Operating CrewVenchi · October 19, 2026

Packaging analysis · Process & confidence

A starting point for pricing.
Final box sizes still need validation.

We reconstructed 21,587 Venchi DTC orders from September 2025 through August 2026, modeled how the items fit together, and compared ranges of six to eight box sizes. The result supports prototype selection and indicative quotes.

The layouts are geometrically consistent. Confidence in final box specifications is low until estimated dimensions and physical packouts are validated.

Only 16.8% of orders have stored retail dimensions for every item. Physical packouts have not been checked, so there is no measured accuracy rate.

Output · Candidate box range

Seven sizes to prototype, with six and eight for comparison

The table allocates each historical order to the smallest selected box that contains one of its tested layouts. These are provisional internal carton dimensions, including the model’s generic clearance. External shipping dimensions will be larger. Final liner thickness and gel packs can increase the required size.

Number of box sizes
Internal box, inchesOrders in yearShare of all orders
11 × 8 × 59,50944.0%
13 × 11 × 55,29624.5%
16 × 13 × 52,38611.1%
13 × 11 × 92,21110.2%
19 × 13 × 91,4966.9%
21 × 15 × 81450.7%
24 × 24 × 245072.3%
Separate packout / no qualifying layout370.17%
Total positive-demand orders21,587100%

21,550 orders fit this range (99.83%). Mean assigned box volume: 1,114 in³ per fitted order. Shares use all 21,587 positive-demand orders; totals may differ slightly due to rounding.

Main demand: the first two sizes in the seven-size base range cover 14,805 orders, or 68.6% of the year’s positive-demand orders.

Large-order warning: the 24-inch cube covers 507 orders (2.3%) but accounts for 29.2% of selected box volume. Test split shipments or an existing bulk carton before making it a regular DTC SKU.

What the extra box size buys

Box sizesMean selected volumeReduction vs previous rangeExtra volume vs tailored fits
61,165 in³Baseline101.7%
71,114 in³4.4%92.8%
81,076 in³3.4%86.2%

All base ranges cover the same 21,550 fitted orders and leave 37 exceptions. Reductions describe carton volume, not freight savings. “Tailored fits” means the smallest feasible layout found for each individual profile. It is a heuristic benchmark; the extra-volume column does not measure empty space.

Seven is a useful candidate for quotes. Its volume reduction over six is modest, and eight offers another small reduction. Carton cost and insulation overlap must decide whether either extra SKU earns its place.

How the items fit together

These examples come from the saved packing coordinates. Each colored shape is one modeled physical unit. The perspective drawing can hide items behind others; the underlying coordinates were checked for overlap and box boundaries. Product descriptions retain the export’s case-pack wording; physical quantities follow the modeled selling unit and bundle expansion.

Profile P00250 · 11 × 8 × 5 in

2 × chocoviar 75% soft bar 170 g 12 pcs; 2 × chocoviar pistacchio soft bar 170 g 12 pcs; 1 × chocoviar stracciatella soft bar 170 g 12 pcs

Stored retail dimensions throughout. Seen in 1 historical order.

Profile P00006 · 13 × 11 × 9 in

2 × hazelnut spread 450 g 8 pcs; 1 × mini 75% extra dark bar golden goose 30 g 32 pcs; 1 × 6x white hazelnut bar kit

Least certain input: generic estimate. Seen in 1 historical order.

The model tests shelf and layer arrangements with multiple base dimensions and item sort orders. It allows rectangular rotations, with jars upright. It includes 0.5 inches of clearance at each wall and 0.125 inches of jar wrap on each surface. A failed search does not prove that an order cannot fit.

How the analysis works

  1. 01

    Reconstruct each order

    Use its SKUs and quantities, separate transport charges and returns, and expand bar bundles into physical items.

  2. 02

    Source dimensions, then match gaps

    Start with individual-product dimensions. Match missing SKUs by name, packaging format and unit weight. Keep each borrowed dimension linked to its source SKU.

  3. 03

    Test physical arrangements

    Test item rotations and stacking. Keep jars upright with room for protection. Every reported fit has a nonoverlapping three-dimensional layout.

  4. 04

    Compare shared box ranges

    Group feasible layouts into six, seven or eight sizes. Allocate historical demand to each size and compare total box volume. Carton prices and insulation availability remain separate inputs.

What the dimensions are based on

Of 495 physical SKUs, 129 received a matched dimension estimate: 41 close product-name matches, 28 strong format matches and 60 broader proxies. All confidence descriptions below are qualitative judgments.

Input basisSKUsHow to treat it
Stored retail dimensions85Best available; website values remain unmeasured in hand.
Strong product or format match69Relatively high confidence in the packaging equivalent.
Broader physical-format proxy60Moderate to low; a similar package may differ in shape.
Unmatched; generic estimate281Low; a planning assumption remains necessary.

Two suspect source records were excluded from the original 87 dimensioned SKUs because their dimensions appeared to describe a praline inside a larger gift box.

A strong match: a renamed 110 g cremino bar can use the dimensions of its equivalent dimensioned bar.

A weaker match: a seasonal gift box can share a named format with another collection while still having a different footprint or depth. Theme alone does not establish a match.

How that affects whole orders

Each order appears once below, according to its least certain item.

Least certain input neededOrdersShare
Direct dimensions for every item3,63116.8%
Strong matches needed; no broader estimates1,9869.2%
Broader proxies needed; no generic estimates2,32010.7%
At least one generic dimension estimate13,65063.2%

Direct dimensions plus strong matches cover every item in 26.0% of orders. Including broader proxies raises that to 36.8%. Percentages are rounded.

Sensitivity to estimated dimensions

Increase inferred edges by 25%: average box volume rises about 39%.

We enlarged every borrowed or otherwise estimated dimension and reran the packing and box selection. The seven-size range required more volume, and 59 additional orders fell outside its fitted cohort. This is an illustrative stress scenario; it does not establish the true error range.

What changed in the stress run

For seven sizes, mean selected box volume increases from 1,114 to 1,547 cubic inches per fitted order. Exceptions rise from 37 to 96. The averages therefore cover slightly different sets of fitted orders.

The original seven-size range still accommodates 99.56% of positive-demand orders under the larger-item assumptions because the larger cartons absorb demand. That coverage does not show that the small-box assignments are stable. Use the scenario selector above to inspect the sizes selected after the stress run.

Why the result could change

Package shape and source quality

Similar names can hide a different package shape or presentation. Flexible bags behave differently from rectangular models. Pick & Mix bags have no dimensioned bag equivalent in the saved catalog.

Selling units and kit contents

Case counts can be confused with customer units. Eight mixed-kit SKUs still use an inferred six-bar count, affecting 155 orders (0.7%). Those assumptions remain flagged.

Protection and thermal packaging

The fit model uses generic clearance and jar-wrap allowances. The selected liner and gel packs still need their own space. Physical handling and transit protection can also change the arrangement.

Commercial choices and future demand

Box volume alone does not determine the best price. Minimum runs and split shipments may favour a different range. The large catch-all carton inflates fit coverage, and the next assortment can change demand by size.

What this supports for pricing

Use the proposed sizes for indicative comparisons with existing insulation SKUs and custom-run minimums. Keep the dimensions and volume allocations provisional. Before locking custom specifications, measure the high-volume gaps and test representative orders with the intended protection and insulation.

Modeled fit coverage is a measure of how many assumed item layouts fit the candidate range. It provides no empirical accuracy estimate. Seven sizes remain a candidate for testing; the analysis does not establish a final commercial optimum.

What to measure first

Start with estimated SKUs that appear in the most orders. Counts overlap because one order can contain several of these products.

SKUFirst dimensions to confirmOrders containing itCurrent basis
116703latte nero milk chocolate 37% cocoa promo bar 70 g 24 pcs _ 20261,015Strong match
638031bubbles metal tin christmas corousel of dream design 2 83 mm 12 pcs _2027815Generic estimate
199016126253 - 9 unit - cremino extra dark chocolate gianduja - p&m726Generic estimate
925355sorpresa interna - pouch cuoio rigenerato _ 2023608Generic estimate
116211pistachio cremino bar 110 g 24 pcs605Strong match
116030pistachio gourmet bar 95 g 24 pcs598Strong match
199004126223 - 9 unit - cremino white chocolate gianduja - p&m585Generic estimate
108427gift box christmas vintage dream 219 g 8 pcs572Broader proxy
116209cremino 1878 bar 110 g 24 pcs565Strong match
199073126397 - cremino coffee 40 pcs - p&m560Generic estimate
Excluded dimension donors

These source records were withheld from matching. The first two also lost their direct-dimension status.

Source SKUProductWhy excluded as a donor
197141197141-blue box 20 assorted pralines 252 gStored tiny dimensions appear to describe a praline, not the252g gift box
197142197142-orange box 30 assorted pralines 372 gStored tiny dimensions appear to describe a praline, not the372g gift box
105147105147-chocoviar soft bar g 200 pcs 12Stored200g versus current170g and inconsistent soft-bar envelope
116652116652-snack chocobiscuit 35 g 32 pcs35g snack dimensions conflict with corroborating35g snacks
143052143052-mini bag mother's day 42 g 20 pcsMini-bag name conflicts with mini-book metadata
122094122094-chocolate cigars in box 1 pcs5400g boxed cigar object, not an individual100g cigar

All SKU assumptions and order profiles

The detail views below contain the full model behind this page. Order profiles combine identical SKU-and-quantity lists; their frequencies add to 21,587 orders. Profile IDs are anonymous. Customer details and source order numbers are omitted.

SKU dimension model · 495 products

Dimensions are inches per modeled physical unit. Bundle rows expand each selling unit into the stated constituent count. Orders containing a SKU are overlapping counts.

Order profiles · 14,495 distinct item combinations

Sorted by historical order frequency. Box assignments and individual fits below use the base dimension scenario. “Individual fit” is the smallest-volume layout found. It is not a proven minimum.

Source reconciliation and model boundaries

The full-year export contains 21,606 source order numbers. 21,587 have positive physical demand; 19 have none. The model uses all positive-demand orders, including bulk orders. The largest contains 796 selling units.

Positive quantities are combined by order and SKU. The 1,029 negative-quantity rows remain in the source but do not subtract from positive outbound packing demand. Transport charges and zero quantities are excluded from physical contents. Fractional positive rows resolve to whole units after aggregation. The source footer differs from the detailed quantity rows by 20 units; the calculations use the detail rows.

The older January–April export overlaps the full-year file and is not added to it. The sample-orders PDF is a separate sample and is not the denominator for this analysis. Stored product dimensions are retail values from the saved catalog. Marco’s outer-case measurements are unsuitable as retail-item dimensions.

Pick & Mix descriptions can say “40 pcs” while selling-unit and weight metadata indicate about 100 g. The model follows that selling-unit evidence; the earlier sample scenario used a larger 40-piece bag. 19 bundle SKUs expand to six physical constituents each. Inferred mixed-kit counts still need confirmation.

Candidate shared boxes are limited to 24 inches on each side as a modeling choice. This is not a carrier rule. Oversize or unresolved profiles remain exceptions. All 99,244 base layouts and 100,606 stress layouts were checked for item conservation, permitted rotations and nonoverlap. Those checks establish geometric consistency under the assumed inputs.

Weights helped assess product equivalence. Actual packed shipment weights and carrier charges were not calculated. Carton strength and fragility rules remain unverified. There is no completed comparison with Justin’s insulation catalog or custom-run pricing, so the page cannot yet establish a final cost-saving range.

Order source: Ecom orders 9-1-2025 8-31-2026.xlsx, Export tab. Dimension source: Venchi_SKU_Physical_Dimensions.xlsx, SKU Physical Dimensions tab. Analysis date: October 19, 2026.

Scope: Venchi DTC orders only. Sources: September 2025–August 2026 e-commerce order-line export and stored retail-product dimension catalog. Each matched estimate retains its source SKU in the working analysis.