Wholesale revenue is four numbers multiplied together: how many active accounts you have, how often each one reorders, how much a reorder is worth, and how many accounts you lose each year. Get those four from your own records and you can forecast the channel more accurately than you can forecast your own store, because a shop buying every eleven weeks is far more predictable than a stranger clicking an ad.
That predictability is the underrated part of the channel. A DTC brand lives with revenue that moves when an auction moves. A brand with forty reordering stockists starts each quarter knowing roughly what will land.
The four numbers
Active accounts. Shops that have ordered in the last two reorder cycles. Not everyone who ever bought; an account that bought once eighteen months ago is not a customer, it is a line in a spreadsheet.
Reorder cycle. The average gap between orders for accounts that reorder. Measure it in weeks. For a consumable or a gift item it is often eight to fourteen weeks; for something durable it can be six months.
Reorder value. The average value of a repeat order, which is usually different from the opening order. Opening orders are often larger, because the shop is buying a range; reorders are narrower and focused on what sold.
Attrition. The share of accounts that stop buying over a year. Some shops close, some change direction, some just never come back. Assume a meaningful number and be pleasantly surprised.
A worked example. Forty active accounts, a reorder cycle of eleven weeks, a reorder value of 420 dollars, and a quarter of accounts lost over the year.
Each account places roughly 4.7 reorders a year, at 420, which is about 1,970 dollars. Forty accounts is about 79,000 dollars, less attrition, so call the base 65,000 to 70,000 dollars. Then add the opening orders from new accounts you open during the year: twenty new accounts at an opening order of 340 adds about 7,000, and the ones opened early in the year begin their own reorder cycle on top.
That is a forecast built from things you can measure rather than a target you hope to hit.
Where the numbers come from
If you have been keeping an account record from your first stockist, all four numbers are already in it. If you have not, start now, because a year of this data is worth more than any forecasting tool.
One row per account, with: shop name, buyer, city, first order date, first order value, the mix they bought, every reorder date and value, and a note on what sold. That is it.
From that sheet you can calculate all four inputs in an afternoon and answer the two questions that matter operationally: which accounts are due to reorder in the next month, and which are overdue and need a call.
The overdue list is the single most valuable output. An account at fourteen weeks when your cycle is eleven has either sold through and forgotten, or not sold through at all. Both need an email, and the email is different depending on which, which is why the note about what sold matters.
Forecast from cohorts, not from a total
The mistake in most wholesale forecasts is treating all accounts as one blob. Accounts behave by cohort: the shops you opened in spring are on a different clock from the ones you open in autumn, and the older cohorts are both more reliable and more valuable.
Lay it out by quarter. Accounts opened in the first quarter, how many are still active, what they have spent. Same for each following quarter. Within two years the pattern is obvious: early cohorts settle into a steady annual value per account, later cohorts are still ramping, and you can project a new cohort onto the curve of the old ones.
This also tells you the truth about a bad quarter. If new accounts slowed but existing cohorts held, that is an outreach problem. If existing cohorts dropped, that is a product, price or service problem, and no amount of outreach fixes it.
Cohort thinking is also how you know whether expansion is working. Adding thirty doors that behave like your best cohort is real growth; adding thirty that churn within two orders is churn with extra shipping, which is the risk running through retail expansion.
Planning production against lumpy orders
Direct demand is smooth: hundreds of small orders averaging out. Wholesale demand is lumpy: nothing for two weeks, then three purchase orders in a day. Production has to be planned against both.
Do it in three steps.
First, take the reorder forecast above and spread it across the calendar by expected reorder date, not evenly. If eleven accounts are due in March, March is a big month whatever else happens.
Second, layer your direct forecast on top, which for most brands is seasonal and reasonably stable year over year.
Third, add a buffer for the wholesale orders you cannot predict: new accounts, an unexpectedly large order, a shop that suddenly wants a second location supplied. A buffer of a few weeks of typical wholesale volume in finished stock prevents most missed ship dates.
Then commit stock properly. When a purchase order is confirmed, that inventory belongs to it and should not be visible to your storefront. Selling promised stock online is the most common way a wholesale forecast turns into an apology.
The seasonality you inherit
Retail buys on a calendar that has nothing to do with your store's calendar, and a forecast that ignores it will be wrong twice a year.
Gift, home and food categories do much of their buying for the winter season in late summer and early autumn. If your product sells at the holidays, the orders land in September and October, and a brand that starts its outreach in November has missed the year.
Then there is a quiet period after the holidays, followed by a spring buying window, and for many categories a market or trade show calendar that concentrates decisions into a few weeks.
Build the shape into your model: heavier months where your category buys, lighter months after. And build your outreach around it, because an email to a buyer two weeks before their buying window is worth ten sent in the dead season.
Turning a forecast into a plan
A forecast is only useful if it changes what you do. Three decisions it should drive.
How many new accounts you need. If the base is 68,000 dollars and you want 110,000, the gap is 42,000. At an average annual account value of around 1,900 dollars, and with new accounts only contributing part of a year, that is somewhere near thirty to forty new stockists, which tells you how much outreach the year requires. Whether that is the best use of the effort compared with other growth options is the question in wholesale as a growth channel.
How much stock to hold. The forecast plus the buffer is your production plan, and it is far better than ordering on instinct.
Whether the channel is worth the effort. Multiply forecast units by the contribution per unit you calculated in the profitability comparison, and compare it against what the same effort would produce elsewhere. That single sum settles most arguments about whether wholesale deserves a day a week.
To find the accounts the forecast asks for, paste a product page into WholesalePilot and the preview shows the kinds of retailers and distributors that plausibly stock something like it, which turns a number in a model into a list of names.
Questions brands ask about wholesale forecasting
How long before the forecast is reliable? About a year, because you need a full cycle of reorders and a look at seasonality. In the first six months, treat every projection as a guess.
What if the reorder cycle varies wildly by account? Segment it. Gift shops and grocers will differ, and two averages are far more useful than one blended number that describes nobody.
Should you count opening orders in the base? No. Opening orders are acquisition revenue and they only recur if you keep opening accounts. Keep them in a separate line so the underlying base is visible.
How do you forecast a distributor? Differently, and with less confidence. A distributor orders in large, irregular blocks based on their own stock position, so plan around their stated schedule and treat anything beyond it as upside.
What is a realistic first-year wholesale number? For a small brand doing outreach seriously, something in the tens of thousands, with the second year considerably larger because the base compounds. Brands who expect first-year wholesale to replace direct revenue are usually disappointed and quit right before it starts working.