cycle simplifies to inventory days minus supplier payment terms.
A brand on Net 30 terms with 110 inventory days runs an 80-day cash conversion cycle. That means every dollar of growth requires 80 days of float financing. At $3M GMV growing 60% YoY, that is roughly $400K of incremental working capital needed in the next 12 months — separate from the operating cash burn. Investors back into this calculation in their heads and ask whether your raise size matches your cash conversion math. If you are raising $1.5M with $400K going to working capital and $400K to inventory build, you have $700K of actual operating runway. That is 7 months at $100K monthly net burn. That is not enough to hit the milestones in your deck.
This is also where Net 60 / Net 90 supplier negotiation becomes a real lever, not a nice-to-have. Pushing payables from Net 30 to Net 60 on $1.8M of annual COGS frees up roughly $150K of working capital permanently. The full mechanics of how to model this — and how the cash conversion cycle interacts with growth rate — are in Cash Flow Forecasting Model.
Number 7 — Return Rate by SKU
Blended return rate is a vanity number. A 22% blended return rate for apparel sounds bad in isolation, but if your top-5 SKUs return at 9% and your bottom-30 SKUs return at 41%, you have a merchandising decision, not a business problem.
Investors pull SKU-level returns because the data tells them whether you are running a brand or running a clearance operation. A brand with 8% return rate on hero SKUs and 35% on long-tail size-extended SKUs is one merchandising pruning away from 14% blended return, which moves contribution margin by 280–400 basis points. That is a real lever, and a partner who sees it before you do will price the deal lower because they will assume you are not paying attention.
The categories where this number matters most: apparel, footwear, eyewear, and any size/fit-driven product. Categories where it matters less but still gets pulled: consumables (return rate near zero, but refund/credit rate matters), beauty (chargeback rate), hardgoods (warranty claim rate). Whichever variant applies to your category, segment it by SKU, sort it, and bring the segmentation to the meeting.
Number 8 — Marginal Ad Spend Efficiency
Blended ROAS is the most misleading number in DTC. A 2.4x blended ROAS could be the result of an 8x branded-search ROAS, a 1.6x prospecting ROAS, and a 0.9x retargeting ROAS — averaged together to look acceptable. Investors do not want the blend. They want the marginal dollar.
The question they are asking: if you take the raise and put another $500K into Meta, what ROAS does that incremental $500K deliver? Because the incremental dollar always underperforms the average — that is the structure of paid acquisition. The dollar after that underperforms further. At some point the marginal dollar drops below contribution-margin breakeven, and every dollar above that point destroys gross profit on a contribution basis.
The way to present this is a curve, not a number: ad spend on x-axis, ROAS on y-axis, with a horizontal line at your contribution-margin breakeven ROAS. The point where the curve crosses the line is your acquisition ceiling — the maximum monthly ad spend that earns positive contribution. If your current spend is at $180K/month and the ceiling is at $240K/month, you can absorb $60K of incremental spend before the model breaks. That is the answer to "how much of the raise goes to growth marketing." If the curve crosses below your current spend, you are already past the ceiling and the raise should not fund more paid acquisition — it should fund retention, AOV expansion, or category expansion.
Build the Dashboard Before You Pitch
The eight numbers above are not a checklist — they are a triangulation grid. Investors are not pattern-matching for any single metric; they are pattern-matching for whether the eight numbers tell a coherent story.
Coherent looks like this: 24% contribution margin, 6.2-month CAC payback, 14%/22%/31% repeat at 90/180/360 days, cohort revenue retention curves that flatten at month 4, 95 inventory days, 65-day working capital cycle, 11% blended return rate driven by 30 long-tail SKUs you are pruning, and marginal ad ROAS at 1.9x against a 1.6x breakeven. That business raises capital at a fair multiple.
Incoherent looks like this: 58% gross margin, $4M GMV, 70% YoY growth — and a partner finds 8% contribution margin, 19-month CAC payback, cohort retention decay across 4 straight quarters, and 165 inventory days during diligence. That business does not raise. Or it raises at a 60% discount to the comp set and the operator does not understand why.
The action is straightforward, though not easy. Spend the next 30 days building the 8-number dashboard. Pull it monthly going back 18 months — you need the cohort decay visibility, and you need the trailing-90 vs trailing-12 contrast. Run the finished dashboard past three friendly investors who are not going to invest — current cap table angels, an operator-turned-VC you know socially, a category-specific advisor. Their job is to find the gaps and the incoherence. Fix what they find. Then start the formal raise.
The same discipline applies to thinking about your eventual exit math — because the metrics that get you funded are the metrics that get you acquired. The structural overlap is covered in Exit Math Nobody Does, which is worth reading before the term sheet stage, not after.
The single most-skipped step in this entire process is the cohort retention chart. Build it first. If it is decaying, you have a quarter of work to do before the raise — fix the acquisition channel mix, prune the discount stacking, retarget the high-LTV cohorts. Raising into a decaying cohort curve is the most expensive mistake a $1M–$5M GMV operator can make. The capital comes in at the wrong price, the milestones miss, and the next round either does not happen or happens at a flat or down mark. Two quarters of dashboard discipline before the pitch is the cheapest insurance in the entire raise process.
