A homeware store in Ankara sells two ceramic planters. They are close to the same size, close to the same weight, and they sell at roughly the same rate. One of them lands a full desi step higher on every single order, so its volumetric weight costs more to ship, and it has done so since the day it was listed.
Nobody in the business can say why, because nobody has ever had a reason to look. The shipping cost arrives at the end of the month, aggregated, on an invoice that is mostly correct. The difference per order is small enough to disappear into it, and large enough that over a year it is a line item somebody would notice if it had a name.
The reason is a box that is a little over one centimeter too tall.
This guide is about volumetric weight — desi, in Turkey — and specifically about what happens to it after you calculate it. Every merchant knows the formula. Very few know that between the tape measure and the invoice line there are three separate transformations, that one of them is a rounding rule whose boundary probably isn't where you think it is, and that a parcel's cost is decided by which side of that boundary it lands on.
If you just need the arithmetic, use the desi calculator — enter the dimensions and it gives you the exact and the rounded value. This guide is about the decisions you make once you have the number. It sits alongside packaging optimization, which is about the boxes themselves, and carrier invoice reconciliation, which is about catching it when the carrier's number and yours disagree.
The Number You Calculate Is Not the Number You Pay
Between measuring a box and paying for it, your parcel goes through three transformations. Each one is a rule somebody else wrote, each one is in your tariff, and most merchants have read none of them.
First: the volume becomes a unit, and the unit depends on the border
Yurtiçi Kargo sets this out plainly on its own pricing page. It computes volume in cubic decimeters — width times length times height in centimeters, divided by 1,000 — and then divides that volume by three. That is the same arithmetic as the familiar formula, written the way the carrier actually thinks about it: one desi is three liters of space.
For international shipments the divisor changes from three to five. Same box, fewer units, because the economics of an aircraft are not the economics of a truck.
This is worth stating clearly because imported advice gets it wrong constantly. English-language guides talk about dimensional weight with divisors of 5,000 or 6,000, and merchants reading them assume the same applies at home. It does not. A domestic Turkish parcel is measured against a divisor of 3,000, which makes the same box between about 1.7 and 2 times more expensive in units than the imported guide suggests. If you have ever read a foreign packaging article and concluded your boxes were fine, re-run the numbers.
Second: the fraction is rounded, and the boundary may not be where you think
Here is the rule Yurtiçi Kargo publishes: when the result has a fraction, a remainder between 0.1 and 0.49 is dropped and only the whole number is used; a remainder between 0.50 and 0.99 rounds the whole number up by one.
That is conventional half-up rounding. It is not what most people believe. Nearly every online calculator and explainer in Turkish states the rule as the fraction always rounds up — 3.2 desi is billed as 4. Those are two genuinely different rules, and they put the expensive boundary in two completely different places.
Take a parcel that measures 25 × 20 × 34 cm. That is 17,000 cm³, or 5.67 desi.
- Under half-up rounding, it is billed as 6. To be billed as 5, you need to get the fraction below 0.50 — under 16,500 cm³ — which on that base means bringing the height below 33 cm. A little over one centimeter.
- If the rule were always-round-up, you would need to get under 5.00 desi outright — 15,000 cm³ — which means a height of 30 cm. Four centimeters.
One of those is a tape-and-filler adjustment you can make this afternoon. The other means sourcing a different box. A merchant who believes the wrong rule looks at the four-centimeter version, concludes it can't be done, and never touches it again.
Fix it: find the rounding clause in your own tariff and read it. It is one sentence. If your agreement doesn't state it, ask your account contact in writing and keep the reply. Until you know which rule applies to you, every packaging decision you make about desi is aimed at the wrong target.
Third: desi is compared to actual weight, and the larger one wins
The carrier weighs the parcel and calculates its desi, then bills whichever number is larger. On a Turkish invoice this is the ücrete esas ağırlık — the chargeable weight — and it is the only number in the whole chain that turns into money.
Practically, this splits your catalog in two. Light, bulky products are priced by their dimensions and every centimeter matters. Small, dense products are priced by the scale and no amount of packaging work will help them. Before you spend an afternoon on boxes, find out which half each product is in. Optimizing the packaging of a product that bills on actual weight is effort with a guaranteed return of zero.
Where the Money Actually Is: the Parcels Sitting Just Past a Boundary
Once the price is a step function, the value of a centimeter stops being constant. Shave a centimeter off a parcel sitting comfortably in the middle of a step and you have saved nothing at all. Shave the same centimeter off a parcel sitting just past a boundary and you have saved a full step on every order of that product, forever.
This is the part almost nobody works on, because it looks like packaging work and it is really data work. You cannot see it on a parcel. You can only see it in a list.
Rank your products by distance past the boundary
The procedure takes about an hour for a normal catalog and it is the highest-return hour available in shipping cost management.
- Take your top 20 products by monthly order count. Not by revenue — by parcel count, because that is what you pay per.
- For each one, pack it exactly the way it actually ships: the real box, the real filler, the tape, the invoice pouch.
- Measure the packed parcel at its widest points and calculate the desi.
- Write down the fraction. Not the rounded figure — the fraction.
- Sort by how far past the nearest boundary the fraction sits, smallest gap first.
The products at the top of that list are ones where a small physical change buys a whole step. Products sitting just under a boundary need protecting rather than improving: a slightly thicker box or an extra layer of bubble wrap will quietly push them over, and nobody will connect the invoice to the decision three months later.
One centimeter is not one price
The consequence of a step function is that the same intervention has wildly different value depending on where it lands. That cuts both ways, and the downside is the one that catches people out.
A supplier changes their carton spec by 5 mm. A new filler material is a little loftier than the old one. Someone starts double-boxing fragile items after a run of damage claims. Each of these is a sensible, small decision made by someone who is not thinking about tariffs, and any of them can move a high-volume product across a boundary. The cost shows up a month later, spread across hundreds of parcels, in an invoice that nobody reads line by line.
Fix it: for your ten highest-volume products, record the packed desi and the fraction somewhere durable. When anything about the packaging changes, re-measure those ten. It takes ten minutes and it is the only early warning you will get.
Measure the Parcel You Ship, Not the Product You Sell
Almost every desi number that turns out to be wrong is wrong for the same reason: somebody measured the product, or the flat box, and not the thing that actually goes to the branch.
The carrier measures the outside, at its widest
A carrier measuring your parcel does not care about your intentions. It measures the outermost points of the object it is handed. That includes the bulge where the box is over-filled, the tape ridge, the document pouch on the side, and any handle or protrusion. A box rated 30 × 20 × 15 that is packed until the lid domes is not a 30 × 20 × 15 parcel any more.
This is also why measurements taken from a supplier's spec sheet are unreliable. The spec is the internal dimension, or the flat dimension, or the nominal one. The carrier bills the assembled, packed, taped reality.
Multi-item orders don't add up
Two products with a desi of 3 each do not make a parcel of 6. Packed together in one box they might come to 5, or to 7 if the shapes fight each other and the box has to be a size up. Desi is a property of the parcel, not of the order's contents, and it is not additive.
For stores where multi-item orders are common, this is a genuine planning problem rather than a rounding curiosity. The combinations that occur most often are worth measuring as combinations. If your two best-selling items are frequently bought together, the parcel that carries both is effectively a third product as far as your shipping cost is concerned, and it has never been measured.
Your Stored Desi Is a Forecast. The Carrier's Is the Outcome.
There are three desi figures in a typical store, and nothing in the ordinary flow of work ever brings them together.
The one you calculated when the product was added, from a measurement someone took once, possibly of the product rather than the parcel.
The one your platform uses to price shipping at checkout or to pick a rate — usually the stored figure, typed into a field and never revisited. This one determines what you charge the customer, which is a pricing decision, not a cost one.
The one the carrier measured at the branch or the sorting hub, which becomes the chargeable weight on the invoice.
Only the third one is real. The first two are estimates, and an estimate that is never compared to the outcome is not an estimate — it is an assumption. If your stored desi for a product is one step below what the carrier consistently bills, you have been quoting shipping, planning margin and choosing carriers on a wrong number for every order of that product since it was listed. Nothing in the process will tell you. The invoice is aggregated, and the difference per parcel is small.
This is the same mechanism described from the platform side in the IdeaSoft shipping guide, where a per-product desi field is typed once and the carrier re-measures every time. It applies wherever the number is stored, regardless of platform.
Fix it: capture what you expected to pay at the moment the label is created — the parcel, the carrier, the service, the desi you calculated, the chargeable weight you expected. Then compare it to the invoice monthly. The full procedure, including how to group differences by product rather than by shipment so that repeating errors surface, is in the carrier invoice audit guide. Systematic differences are worth far more than one-off ones, because they repeat on every order until someone notices.
Minimum Desi: the Floor You Pay Even When You Ship Air
There is a term in many Turkish carrier agreements that never appears on a parcel, never appears in a panel, and quietly resets the economics of a whole category of store: a minimum.
It takes more than one form. Some agreements set a floor per parcel, so a genuinely small shipment is billed as though it were larger. Others set a minimum monthly volume you commit to, whether or not you ship it. Some tariffs simply start their table at a level that functions as a floor. In each case, the effect is the same: below a certain size, making your parcel smaller stops saving you anything.
For a store shipping jewelry, cosmetics, spare parts or anything else that is small and dense, this can be the single most important line in the contract. It is entirely possible to negotiate an excellent headline rate and a poor effective one, because the discount applies to steps you never use and the floor applies to every order you ship.
Fix it: work out the desi distribution of your actual parcels — what share sit at 1, at 2, at 3, and so on. Then look at your agreement's floor and at where the discount in your rate table actually bites. If most of your volume sits below the floor, you are not buying shipping, you are buying a subscription. That distribution is also the single most useful thing you can bring to a renegotiation; the shipping agreement guide covers the rest of that conversation.
Desi Is Only Half the Price. The Other Half Is Distance.
A desi figure on its own does not produce a price. Turkish domestic tariffs are a grid: chargeable weight on one axis, distance band on the other. Yurtiçi Kargo publishes its bands as near line for 1–200 km, short line for 201–600 km, medium line for 601–1,000 km and long line beyond that, with in-city and Cyprus handled separately.
Two consequences follow, and both are routinely missed.
The first is that the cheapest carrier is a property of the parcel and the route together, not of the carrier. A carrier that is excellent on your compact, high-volume products across short distances may be poor on the bulky ones going cross-country. Comparing carriers on a single headline rate answers a question nobody in your operation ever asks. This is the practical argument behind running more than one carrier, and the reason per-carrier cost reporting has to be split by route before it means anything.
The second is that flat national pricing hides a transfer. Some sales-platform shipping programs charge the same price for a given desi regardless of distance. That is genuinely simpler, and for some stores it is the right trade. But it means your in-city orders are subsidizing your cross-country ones, invisibly, and the size of that subsidy depends on a geographic mix you can measure and they don't show you.
Note also that the grid does not run forever. Tariff tables stop at a top row — Yurtiçi's domestic table tops out at 50 kg of chargeable weight — beyond which an incremental per-unit rate applies. If you ship anything large or bulky, the step function you have been optimizing against turns into a straight line at the top of the table, and the rules change with it.
Four Levers, in Payback Order
Everything above narrows down to four things you can actually do. They are listed in the order that pays back fastest, which is not the order most stores attempt them in.
1. Correct the desi figures that are simply wrong. Costs nothing, takes an afternoon, and fixes an error that repeats on every order. This is first because it is the only lever that requires no purchase, no negotiation and no change to how anyone works.
2. Move the boundary products back under the line. The ranked list from earlier. A smaller box, less filler, a flatter void-fill material, or packing an accessory beside the product instead of on top of it. Small physical changes, targeted at the specific products where a step is within reach.
3. Rework the box portfolio. Real money, slower, and the one most stores start with. Fewer size gaps means fewer products shipping in a box a size too big. This is proper packaging work and it is covered in depth in the packaging optimization guide — right-sizing, matching products to boxes, and the damage trade-off that comes with going smaller.
4. Renegotiate the parts of the tariff you actually use. Not the headline discount: the rounding rule, the minimum, and the specific steps and distance bands where your volume concentrates. This is last only because it works far better once you have the data from the first three. Walking in with a measured desi distribution is a different conversation from walking in with a request.
What Not to Do
The restraint list matters as much as the action list, and most of these mistakes are made by people trying hard to do the right thing.
Don't under-declare desi to get a cheaper label. It does not work and it costs more than it saves. The carrier re-measures, the difference is re-invoiced, and you have spent the credibility you need for every legitimate dispute you will ever raise. A merchant whose declared figures are reliably accurate is in a much stronger position when a charge really is wrong.
Don't buy fourteen box sizes. Every additional size adds storage, procurement and a decision at the packing table, and the decision is where errors enter. Most catalogs are well served by a handful of sizes chosen deliberately.
Don't optimize a product you ship three times a month. The saving is a step on three parcels. Sort by volume and stop when the list stops being worth your time.
Don't shave packaging into damage. A smaller box that arrives broken costs the outbound leg, the return leg, the replacement and the customer. The economics of that trade are in the lost and damaged parcels guide — check them before you take protection out of a fragile product.
Don't trust a stored desi you didn't measure. Not the supplier's spec, not the value inherited from an old catalog import, not the figure someone typed in a hurry two years ago.
Don't apply an imported dimensional weight playbook. The divisors are different, the tariff structure is different, and the accessorial charges are different. This is the same lesson as the cross-border shipping guide: the rules that govern a parcel are the rules of the country it moves in.
Why September and October Are the Window
This is unglamorous work and it will always lose to something more urgent, so it is worth being specific about the deadline.
Boxes have lead times. If the ranked list tells you that three of your best sellers need a slightly different carton, that is a supplier conversation, a sample, and a delivery — comfortably several weeks before the first parcel goes out in the new size. Starting that in November is starting it too late.
A per-parcel error also multiplies with volume, and volume is about to multiply. A single step of overcharge is a rounding error in September and a meaningful number across a peak-season November. The work is the same either way; only the payback changes.
And packaging is exactly the kind of change that should be frozen before the season starts. The peak season preparation guide makes the general argument: the last safe date to change anything structural is roughly a month out, because you need live volume to find out whether the change worked. A new box introduced in the middle of peak is a variable added at the worst possible moment. Measure now, change in September and early October, and spend the rest of October confirming rather than adjusting.
How Shipink Handles the Desi Side
Four parts of this are software problems, and this is where Shipink fits.
Desi is calculated for every order automatically, from the dimensions of the parcel rather than a figure recalled from memory at the packing table, and it feeds the rate comparison directly.
Rates are compared across 15+ carriers for the actual parcel and the actual destination, which is the only way the desi-and-distance grid can be resolved per order rather than per carrier. You can run your own carrier agreements or Shipink's, or both side by side.
Shipink's own agreements carry no minimum desi and no volume commitment — shipments start from 1 desi. For a store whose parcels sit at the small end, that removes the floor problem described above rather than negotiating around it.
Per-carrier reporting includes cost and desi, so the question "which carrier is expensive for which kind of parcel" has an answer that comes out of a report instead of a spreadsheet someone maintains by hand.
Two honest limits. Shipink does not measure your boxes — the dimensions that go in are the dimensions you provide, so the measuring discipline in this guide stays yours. And nothing prevents a carrier from re-measuring a parcel and re-invoicing the difference. What the platform can do is give you the expected figure at label time, which is exactly the record you need to notice when that happens.
The Desi Audit: a One-Hour Checklist
Know the rules that apply to you
- Read the rounding clause in your own tariff, and write down which rule it uses
- Confirm the divisor for domestic and for international shipments
- Find any minimum — per parcel or per month — and note it
- Note where your tariff table ends and the incremental rate begins
- Check which distance bands your orders actually fall into
Measure what you actually ship
- List your top 20 products by monthly parcel count, not by revenue
- Pack each one exactly as it ships, filler and tape included
- Measure the packed parcel at its widest points
- Record the exact desi and the fraction, not the rounded figure
- Measure your three most common multi-item combinations as parcels
Find the money
- Sort the list by distance past the nearest rounding boundary
- Mark which products bill on desi and which bill on actual weight
- Identify the products sitting just under a boundary, and protect them
- Estimate your desi distribution: what share of parcels sit at each step
Close the loop
- Record the expected chargeable weight at the moment each label is created
- Compare expected against invoiced monthly, grouped by product
- Re-measure your ten highest-volume parcels whenever packaging changes
- Bring the measured distribution to your next rate negotiation
The Number Was Never the Point
Desi is easy arithmetic and every merchant can do it. That is precisely why it gets so little attention: it looks solved.
What is not solved is everything the number passes through afterwards. A rounding rule that decides which side of a step you land on. A comparison against actual weight that makes half your catalog immune to packaging work. A stored figure that nobody has ever checked against an invoice. A floor in a contract that makes small parcels stop getting cheaper. A distance grid that makes "our cheapest carrier" a sentence with no fixed meaning.
None of that requires new software, a new carrier or a bigger team. It requires one afternoon with a tape measure, a list sorted the right way, and a habit of comparing what you expected to pay against what you were billed.
Start with the desi calculator and your twenty best sellers. If you would rather the calculation, the comparison and the reporting happened for every order without anyone thinking about it, see how Shipink prices a parcel or start free.
