Two theses in v0.1 are dead. Read this first.
Between v0.1 and this document we stopped modeling and started measuring. We pulled Golden Glaze's actual purchase ledger, obtained a real vendor cost sheet, and scraped the incumbent's complete published price book. Two load-bearing assumptions did not survive contact with the data.
Dead thesis #1 — "we win on price transparency"
v0.1 argued that bakery distribution runs on opaque, rep-negotiated pricing and that a flat published list would be genuinely disruptive. Texas Bakery Supply already publishes its entire price list publicly. We scraped 952 priced SKUs off tbshtx.com. Transparency is table stakes in this category, not a wedge. Any pitch built on it is a pitch the incumbent already answers.
Dead thesis #2 — "we displace a sleepy incumbent on price"
Normalized to dollars per pound, same pack size, comparable brand tier, Golden Glaze currently pays more than Texas Bakery Supply's public shelf price on 8 of 14 clean matches. Not marginally — on raised mix we are 28.9% above their shelf, on glaze sugar 49.2% above, on roll icing 48.8% above. Texas Bakery Supply is an efficient buyer with real purchasing scale, not a lazy incumbent waiting to be taken.
At our current vendor pricing, gross margin on the core basket is not merely below the 15% gate — it is negative. We would be paying more for the goods than the incumbent charges the customer.
| Item (normalized $/lb) | Golden Glaze pays | TBS shelf | Delta |
|---|---|---|---|
| Pillsbury yeast-raised donut mix, 50# | 0.760 | 0.540 | −28.9% |
| Glaze / icing sugar, 50# | 1.100 | 0.559 | −49.2% |
| White roll icing, ready-to-use | 2.798 | 1.433 | −48.8% |
| Powdered sugar 12X, 50# | 0.845 | 0.529 | −37.4% |
| Instant dry yeast, 20# | 3.495 | 2.647 | −24.2% |
| Lemon filling, ~37# | 1.314 | 1.135 | −13.6% |
| Bavarian creme, ~37# | 1.208 | 1.103 | −8.7% |
| Donut fry shortening, 50# | 0.987 | 0.920 | −6.8% |
| Bread / donut flour, 50# | 0.358 | 0.339 | −5.3% |
| Pillsbury Tender Taste base, 50# | 0.840 | 0.910 | +8.3% |
| Powdered sugar 10X, 50# | 0.699 | 0.836 | +19.5% |
| Powdered sugar 6X, 50# | 0.679 | 0.926 | +36.4% |
| Granulated sugar, 50# | 0.690 | 0.739 | +7.1% |
| Apple filling | 1.981 | 2.172 | +9.7% |
Known defect in the TBS data. Exactly one item number carries two prices: 38159, Pillsbury 365 raised mix 50#, listed at both $27.00 and $40.45. We used $27.00. One further near-duplicate ($26.45 vs $26.95 on Velvet 12X). Everything else is internally consistent. Separately, 920 of 929 products render
schema.org/OutOfStock — on a Wix storefront this normally means the catalog is not wired to live inventory, so treat this as a published list, not a live quote. Someone should call TBS as an ordinary shop and verbally confirm two or three prices before we bet on them.
What survived, and one thing that got better
Survived: digital ordering as a cost-to-serve engine rather than a customer pitch; Golden Glaze's operating credential as the trust wedge; the "we will not open a shop near a customer" commitment as a necessary and rehearsed answer.
New and stronger: Texas Bakery Supply is open Monday to Friday, 8:30 a.m. to 3:30 p.m., closed Saturday. Donut shops bake overnight and run out on weekends. That is a service-window gap we can attack without touching price, and it is now the primary differentiator.
The audit paid for itself before we sold anything
The same comparison that killed the price thesis found real money inside the business we already run. Against Golden Glaze's measured ~$310K/yr distributor run rate, correcting the three worst lines — glaze sugar, roll icing, 12X powdered — is on the order of $31K/yr straight to Golden Glaze EBITDA. That is worth doing this month whether or not Golden Star Supply ever launches.
One number decides whether this business exists
Everything downstream — trucks, warehouse, hiring, the whole plan — is contingent on landed cost. Here is the arithmetic, stated as plainly as it can be stated.
To sell a shop something at a price below what Texas Bakery Supply already charges, and still keep a gross margin that covers a truck and a driver, our landed cost has to sit far enough below their shelf price to leave room for both the discount and the margin.
| If our landed cost is… | …we must sell at (20% GM) | vs TBS shelf | Verdict |
|---|---|---|---|
| 60% of TBS shelf | 75% of TBS | −25% | Crushing. Whole plan works. |
| 65% of TBS shelf | 81% of TBS | −19% | Strong. Clear reason to switch. |
| 70% of TBS shelf | 87% of TBS | −13% | Workable. Switch-worthy on price alone. |
| 72% of TBS shelf | 90% of TBS | −10% | The target. Minimum credible discount. |
| 75% of TBS shelf | 94% of TBS | −6% | Thin. Not enough to move a loyal buyer. |
| 80% of TBS shelf | 100% of TBS | 0% | Price parity. Must win on service only. |
| 85% of TBS shelf | 106% of TBS | +6% | Dead. We are more expensive. |
Where we stand today
On the matched basket, Golden Glaze's current BakeMark dealer pricing averages roughly 118% of TBS shelf. The target is 72–80%. That is a required landed-cost reduction of ~35–40% from where we buy today.
That gap is not closed by negotiation, volume commitments, or being clever. It is closed by moving from the dealer/bakery price tier to the distributor price tier — a different price book entirely, which manufacturers extend to resellers, not to bakeries. Either that tier exists at the required level or this business does not open.
Phase 0 action — two phone calls, this week
This is not more analysis. It is two conversations with reps we already have relationships with. The question is identical for both: "What does your distributor tier look like if I'm buying to resell to fifty-plus independent shops in DFW, not to bake with? Send me the reseller price book and the volume commitment it requires."
- Wayne Clayton — Dawn Foods, DFW Market Specialist — 214-878-4577 — DC at 3353 Miller Park South, Garland TX 75042
- Peter Song — BakeMark, Sales Rep — 214-558-3014 — 3115 N Great Southwest Pkwy Ste 200, Grand Prairie TX 75050 — peter.song@bakemark.com
Ask both. Do not accept a verbal indication — get the sheet. Then re-run the table above with real numbers. Nothing else in this plan should be funded until that sheet is in hand.
Markup-to-margin conversion, for the rep conversation
Distributors quote margin; suppliers quote markup. Do not confuse them in a negotiation.
| Gross margin | = markup on landed cost | = price multiplier |
|---|---|---|
| 15% | 17.6% | 1.176× |
| 18% | 22.0% | 1.220× |
| 20% | 25.0% | 1.250× |
| 22% | 28.2% | 1.282× |
| 25% | 33.3% | 1.333× |
| 28% | 38.9% | 1.389× |
| 30% | 42.9% | 1.429× |
For orientation: Sysco runs 18.4% gross margin and keeps 3.8% operating. US Foods runs 17.4% and keeps 3.0%. Both figures are audited FY2025 10-K numbers, not press releases. The industry converts roughly one fifth of gross margin into operating profit — operating expense eats the other 80%. That ratio is the physics of this business and we should assume it applies to us.
Because we are single-category and selling only to independents — the highest-margin channel in foodservice — we should target 22–28% gross margin, not 18%. If the distributor tier will not support 22%, the route economics in Section 07 do not cover their own overhead.
Five phases, each with a gate that can stop the next one
Nothing here is a milestone we hope to hit. Each phase ends in a binary test. If a gate fails, the correct action is to stop and either fix the input or shut the project — not to proceed on optimism.
Phase 0 — Cost basis Weeks 1–3 · ~$0
- Call Wayne Clayton (Dawn) and Peter Song (BakeMark). Obtain written distributor-tier price books and the volume commitment each requires.
- Call at least two more: Ardent Mills (flour), Ventura / Stratas Foods (shortening), and a sugar refiner rep (Imperial or Domino). These are the four commodity lines where a manufacturer-direct tier is most likely to exist.
- Have someone call Texas Bakery Supply as an ordinary shop and verbally confirm three shelf prices against our scrape.
- Re-run the Section 02 table on real numbers. Publish the result.
Gate 0 — the kill gate
Landed cost on the core basket must come in at ≤80% of verified TBS shelf, and the resulting gross margin at a 10%-under-TBS sell price must be ≥20%. If landed cost lands above 85% of TBS shelf, stop. Take the Golden Glaze savings from Section 01 and close the file.
Phase 1 — Internal proof Months 1–3 · ~$15K
Before selling a single outside shop, run Golden Glaze's own seven stores as the first customer. This is the entire de-risking move and it is available to us and to nobody else.
- Buy the top 10 SKUs at the distributor tier. Receive into one Golden Glaze location. Distribute to the other six on the existing inter-store run.
- Measure, do not estimate: actual landed cost including freight and pallet fees, actual shrink, actual turn rate per SKU, actual receiving labor minutes.
- Stand up the ordering pipe: shop texts → structured order on our side → auto-confirm with line-item total. Test it internally where a bug costs nothing.
- Register the entity, obtain the Texas sales-tax permit, file Form 01-339 with every supplier, and register with FDA (free).
Gate 1
Realized landed cost is within 5% of the quoted distributor sheet after freight and fees. Golden Glaze's ingredient cost line falls measurably. The order pipe runs one full month with zero missed deliveries and zero transcription errors. If freight and minimums eat the discount, the tier was theatre — go back to Gate 0.
Phase 2 — Beachhead Months 4–9 · ~$90K + working capital
- Work the named 30-shop West Fort Worth / Camp Bowie list in Section 09. Target 10 paying accounts.
- Buy one used 16-ft box truck with a liftgate. Hire one non-CDL driver. Deliver twice weekly.
- Ship the published price sheet on the core SKUs only. No custom pricing, ever, for anyone.
- Collect a completed Form 01-339 resale certificate at onboarding — no certificate, no account. Four-year retention.
- Run credit properly from account one: net-15, personal guarantee on the first three orders, hard stop at 30 days past due.
Gate 2
10 active accounts, ≥20% realized gross margin after freight and shrink, and bad debt under 3% of receivables. Contribution per stop must be positive at the measured drop size. If we cannot convert 10 of 30 shops that are physically 8 miles from nothing we own, the pitch is wrong — fix the pitch before buying a second truck.
Phase 3 — Route density Months 10–18 · ~$120K + working capital
- Fill the Camp Bowie route to 25–30 accounts before opening a second geography. Density, not coverage, is what pays.
- Open route two in NE Tarrant (Haltom City / North Richland Hills / North Fort Worth) — the densest cluster in DFW, deliberately held back because it sits 1.6 miles from the Golden Glaze North Richland Hills shop and the competitor objection is hardest there. By month 10 we will have references to answer it with.
- Second truck, second driver, first warehouse associate.
- Expand the basket only into items existing customers already ask for. Do not add SKUs speculatively.
Gate 3
50 active accounts. Positive EBITDA at the route level. Cash conversion cycle measured and stable. This is the point at which a standalone facility first pencils — see Section 08.
Phase 4 — Facility and scale Months 19–36
- Standalone 8,000–12,000 SF warehouse, or the Golden Glaze centralized kitchen if that raise has closed and the building can carry both.
- Third truck. Warehouse supervisor. Dedicated AR and collections.
- Push toward 120–150 accounts, which is where the model produces real operating profit.
What it costs to open, line by line
Two builds. The lean build rides Golden Glaze's existing footprint and is what Phase 2 actually requires. The standalone build is what Phase 4 looks like. Every figure is sourced; assumptions are marked.
Build A — Lean launch (Phase 2, 10–30 accounts, riding Golden Glaze)
| Line item | Low | High | Note |
|---|---|---|---|
| Used 16-ft box truck, liftgate, ~5 yr old | $33,000 | $45,000 | Verified 2020 Isuzu NPR-HD listing at $33,252. Non-CDL at 14,500 GVWR. |
| Truck prep — wrap, shelving, hand trucks, straps | $3,500 | $7,000 | Estimate |
| Pallet rack, ~120 positions, installed | $7,800 | $12,600 | $65–$105/position installed |
| Used forklift or electric pallet jack | $3,000 | $18,000 | Pallet jack $3–8K; used forklift toward the high end |
| Texas DSHS food wholesaler license (2-yr) | $250 | $900 | Fee tiers by gross food sales. $250 under $200K. |
| Texas DSHS warehouse operator license (2-yr) | $350 | $700 | $350 under 6,000 SF |
| FDA food facility registration | $0 | $0 | No fee. Renews every two years; 2026 window opens 1 Oct. |
| Entity, legal, contracts, credit application forms | $3,000 | $6,000 | Estimate |
| Ordering pipe + accounting + routing software | $4,000 | $12,000 | Built on Golden Ops. Mostly internal build cost. |
| Opening inventory — 10 core SKUs, ~3 weeks cover | $25,000 | $45,000 | See material build-up below |
| Incremental space at a Golden Glaze site | $0 | $24,000 | Internal charge if any |
| Fixed + inventory subtotal | $79,900 | $171,200 | |
| Working capital reserve (see §10) | $120,000 | $180,000 | ~63-day cash conversion cycle at ~$1M revenue |
| TOTAL — Build A | $199,900 | $351,200 |
Build B — Standalone facility (Phase 4, 70+ accounts)
| Line item | Low | High | Note |
|---|---|---|---|
| 2× used 16-ft box trucks w/ liftgate | $66,000 | $90,000 | |
| Pallet rack, ~600 positions, installed | $39,000 | $63,000 | |
| Forklift + electric pallet jack | $25,000 | $63,000 | New IC forklift $20–50K; electric $22–55K plus $7.5–9.5K battery |
| Lease deposit — 2 to 6 months gross rent | $23,000 | $85,000 | Assumption. Not sourced — expect a personal guarantee as a no-history tenant. |
| Tenant improvements net of allowance | $15,000 | $60,000 | Industrial TI allowance $5–15/SF on 5-yr terms |
| Licensing — wholesaler at $1M+ tier + warehouse 6–24K SF | $1,600 | $1,600 | Two-year term, i.e. $800/yr |
| Opening inventory — expanded basket | $110,000 | $180,000 | |
| Systems, WMS, integration | $20,000 | $45,000 | |
| Fixed + inventory subtotal | $299,600 | $587,600 | |
| Working capital reserve | $400,000 | $550,000 | ~63-day CCC at ~$3M revenue |
| TOTAL — Build B | $699,600 | $1,137,600 |
The single largest capital lever we control
Truck finance is priced off operating history. A new business with excellent credit and no industry experience pays roughly $2,800/month on a 60-month note for a new 26-ft box truck; an established company with two years of history pays about $2,420. Thin credit pushes it to $3,450+.
Put Golden Glaze's seven-store operating history behind the truck paper rather than a newco. That is worth $380–$1,030 per truck per month — $68K to $185K across three trucks over five years. Raise it explicitly with the lender.
Material — opening inventory build-up
Ten SKUs is the right opening basket. It covers the overwhelming majority of an independent donut shop's daily burn and keeps working capital small enough to survive a slow start.
| Core SKU | Why it is in the opening ten |
|---|---|
| Yeast-raised donut mix, 50# | Highest-volume single line. The anchor SKU — whoever supplies mix supplies everything else. |
| Cake donut mix, 50# | Second-highest volume. Different buyer preference by shop; carry two brands. |
| Donut fry shortening, 50# | Heavy, bulky, burned continuously. High freight cost is a barrier to shops buying it themselves — a real service value. |
| Glaze / icing sugar, 50# | Where Golden Glaze is 49% over TBS shelf today. Highest-leverage line to fix. |
| Powdered sugar 6X / 10X / 12X, 50# | Three grades, one supplier relationship. We already buy 6X and 10X well. |
| Granulated sugar, 50# | Commodity. Low margin, but nobody switches suppliers for nine items out of ten. |
| Bread / donut flour, 50# | Lowest margin in the basket at ~5%. Carried for completeness, not profit. |
| Instant dry yeast, 20# | Small, light, high value per pound — the best freight economics in the basket. |
| Fillings — bavarian creme, lemon, apple | Higher margin than dry goods. Shops buy on brand preference, which reduces pure price shopping. |
| Donut boxes, bags, tissue | Non-food, no DSHS exposure, decent margin, and every shop needs them weekly. |
Do not break bulk. Not in year one.
A warehouse that stores only sealed, shelf-stable manufacturer packages qualifies for modified requirements under FDA Preventive Controls for Human Food (21 CFR 117). The moment we repack a 50# flour sack into 10# bags, private-label anything, or perform any "value-added service," we convert into a full PCHF facility requiring a written food safety plan and a certified PCQI on staff.
That is a real, ongoing cost for a marginal revenue idea. Sell only sealed manufacturer packages in year one. Revisit repacking when there is a customer paying for it.
What it costs to run, per year
Labor — fully loaded, Texas 2026
Texas is a cheap payroll state: no state income tax, unemployment tax capped at the first $9,000 of wages, workers' comp not mandatory. Health insurance, not taxes, is 60–75% of the entire burden load.
| Role | Base | FICA | Workers' comp | SUTA+FUTA | Health | Loaded | Multiple |
|---|---|---|---|---|---|---|---|
| Non-CDL driver @ $20.38/hr | $42,390 | $3,243 | $2,844 | $314 | $7,885 | $56,676 | 1.34× |
| Non-CDL driver, no benefits | $42,390 | $3,243 | $2,844 | $314 | $0 | $48,791 | 1.15× |
| Warehouse associate @ $17.20/hr | $35,776 | $2,737 | $676 | $314 | $7,885 | $47,388 | 1.32× |
| Warehouse associate, no benefits | $35,776 | $2,737 | $676 | $314 | $0 | $39,503 | 1.10× |
| Warehouse supervisor @ $27.73/hr | $57,683 | $4,413 | $1,090 | $314 | $7,885 | $71,385 | 1.24× |
| CDL-A driver @ $32.42/hr (avoid) | $67,433 | $5,159 | $4,525 | $314 | $7,885 | $85,316 | 1.27× |
Two labor decisions that are worth real money
Stay non-CDL. A CDL is required at 26,001 lb GVWR. Manufacturers deliberately spec 26-ft boxes at 25,950 lb precisely to stay under it — a 2026 Isuzu FTR with a 26-ft box and a 3,300 lb liftgate is 25,950 GVWR and runs on an ordinary Texas Class C licence. Payload is still ~10,000–12,000 lbs, which at 40–50 lbs a case is 200–300 cases. Plenty. Staying under also keeps us below the TxDMV motor carrier registration trigger, and a non-CDL driver costs $56,676 loaded against $85,316 for CDL-A. Do not default to a Freightliner M2 — it is normally spec'd over the line.
Classify carefully. Workers' comp for drivers (class 7380) is $6.71 per $100 of payroll. Warehouse wholesale (class 8018) is $1.89 — 3.55× cheaper for the same wage. Because we own the inventory we store, we are 8018, not the higher warehousing code 8292. Get that classification right on the application.
Vehicle — per truck, per year
| Line | Figure | Note |
|---|---|---|
| Commercial auto insurance | $6,000–$8,500 | Local <50-mi DFW radius, Class 5–6, non-CDL. Do not model the $4,500 floor in year one with no loss history. Premiums +6.4% industry-wide in Q1 2026. |
| Finance — used truck, newco tier | ~$16,800 | ~$1,400/mo. Roughly half that if Golden Glaze's history backs the paper. |
| Fuel | $0.53/mi | Gas at $3.69/gal ÷ 7 mpg loaded |
| Repair & maintenance | $0.19–$0.22/mi | ATRI South-Central region. Rising fast — R&M was +8.6% in 2025. |
| Variable, all-in | $0.72–$0.75/mi |
Spec gas, not diesel
Dallas diesel is $4.96/gal as of 27 July 2026 — up 52.4% year over year. Texas regular gasoline is $3.69. That $1.27 spread (34%) more than erases diesel's roughly 1.5 mpg advantage on a stop-and-go metro route. Run the gas V8 Isuzu NPR-HD unless annual miles per truck exceed about 30,000 — and on a 44-mile route run twice a week, we will be nowhere near that.
This is also the single biggest macro risk in the model. Fuel is the one input we cannot hedge and it moved 52% in a year.
Facility and overhead
| Line | Annual | Note |
|---|---|---|
| Occupancy — 10,000 SF standalone, all-in | $140,000–$170,000 | Base $11.50–13.50/SF NNN for small-bay in Garland / Irving / Carrollton / Arlington, plus $2.50–3.50/SF NNN load. Texas property tax is the largest component and it reassesses. |
| General liability + product liability | $5,000–$15,000 | Judgment, not sourced. Published "$525/yr food business" quotes price a farmers-market booth, not a wholesale distributor with a fleet. Get three real broker quotes. |
| DSHS licensing | $800 | Two licenses on two-year terms |
| Software — ordering, routing, accounting | $9,000 | Built on Golden Ops |
| Admin / AR / billing, loaded | $65,000 | Phase 3 onward |
DFW is not a tenant's market, whatever the national narrative says
DFW industrial vacancy is falling — 9.3%, down 100bps year over year — on the highest net absorption of any US industrial market (17.9M SF in H1 2026). JLL states plainly that concessions are holding stable rather than widening. And small-bay, which is our size band, is the tightest segment in the country at 4.2% vacancy; sub-10,000 SF asking rents are up more than 40% since 2020.
We are the weaker party in that negotiation. Which is another argument for Section 08.
Do not build cold storage
Refrigerated construction runs $185–$235/SF and freezer $245–$330/SF; DFW-specific cold builds with insulated metal panel come in at $275–$495/SF. A 1,000 SF freezer box is $245K–$330K of capital before we know our shortening turn rate.
Renting cold is $35–$75 per pallet per month. Forty pallets is $17K–$36K a year — a 8-to-16-year break-even against building. Rent cold, own dry. And note that shortening and most donut fillings are shelf-stable at controlled ambient anyway; a conditioned dry room with insulation and a mini-split costs a few thousand dollars, not a cold-storage buildout.
The price sheet, and how it gets built
Pricing rules — non-negotiable
- One published price. No exceptions, no volume tiers, no rep discretion. Not because transparency wins deals — the incumbent already publishes — but because it removes the entire cost of a negotiating sales force, which is what makes the low price affordable in the first place.
- Price against verified TBS shelf, per pound, not per case. Pack sizes differ across 20#, 25#, 36#, 37#, 38#, 40#, 45# and 50#, and case-to-case comparison is how you accidentally sell below cost.
- Minimum order $250. Sysco and US Foods sit at $250–$500 for independent restaurant accounts. Below $250 the stop does not cover itself.
- Reprice monthly, publish the change a week ahead. Commodity inputs move. Absorbing swings quietly is how distributors discover they have been losing money for a quarter.
- Never quote a price Jack has not approved. Standing rule.
Required landed cost by SKU, to sell 10% under TBS at 20% GM
| Item ($/lb) | GG pays now | TBS shelf | Our price | Required cost | Cost gap to close |
|---|---|---|---|---|---|
| Pillsbury raised donut mix 50# | 0.760 | 0.540 | 0.486 | 0.389 | −48.8% |
| Glaze / icing sugar 50# | 1.100 | 0.559 | 0.503 | 0.402 | −63.4% |
| White roll icing, RTU | 2.798 | 1.433 | 1.290 | 1.032 | −63.1% |
| Powdered sugar 12X 50# | 0.845 | 0.529 | 0.476 | 0.381 | −54.9% |
| Instant dry yeast 20# | 3.495 | 2.647 | 2.382 | 1.906 | −45.5% |
| Lemon filling ~37# | 1.314 | 1.135 | 1.022 | 0.817 | −37.8% |
| Bavarian creme ~37# | 1.208 | 1.103 | 0.993 | 0.794 | −34.3% |
| Donut fry shortening 50# | 0.987 | 0.920 | 0.828 | 0.662 | −32.9% |
| Bread / donut flour 50# | 0.358 | 0.339 | 0.305 | 0.244 | −31.8% |
| Pillsbury Tender Taste base 50# | 0.840 | 0.910 | 0.819 | 0.655 | −22.0% |
| Powdered sugar 10X 50# | 0.699 | 0.836 | 0.752 | 0.602 | −13.9% |
| Powdered sugar 6X 50# | 0.679 | 0.926 | 0.833 | 0.667 | −1.8% |
| Granulated sugar 50# | 0.690 | 0.739 | 0.665 | 0.532 | −22.9% |
| Apple filling | 1.981 | 2.172 | 1.955 | 1.564 | −21.1% |
| Basket (unweighted) | 17.754 | — | — | 10.647 | −40.0% |
Read this table as the shopping list for the Phase 0 phone calls. On powdered 6X we are already within 2% of where we need to be — that line works today. On glaze sugar and roll icing we need to cut landed cost by roughly 63%, which almost certainly means changing brand, not just tier. Those two SKUs are where the negotiation should start, and they are also the two biggest immediate savings for Golden Glaze regardless of what happens to Golden Star Supply.
Margin gates, restated
| Gross margin achieved | What it means |
|---|---|
| Under 15% | Stop. Below the Phase 0 kill gate. Route costs do not clear. |
| 15–20% | Broadline territory. We would be a worse Sysco with one truck. Only viable if volume is enormous, which it is not. |
| 20–25% | Workable. Section 07 shows this covers a route and its overhead at 30+ accounts. |
| 25–30% | Target. Appropriate for a single-category distributor selling only to independents. |
| Over 30% | Check the numbers — either the cost basis is wrong or we are not actually cheaper than TBS. |
Cost per stop, and what a stop has to be worth
This is the operating heart of a distribution business. Everything else is a support function for putting a case of mix on a shelf at 5 a.m. for less than the margin on it.
Cost per stop — built from sourced inputs
| Component | Per stop |
|---|---|
| Driver labor | $9.77 |
| Fuel + maintenance | $1.10 |
| Truck fixed (insurance + finance), allocated | $7.71 |
| Total delivery cost per stop | $18.58 |
Route time works out to 10.8 hours per run — which is a full day and then some. At 30 accounts on two runs a week that is 1,119 driver-hours a year, comfortably inside one truck and one driver. But note that a single 30-stop run does not fit in an eight-hour day: split it into two runs of fifteen, on different days. That is also better for the customer, who would rather receive twice than once.
Break-even revenue per stop
Break-even revenue per stop is simply delivery cost divided by gross margin. This is why the margin gate matters so much:
| Gross margin | Break-even revenue per stop | Contribution at a $300 drop |
|---|---|---|
| 15% | $124 | $26.42 |
| 20% | $93 | $41.42 |
| 25% | $74 | $56.42 |
| 30% | $62 | $71.42 |
At a $250 order minimum and twice-weekly delivery, an average shop spending $600 a week produces a $300 drop — well clear of break-even at any margin above 15%. Drop size is not the constraint. Account count is.
The scale curve, standalone facility
| Accounts | Revenue | Gross profit | Delivery cost | Trucks | Overhead | EBITDA |
|---|---|---|---|---|---|---|
| 30 | $0.94M | $234,000 | $57,974 | 1 | $287,188 | −$111,162 |
| 50 | $1.56M | $390,000 | $76,352 | 1 | $287,188 | $26,460 |
| 70 | $2.18M | $546,000 | $118,781 | 2 | $405,961 | $21,258 |
| 90 | $2.81M | $702,000 | $137,160 | 2 | $405,961 | $158,879 |
| 120 | $3.74M | $936,000 | $164,728 | 2 | $405,961 | $365,311 |
| 150 | $4.68M | $1,170,000 | $216,346 | 3 | $453,349 | $500,305 |
The dead zone: 30 to 70 accounts
A standalone facility does not break even until roughly 48 accounts, and then immediately dips again at 70 when the second truck and the supervisor step in. Between 30 and 70 accounts a standalone build burns cash continuously.
Any plan that opens with a leased warehouse and 10 customers is a plan to fund 18 months of losses through the hardest part of the curve. That is how this business fails, and it is entirely avoidable — see the next section.
Golden Glaze is the anchor customer. Use it.
This is the most important structural decision in the plan and it is worth more than every negotiation in it combined.
Golden Glaze's seven shops already consume a measured ~$310K a year of distributor purchases — roughly $850 per shop per week, which is a larger drop than a typical independent, with zero acquisition cost, zero credit risk, and no switching decision to win. Golden Glaze is not a hypothetical first customer. It is an existing book of business we are currently handing to somebody else.
| GM | Account mix | Revenue | Gross profit | EBITDA |
|---|---|---|---|---|
| 15% | 7 Golden Glaze only | $0.31M | $46,410 | −$32,390 |
| 15% | 7 GG + 15 independents | $0.78M | $116,610 | $23,739 |
| 15% | 7 GG + 30 independents | $1.25M | $186,810 | $79,867 |
| 20% | 7 Golden Glaze only | $0.31M | $61,880 | −$16,920 |
| 20% | 7 GG + 15 independents | $0.78M | $155,480 | $62,609 |
| 20% | 7 GG + 30 independents | $1.25M | $249,080 | $142,137 |
| 25% | 7 Golden Glaze only | $0.31M | $77,350 | −$1,450 |
| 25% | 7 GG + 30 independents | $1.25M | $311,350 | $204,407 |
| 25% | 7 GG + 50 independents | $1.87M | $467,350 | $317,596 |
A $253,000 swing from one decision
At 30 independent accounts, a standalone facility loses $111,162. The same 30 accounts plus Golden Glaze's own seven, riding the existing footprint, makes $142,137. Same customers, same trucks, same margin — the entire difference is whether we sign a warehouse lease before we need one.
Note too that Golden Glaze alone is roughly break-even at 25% margin. The internal capture pays for the infrastructure. Every independent account after that is close to pure contribution. That is the structure that makes this fundable.
Sequencing the building
- Phases 1–2: receive into an existing Golden Glaze location. No lease.
- Phase 3: if the Golden Glaze $1M raise and centralized kitchen proceed, that building is the same dock, the same rack, and the same route a distributor needs. Design it for both from the start — the marginal cost of adding distribution capacity to a kitchen already being built is a fraction of a standalone lease.
- Phase 4: standalone lease only above ~70 accounts, and only if the kitchen cannot carry it.
Two governance points that must be handled properly
Intercompany pricing. Golden Capital Management already runs a "4102 Supply Management" account at $143,822.21 for Jan–Jul 2026 — a central purchasing function already exists. Golden Star Supply selling to Golden Glaze is a related-party transaction. Price it at the published list, document it, and keep it clean. Golden Glaze has Reg D 506(b) investors and this will be looked at.
There is already one external customer on the books — JCT Donuts LLC, $3,550.00 on 9 February 2026. Worth understanding how that happened before designing the sales motion, because it is evidence the demand is real and it arrived without anyone selling.
Thirty shops, by phone and in person, in one geography
The list is smaller than v0.1 said
The v0.1 figure of ~945 independents was too high. After removing franchise units with mandated supply and de-duplicating shops that appear twice under different city labels with near-identical coordinates, the serviceable list is 755. Two data notes that matter operationally: only 191 rows carry a phone number, and the City column is unreliable — one Rufe Snow Drive address is labelled "Grand Prairie" when it is physically in North Richland Hills. Trust latitude, longitude, and street address only.
This is a phone-and-walk-in channel
Only 185 of the original 1,669 shops have a website at all. Email is not the channel. The owner is 55, has been awake since 2 a.m., and answers the shop phone himself between 5 and 10 a.m. Phone coverage is our real asset gap — 191 numbers against 755 shops. The Google Places API would fill it for roughly $40, and that remains an open ask; no billed API gets enabled on Jack's GCP project without an explicit yes.
In the interim, the beachhead has 22 of 30 shops with phone numbers — the best coverage of any cluster in DFW, which is part of why it was chosen.
Why Camp Bowie and not the denser cluster
| Cluster | Shops | Radius | Phones | Nearest GG shop | Decision |
|---|---|---|---|---|---|
| West Fort Worth / Camp Bowie | 30 | 4.79 mi | 22/30 | 8.6 mi | Chosen |
| NE Tarrant — Haltom City / NRH | 30 | 3.95 mi | 21/30 | 1.6 mi | Deferred to route 2 |
The NE Tarrant cluster is tighter and would run about six fewer road miles. It also sits 1.6 miles from the Golden Glaze North Richland Hills shop, where the sentence "we will not open a shop near you" is worthless because we already did. At 30 accounts the binding constraint is not six miles of diesel — it is pitch conversion. Prove the pitch where the competitor objection is weakest, then bring the references to the hard neighbourhood as route two. NE Tarrant becomes margin expansion, not the experiment.
The pitch
Order of arguments — and it is deliberately not price first
- Service window. "Your supplier closes at 3:30 and doesn't open Saturday. You bake at 2 a.m. and you run out on Sunday morning." This is the opening line because it is a real, verifiable gap and it is not a price argument the incumbent can simply match on a sheet.
- Operator credential. "I run seven donut shops. I've fried at 3 a.m. I know what happens when the mix doesn't show up." Nobody at Texas Bakery Supply can say this.
- Price, stated flatly and only third. "Here's the sheet. Same price for you as for everybody. Compare it line by line." Only make this argument if Phase 0 delivered a real discount. Never claim to be cheaper without checking the specific line — they publish, and the owner can check in thirty seconds.
- Zero switching cost. "Keep your current guy. Try me on mix and shortening only. If I'm late once, fire me."
The objection that decides everything — rehearse this
"You own donut shops. Why would I buy from my competitor?"
It will come up in the first two minutes of most conversations. The answer has three parts and all three must be said:
- Separate brand and separate entity. Golden Star Supply, not Golden Glaze. Nothing customer-facing carries a goldenglazedonuts.com URL.
- A written non-encroachment commitment. We will not open a Golden Glaze location within a defined radius of an active supply customer, for as long as they are a customer. Put it in the account agreement. It costs us little — we are not opening in West Fort Worth — and it is the only thing that actually resolves the objection.
- Reframe the credential. "I'm the only supplier in this town who has actually run the machine you're running. That's why my delivery window matches your production window and theirs doesn't."
Cadence — the first ninety days
| Week | Action | Target |
|---|---|---|
| 1–2 | Call all 22 shops with numbers, 5:30–9:00 a.m. Goal is not a sale — it is the owner's name and their current supplier. | 22 called, 15 owner names |
| 3–4 | Walk in on all 30, including the 8 without phones. Bring donuts. Leave a one-page price sheet with a mobile number on it. | 30 visited, 20 sheets left with a decision-maker |
| 5–6 | Follow up on the walk-ins. Ask for a trial on two SKUs only — mix and shortening. | 8 trial commitments |
| 7–10 | Deliver trials personally. Jack on the truck, not a driver. Every complaint gets fixed the same day. | 8 trials delivered, 5 converting |
| 11–13 | Convert trials to standing orders. Ask each convert for one referral — donut shop owners in DFW know each other, and a large share are part of the same community networks. | 10 active accounts |
Jack runs the first ten accounts personally
Do not hire a sales rep for the beachhead. The pitch is unproven, the price sheet is unproven, and the objection handling is unrehearsed. A rep will fail at all three and we will not learn why. The operator credential is also the pitch — it does not transfer to a hired rep in month one.
Hire the rep at Gate 2, once there are ten reference accounts and a script that has been proven to work.
Account onboarding checklist — every account, no exceptions
- Completed Form 01-339 resale certificate — legal name and address exactly as registered, 11-digit Texas sales tax permit number. No certificate, no account. We carry the audit risk if it is defective and we must retain it four years. This is the single most common sales-tax audit failure in wholesale distribution.
- Signed credit application with a personal guarantee. Net-15 to start.
- Signed non-encroachment addendum.
- Delivery window, gate/door access, and a mobile number that gets answered at 4 a.m.
- Standing order template, so the weekly text is a confirmation not a fresh order.
The beachhead — named list
| Shop | Address | ZIP | Phone |
|---|---|---|---|
| Dough Boy Donuts | 4916 Camp Bowie Boulevard | 76116 | — |
| Dales Donuts #9 | 4455 Camp Bowie Blvd #120 | 76107 | (817) 737-9979 |
| Happy Donut Shop | 109 Roberts Cut Off Rd | 76114 | (817) 569-0833 |
| A & H Donuts | 3412 W 7th St | 76107 | (817) 882-9011 |
| Momma's Donuts | 5071 River Oaks Blvd | 76114 | (817) 624-1434 |
| K Donut | 2013 8th Ave | 76110 | (817) 921-9583 |
| Dreamboat Donuts at Spiral Diner | 1314 W Magnolia Ave | 76104 | (817) 332-8834 |
| Donuts Plus | 2201 Jacksboro Hwy Ste 109 | 76114 | (817) 624-4502 |
| Batter and Beans - Cake Donuts and Premium Coffee | 3548 S Hills Ave #18 | 76109 | (682) 224-0749 |
| Circle Donuts | 3506 Blue Bonnet Cir | 76109 | (817) 927-7238 |
| Manna Donuts TCU | 2211 W Berry St | 76110 | (817) 207-0336 |
| Donut Express | 2606 S Cherry Ln C | 76116 | (817) 696-9415 |
| Momma's Donuts🍩 | 823 S Cherry Ln | — | |
| Yummy Yummy Donuts | 2621 Jacksboro Hwy Ste E | 76114 | (817) 378-0188 |
| Super Donuts | 571 W Northside Dr | 76164 | (817) 625-5212 |
| Duke's Donuts | 500 S Cherry Ln | 76108 | (817) 246-0908 |
| Tom’s Donuts | 7933 Camp Bowie W Blvd | — | |
| K Donuts | 4605 Benbrook Blvd #101 | — | |
| Mr. Bakery Donuts | 4400 Benbrook Highway | 76116 | (817) 420-6166 |
| FunkyTown Donuts and Drafts | 132 E 4th St | — | |
| Mousa Donut | 201 W Rosedale St Unit 757 | 76104 | (682) 404-4176 |
| Dusty Biscuit Beignets | 411 S Main St Unit 109 | 76104 | (817) 841-9255 |
| Donuts | 1950 South Las Vegas Trail | 76108 | — |
| Papa Yun's Donuts - Fort Worth | 4861 Bryant Irvin Rd | 76132 | (817) 423-0164 |
| Fort Worth Donuts | 1612 South Fwy | — | |
| Angel's Donuts | 4957 S Hulen St | 76132 | (682) 703-2298 |
| Manna Donuts McCart | 4246 McCart Ave | 76115 | (817) 921-1761 |
| Wonder Donut | 5208 Wonder Dr | — | |
| Lim's Donuts | 9078 Camp Bowie W Blvd | 76116 | (817) 244-0323 |
| Dale's Donuts | 5515 S Hulen St | 76132 | (817) 370-9863 |
Thirty shops, 4.79-mile radius, 22 with phone numbers, ~44 road miles for the full loop, 8.6 miles from the nearest Golden Glaze. Eight have no number and must be walked into. Source rows are in the DFW lead sheet — trust the street address, not the City column.
The thing that actually kills distributors
Distribution does not fail on gross margin. It fails on the gap between paying for the pallet and getting paid for it. At 20–25% margin, every dollar of growth consumes cash before it produces any.
| Component | Days | Driver |
|---|---|---|
| Days inventory outstanding | 30 | Three weeks' cover on ten SKUs plus safety stock. Dry goods, so no spoilage clock — but every extra week of cover is cash on a rack. |
| Days sales outstanding | 33 | Net-15 stated. Independent shops pay late; 33 days is the realistic collected number, not the invoice term. |
| Days payable outstanding | −15 | What a new distributor gets. Established accounts get 30. We will not, at first. |
| Cash conversion cycle | 48 days | Stretches to ~63 if suppliers demand prepay or COD in year one, which is the case to plan for. |
| Annual revenue | Working capital at 48-day CCC | At 63-day CCC |
|---|---|---|
| $0.94M — 30 accounts | $124,000 | $162,000 |
| $1.25M — 7 GG + 30 independents | $164,000 | $216,000 |
| $1.87M — 7 GG + 50 independents | $246,000 | $323,000 |
| $3.74M — 120 accounts | $492,000 | $645,000 |
Credit discipline from account one, not from the first bad debt
- Net-15, personal guarantee on the first three orders. Independent donut shops are thin-margin cash businesses and some of them fail. This is not distrust, it is the standard of the category.
- Hard stop at 30 days past due. Written into the credit application, applied without exception, including to friends. The distributor that keeps delivering to a slow payer is financing a business it does not own.
- Bad debt above 3% of receivables is a Gate 2 failure, not a cost of doing business.
- Take card payment and eat the fee. ~3% on a 22% margin costs 3 points of gross but converts 33 DSO to 2. That trade is worth making on any account that has ever paid late.
- Never fund growth out of the Golden Glaze operating account. Separate entity, separate bank, separate line. Golden Glaze has Reg D investors and commingling working capital is exactly the thing that gets asked about.
The internal book funds the float
Golden Glaze's own seven shops are ~$310K/yr of revenue with zero DSO and zero bad-debt risk — it is an intercompany transfer. That book alone shortens the blended cash cycle materially and is the single cheapest source of working capital in the plan. It is another reason Phase 1 is internal-only.
Who does the work, and when they get hired
The failure mode here is hiring a sales rep and a warehouse manager before there is a proven pitch or a second truck. Every role below is tied to a gate, not to a month.
| Role | Hired at | Loaded cost | Why not earlier |
|---|---|---|---|
| Jack — sales, pricing, first ten accounts | Phase 0 | — | The operator credential is the pitch. It does not transfer to a hire in month one. |
| Non-CDL driver #1 | Phase 2, at ~8 accounts | $56,676 | Below 8 accounts Jack drives. Riding the truck is also how the service failures get found. |
| Warehouse associate #1 | Phase 3, at ~50 accounts | $47,388 | One associate carries roughly 60 accounts of receiving and picking. |
| Sales rep #1 | Gate 2 cleared | Base + commission | Only once there are ten reference accounts and a script proven to convert. |
| Driver #2 + truck #2 | Phase 3, at ~70 accounts | $56,676 | One truck carries ~60–70 accounts at two drops a week. |
| Warehouse supervisor | Phase 4, at ~60–70 accounts | $71,385 | This is the hire that causes the EBITDA dip at 70 accounts. Delay it as long as the floor allows. |
| AR / admin | Phase 3–4 | $65,000 | Collections is the highest-ROI back-office hire in distribution. Do not leave it on Jack past 50 accounts. |
Two structural rules
Stay non-CDL for as long as physically possible. $56,676 versus $85,316 loaded, a shallower hiring pool problem, and no TxDMV motor carrier registration. Spec every truck under 26,001 lb GVWR deliberately.
Classify drivers and warehouse staff separately on the workers' comp application. Class 7380 (drivers) is $6.71 per $100 of payroll; class 8018 (wholesale warehouse) is $1.89. Because we own the inventory we store, warehouse staff are 8018, not 8292. Getting this wrong costs multiples of the premium it saves.
What has to be filed before the first pallet moves
| Requirement | Cost | When | Note |
|---|---|---|---|
| Texas entity formation + registered agent | ~$300 + legal | Phase 1 | Separate entity from Golden Glaze. Non-negotiable for the competitor objection and for clean related-party accounting. |
| Texas sales & use tax permit | $0 | Phase 1 | 11-digit number. Required before issuing or accepting resale certificates. |
| Form 01-339 issued to our suppliers | $0 | Phase 1 | So we buy for resale tax-free. |
| Form 01-339 collected from every customer | $0 | Every account | Four-year retention. We carry the audit liability if a certificate is defective. This is the most common sales-tax audit failure in wholesale distribution. |
| Texas DSHS food wholesaler license | $250–$900 | Phase 1 | Two-year term, fee tiered on gross food sales. $250 under $200K. |
| Texas DSHS warehouse operator license | $350–$700 | Phase 1 | $350 under 6,000 SF. Required for the storage site even if it is inside a Golden Glaze building. |
| FDA food facility registration | $0 | Phase 1 | No fee. Biennial renewal; the 2026 window opens 1 October. |
| Commercial auto insurance | $6,000–$8,500/truck/yr | Before truck #1 | Get three real quotes. Do not model the $4,500 floor with no loss history. |
| General + product liability | $5,000–$15,000/yr | Before first outside sale | Judgment, not sourced. Get three broker quotes. |
| Workers' comp (optional in TX, take it anyway) | Class-rated | First hire | 7380 drivers $6.71 / 8018 warehouse $1.89 per $100 payroll. |
| TxDMV motor carrier registration | Avoided | — | Triggered above 26,000 lb GVWR. Staying non-CDL avoids it entirely. |
The one compliance decision that changes the cost structure
Storing only sealed, shelf-stable manufacturer packages keeps us under FDA modified requirements (21 CFR 117). Repacking a 50# sack into 10# bags, private-labelling, or any "value-added service" converts us into a full Preventive Controls facility requiring a written food safety plan and a certified PCQI on staff. Sell sealed packages only in year one.
Related-party exposure
Golden Capital Management already books $143,822.21 to "4102 Supply Management" for Jan–Jul 2026 — the central purchasing function exists today. Golden Star Supply selling to Golden Glaze is a related-party transaction inside a business with Reg D 506(b) investors. Price it at published list, paper it like a third-party sale, and disclose it. Do not let it become an audit finding two years from now.
The conditions under which we stop
Written down in advance so the decision is not made in the middle of it, with capital already spent and pride attached.
| Gate | Test | Fail action |
|---|---|---|
| Gate 0 — Cost basis Unpassed | Landed cost ≤80% of verified TBS shelf and ≥20% GM at a 10%-under sell price. | Above 85% of shelf → stop. Bank the Golden Glaze savings and close the file. |
| Gate 1 — Internal proof | Realized landed cost within 5% of the quoted sheet after freight and fees. One clean month of the order pipe: zero missed deliveries, zero transcription errors. | Freight and minimums ate the discount → the tier was theatre. Return to Gate 0. |
| Gate 2 — Beachhead | 10 active accounts, ≥20% realized GM after freight and shrink, bad debt <3% of receivables, positive contribution per stop. | Under 10 of 30 → the pitch is wrong. Fix the pitch before buying a second truck. |
| Gate 3 — Density | 50 active accounts, positive route-level EBITDA, measured and stable cash conversion cycle. | Do not sign a facility lease. Hold at one route. |
Risks that are not gates
| Risk | Why it matters | Mitigation |
|---|---|---|
| The distributor tier does not exist at the required level | This is the whole business. We need a ~35–40% landed-cost reduction from where Golden Glaze buys today. | Two phone calls, this week. It is the cheapest possible test and it sits above everything else. |
| The TBS price book is not a live quote | 920 of 929 products render OutOfStock. If their real prices are higher, our gap is smaller; if lower, worse. | Mystery-shop three SKUs by phone before pricing against it. |
| The competitor objection does not clear | Every prospect knows Jack owns seven donut shops and is going to ten. | Separate brand and entity, written non-encroachment addendum, rehearsed three-part answer. Measured at Gate 2 — if 30 shops 8 miles from nothing we own will not convert 10, the objection is not being answered. |
| Fuel | Dallas diesel moved +52.4% year over year. It is the one input we cannot hedge. | Spec gas, not diesel, under ~30K miles/truck/yr. Reprice monthly and publish changes a week ahead. |
| TBS retaliates on price | They are an efficient buyer with scale. They can cut before we have density. | Do not lead on price. Lead on the Saturday service window and the operator credential — neither is matchable from a price sheet. |
| Signing a lease too early | A standalone build burns cash continuously between 30 and 70 accounts. DFW small-bay vacancy is 4.2% and we are the weak party. | No lease before ~70 accounts. Ride the Golden Glaze footprint; design the centralized kitchen for both if that raise closes. |
| Phone coverage | 191 numbers against 755 targets. This is a phone-and-walk-in channel. | ~$40 Google Places backfill — open ask, awaiting Jack's yes. Beachhead has 22/30 in the meantime. |
| The QBO baseline is not clean | The file mixes Golden Glaze and FlexStay; retail sales $824K against $398K of ingredient cost is 48%, which is not a real donut number. | Confirm which entities sit where before any margin claim is published to an investor. |
Everything open, in the order it should be done
Items 1–4 gate every other line on this page. Nothing below item 8 should consume capital until Gate 0 clears.
Blocking — this week
| # | Action | Owner | Blocks | Status |
|---|---|---|---|---|
| 1 | Call Wayne Clayton, Dawn Foods DFW — 214-878-4577. Ask: "What does your distributor tier look like if I'm buying to resell to fifty-plus independent shops in DFW, not to bake with?" Get the reseller price book in writing plus the volume commitment it requires. DC is in Garland, ~20 minutes from the shops. | Jack | Everything | Open |
| 2 | Call Peter Song, BakeMark — 214-558-3014 (peter.song@bakemark.com, Grand Prairie). Same question, same ask. He already sends the dealer sheet; we need the tier below it. | Jack | Everything | Open |
| 3 | Call Ardent Mills (flour), Ventura / Stratas (shortening) and a sugar refiner rep (Imperial or Domino). These four commodity lines are where a manufacturer-direct tier is most likely to exist — and glaze sugar and roll icing are the two worst lines we buy. | Jack | Gate 0 | Open |
| 4 | Mystery-shop Texas Bakery Supply. Have someone call as an ordinary shop and verbally confirm three shelf prices against our scrape. 920 of 929 SKUs render out-of-stock on their Wix store, so the published book may not be live. | Jack / staff | All pricing | Open |
Worth doing regardless of whether this business launches
| # | Action | Owner | Value | Status |
|---|---|---|---|---|
| 5 | Fix Golden Glaze's own buy on the three worst lines — glaze/icing sugar (49.2% over TBS shelf), white roll icing (48.8% over), powdered 12X (37.4% over). Against a measured ~$310K/yr distributor run rate this is roughly $31K/yr straight to Golden Glaze EBITDA. | Jack | ~$31K/yr | Open |
| 6 | Confirm TBSD LLC is Texas Bakery Supply. If it is, Golden Glaze is paying the incumbent ~$117K/yr — that is leverage, first-hand service intel, and an awkward fact to know before we pitch against them. | Jack | Negotiating position | Likely yes, unconfirmed |
| 7 | Resolve the QBO entity question. The file mixes Golden Glaze and FlexStay: $824K retail sales against $398K ingredient cost = 48%, which is not a real donut number. Either some shops sit in another entity or the ingredient account carries material for entities whose revenue is not in this file. | Jack | Every margin claim | Open — needs Jack's answer |
| 8 | Understand JCT Donuts LLC — $3,550 on 9 February 2026, an outside donut shop already on the books. Nobody sold them. How they arrived tells us something about inbound demand. | Jack | Sales motion design | Open |
Decisions only Jack can make
| # | Decision | Why it is open | Status |
|---|---|---|---|
| 9 | The name. "Golden Star Supply" on the repo and this site; the Discord channel says "Golden Bakery Supply." | Same business, two names. Thirty seconds to swap once decided — but it should be decided before anything is printed or filed. | Awaiting Jack |
| 10 | Google Places API key, ~$40 one-time. | Phone coverage goes from 191 to roughly 755. This is a phone-and-walk-in channel, so that number is close to the whole asset. No billed API gets enabled on Jack's GCP project without an explicit yes. | Asked four times, no answer |
| 11 | Domain. This site is on pages.dev with no domain attached, per instruction. | Fine while it is internal and investor-facing. If any of it goes to prospects, buy goldenstarsupply.com — nothing customer-facing should carry a goldenglazedonuts.com URL. | Deferred |
| 12 | Whether Golden Glaze's credit history backs the truck paper. | Worth $380–$1,030 per truck per month versus a newco — $68K to $185K across three trucks over five years. Raise it explicitly with the lender. | Open |
Queued behind Gate 0 — do not start early
| # | Action | Phase |
|---|---|---|
| 13 | Register the entity, obtain the Texas sales-tax permit, file Form 01-339 with every supplier, register with FDA, apply for both DSHS licenses. | Phase 1 |
| 14 | Buy the top ten SKUs at the distributor tier and run Golden Glaze's seven shops as customer number one. Measure landed cost, shrink, turn rate and receiving labour — do not estimate them. | Phase 1 |
| 15 | Build the order pipe on Golden Ops: shop texts → structured order on our side → auto-confirm with line-item total → pay by link. Test it internally where a bug costs nothing. | Phase 1 |
| 16 | Draft the account packet: credit application with personal guarantee, non-encroachment addendum, Form 01-339, standing-order template. | Phase 1 |
| 17 | Get three broker quotes each for commercial auto and general/product liability. The published "$525/yr food business" quotes price a farmers-market booth, not a fleet. | Phase 1–2 |
| 18 | Work the named 30-shop Camp Bowie list on the ninety-day cadence in Section 09. Jack runs the first ten accounts personally — no rep. | Phase 2 |
| 19 | Buy one used 16-ft non-CDL box truck with a liftgate. Gas, not diesel. Under 26,001 lb GVWR, deliberately. | Phase 2 |
Where every number came from
| Input | Source | Confidence |
|---|---|---|
| Golden Glaze ingredient cost — $29,193 Dawn, $19,543 TBSD, $2,942 BakeMark, $559 Restaurant Depot (May–Jun 2026); ~$310K/yr distributor run rate; $397,812 trailing-twelve ingredient account | QuickBooks Online general ledger, pulled 1 Aug 2026 | Measured |
| "Golden Glaze pays" per-SKU cost | BakeMark price sheet dated 2 Jul 2026, 87 priced SKUs, rep Peter Song — from the Golden Glaze inbox | Measured |
| "TBS shelf" per-SKU price | tbshtx.com product pages, 952 priced SKUs scraped 1 Aug 2026 | Published list, not a live quote — 920/929 render OutOfStock |
| Intercompany supply management $143,822.21 (Jan–Jul 2026); JCT Donuts LLC $3,550 (9 Feb 2026) | Golden Capital Management ledger | Measured |
| 1,669 DFW leads; 755 true independents; 191 with phones; Camp Bowie 30-shop cluster | Google Maps + OpenStreetMap scrape, geo-verified by lat/lon, chain-stripped and deduped. Sheet 1Be3XXojq9rlVq0M9qFrAstQ5uGhxLKi… | Measured |
| Truck pricing, finance tiers, fuel, R&M, insurance bands | Live 2026 dealer listings, ATRI South-Central operational cost data, EIA/AAA Texas fuel prices as of 27 Jul 2026 | Sourced |
| Texas wage rates, burden, workers' comp class rates 7380 / 8018 | BLS Texas occupational wage data 2026, Texas Department of Insurance class rates, TWC unemployment wage base | Sourced |
| DFW industrial rents, 4.2% small-bay vacancy, 17.9M SF H1 net absorption | JLL and CBRE DFW industrial market reports, H1 2026 | Sourced |
| DSHS license fees, FDA registration, 21 CFR 117 modified requirements, Form 01-339 retention | Texas DSHS fee schedule, FDA FFR guidance, Texas Comptroller rules | Sourced |
| Sysco 18.4% GM / 3.8% operating; US Foods 17.4% / 3.0% | Audited FY2025 Form 10-K filings | Sourced |
| Cash conversion cycle, scale curve, EBITDA by account count, working capital by revenue | Modeled from the sourced inputs above | Modeled |
| Lease deposit range, general/product liability premium, per-shop weekly spend of $600 | Judgment | Assumption — not sourced |