Customer Segmentation Strategy for Distributors

A $1.1 million door at 11% with twice-weekly specials and 55-day pay is not your best customer. A $280,000 door at 24% on a dense Thursday route with 18-day pay might be. Rank the book by trailing-12 GP$ after an allocated freight and returns haircut. That ranking will embarrass a few logos and surface doors the hunters ignore.

Add operational pain to the rank: lines per order, return rate, appointment-window misses, and whether they require a custom label. Pain is capacity. A mid-GP$ account that burns 40 minutes of CS time per invoice is a different segment than a quiet EDI buyer at the same GP$.

The Census AWTS inventory and sales figures will not segment your book, but they will keep you from treating a 4-turn commodity lane like a 10-turn perishable lane when you set service promises.

A useful door record is four numbers: GP$ per stop, SKUs per door in 90 days, drop frequency, and days to cash. A c-store at $95 GP$/stop, 7 SKUs, weekly, 22 DSO is a candidate for penetration. The same GP$/stop at 22 SKUs already bought through you is a retention problem if fill slips. The $40 GP$/stop twice weekly is a fee or a pass.

Worked cut. 180 doors. Top 25 produce 61% of GP$. Next 55 produce 29%. Bottom 100 produce 10% and 44% of the stops. That bottom pile includes 18 doors under $40 GP$/stop. Move those 18 to will-call, a higher minimum, or a $25 delivery add. You will lose some. The truck will get faster. The top 80 will notice.

SKU-per-door is the penetration metric that sales can understand. Publish it by segment. Core independents in your book might sit at 9; the ones you would call healthy sit at 16–20. The gap is a list of items, not a motivational speech.

Hunters open. Farmers expand SKUs and defend fill. House accounts order through the desk with a quarterly visit. If one person is paid like a hunter and given a 90-door farm book, they will ignore the farm. Write the seat, then assign the doors.

A practical split for a $12M house: 1 hunter on a 200-door open list, 2 farmers on 50–70 doors each, inside covering the long tail with a $350 minimum. When a hunter's starter PO survives 90 days, it moves to a farmer with a documented SKU plan. If it stays with the hunter, it will freeze at 6 items.

Enterprise and bid accounts need a named owner who can sit through a QBR and a chargeback fight. Do not leave a $900k banner on an inside queue because 'they just EDI.' Those accounts fire vendors in a portal.

A-segment doors get a standing window, a named backup, and first claim on A-item inventory in a shortage. B-segment gets the published cutoff and the route day. C-segment gets will-call or a once-a-week consolidated ship. If everyone gets A-segment promises, nobody does.

Price follows the cut. A-segment can earn a tighter multiplier in exchange for EDI, a quarterly minimum, and 15-day pay. C-segment pays list-plus-freight. Mixing those two on the same matrix is how the C door trains the A door to ask for the same number.

Write the gates. Upgrade to A if trailing-6 GP$ clears $18,000, DSO is under 30, and return rate is under 2%. Exile to C if GP$/stop stays under $45 for a quarter after a coaching visit, or if they go 45 days past terms twice. Put the rule in the ERP so a friendly rep cannot silently keep a dead door on the premium route.

Review the cut every quarter with invoices, not opinions. Doors migrate. A seasonal garden account that looks C in January may be A in April — plan the inventory, do not change the segment on a feeling in February.

The SBA advice on knowing your customer is generic and still right on one point: you cannot treat every buyer as the same work. In wholesale the difference shows up as minutes on the dock, not as a persona workshop.

A farmer swore the book was healthy. Sales were flat-up 4%. The cut showed eight doors under $38 GP$/stop, two of them 62-day payers, one of them a 9% return rate because they ordered cases and returned inners. Those eight consumed Thursday morning. After minimums and a will-call move, four left and four paid the fee. Thursday opened 70 minutes. The remaining 32 doors got a 2-hour QBR slot that had never existed.

That is the job of segmentation: free capacity for the doors that fund the building. If your cut does not change a route or a price, it was a slide.

A door that pays in 14 days and a door that pays in 52 are not the same A-segment, even if GP$ matches. Put credit tier on the cut: COD / Net 10 / Net 30 / watch / hold. A watch-tier 'A' account does not get extra dating to win a bid. They get the product if the card clears.

Concentration is a silent segment. If one banner is 22% of GP$, they are not just A — they are a going-concern risk. Dual-source their category in your own head: what happens to the truck if they move the primary in 45 days. Do not give them a custom warehouse process that you cannot reuse.

New doors start in a probation segment for 90 days: limited terms, starter SKU set, no custom labels. Graduation is earned with two paid cycles and a return rate under 2%. Hunters hate this. Credit managers sleep.

Independents want a person and a standing day. Banners want EDI, appointments, and a portal owner. Bid accounts want a packet and a penalty clause. Marketplace resellers want MAP-safe pricing and will burn your authorization if you get sloppy. Those four channels do not share a price list or a CS queue.

If you sell to a reseller who also lists on a marketplace, MAP compliance is their problem and your risk. Segment them, watch advertised price weekly, and keep the paper. Mixing them into the independent farm book is how a farmer gets blamed for a brand letter they never saw.

Use a simple tag in the ERP — IND / BAN / BID / RES — and make every report able to filter it. A blended GP% across those four will lie to you every month. The Census wholesale kind-of-business cuts are coarser than this; yours should not be.

Two doors with identical GP$ are not the same if one is on the ring and one is 48 minutes off it. Add a density tag: on-ring / adjacent / off-ring. Off-ring A-accounts can stay A on price and still move to a once-a-week consolidated ship. Off-ring C-accounts should not be on the truck at all.

Drop-size brackets belong on the segment, not as a whispered exception. Under $200: will-call or a fee. $200–$400: on-ring only. Over $400: full service if DSO is clean. Publish it. Sales will complain for two weeks and then use it as a reason to raise the order instead of begging for a free stop.

Re-cut after a fuel spike or a driver loss. The economics of the off-ring door change when diesel moves $0.80 or when you lose the person who 'did not mind' the deadhead. Segmentation that cannot survive a cost shock is a poster, not an operating rule. Put the new math on one page and walk the affected doors with a date, not a surprise on the invoice.

Farmers will defend a dog because they remember the year it was good. Bring the last 12 invoices, the GP$/stop, and the minutes on the dock. If the door still belongs on the premium route, the numbers will say so. If they do not, give the farmer a 30-day save: raise the drop, collect the aging, or accept the fee. After 30 days the system moves the door. Do not let a hallway conversation undo the cut.

New hires get the cut on day one, not 'once they know the book.' The book they will invent otherwise is last year's sales ranked by whoever shouts. Hand them the A list, the penetration list, and the will-call list. That is the job.

Publish wins when a cut works: Thursday saved 70 minutes, four doors paid the fee and stayed, contribution on the ring rose $1,100 a week. People copy what you celebrate. If you only celebrate new logos, the cut will die in a quarter.