The napkin test
I've spent a bit over two million dollars on ads across Meta and TikTok. The single most valuable habit I've built isn't a targeting trick or a creative formula, it's refusing to touch a product until it survives one napkin. Selling price, minus everything, minus the CPA I'll actually get (not the one I'm hoping for). If the napkin says no, no amount of optimism fixes it.
$34
My blended CPA across all of 2025, plan for this
$19
The CPA everyone puts in their projections
3x
Minimum sell/cost ratio before I even run the napkin
That gap between $19 and $34 is where most first stores die. Not from bad products, from projections written by their own hope.
The cost stack nobody writes down
Say you sell at $39.99 and the product costs $9 landed. 'Thirty dollars of margin!' No. Watch what actually happens to it:
- Payment processing: ~$1.45. Every sale, forever, off the top.
- Refunds and chargebacks: budget 5 to 8% of revenue. On a $40 product that’s $2 to 3 per order averaged out. Yes, even with a good product.
- Apps and tools: your $150/month stack spread over 300 orders is 50 cents each. Spread over 40 orders it’s nearly $4. Volume changes your costs.
- Shipping surprises: remote surcharges, reshipments for lost packages, the occasional angry-customer overnight. Call it $1 averaged.
- The CPA. The real one. If you haven't run ads before, use $30+ for a $40 product and argue with me later.
Run it: 39.99 − 9 − 1.45 − 2.50 − 1 − 1 − 30 = a loss of five dollars per order. That product with 'thirty dollars of margin' loses money at a completely normal CPA. This exact math, un-run, is the most expensive blank spot in dropshipping.
Check margin room before you fall in love
Majorka's verdict includes a margin-room bar that runs a version of this stack against typical CPAs for the product's category. If that bar is short, the napkin math will hurt. Look at it first, not after the creative is shot.
AOV is the cheat code, not CPA
Everyone attacks the wrong side of the equation. They fight for a $2 lower CPA, a brutal, competitive fight against every other advertiser, when a $12 higher average order value does the same thing to the napkin and nobody's competing with you for it. Your bundle is between you and your customer.
- Two-pack default: for consumables and anything people share, make quantity two the pre-selected option. This one change is regularly worth 20% of AOV.
- The $9 companion: find the accessory the source data says people buy anyway (the variant and bundle data shows you), and offer it post-purchase so it can’t hurt conversion.
- Shipping threshold just above your AOV: if the average order is $42, free shipping at $50 drags the whole distribution rightward.
“I've never scaled a product with great CPA and bad AOV. I've scaled several with mediocre CPA and great AOV.”
Redo the napkin with AOV at $52 instead of $39.99: suddenly the same product, same CPA, makes seven dollars a unit. Nothing about the market changed. You just did the math on the side of the equation you control.
Written by
MaximusFounder of Majorka
Built Majorka after years of running his own stores and being tired of guessing. Every post here comes from real tests, real ad spend, and real mistakes he paid for so you do not have to.
Share article
Put this playbook to work in 30 seconds
Paste any AliExpress link and see the order curve, velocity, and verdict for yourself. Plans from $39/mo.
See it in Majorka4.9/5 from 1,632 sellers
Orders · 24 wks
Scale23,491
