Sign in
Set up for Distributors

AI for insights and analytics

AI analytics for distributors

A distributor asks the same questions every Monday: who has gone quiet, which accounts cost more than they earn, where the price is leaking. AI analytics answers them in your own words, from the invoice ledger and the delivery file, with every figure computed by a tool and every line linking to its rows.

Ask it in your own words

These are the questions distributors ask. Each one maps to a measure our tools compute from your files. The AI model chooses the measure and explains the result; the arithmetic is done by our code, and every total is checked.

The questionThe measure behind itWhat comes back
Which customers have stopped ordering, and how much were they worth?Dormant accounts, by prior valueA list by prior-year revenue, each account against its own order gap, with the rep who owns it.
Which accounts lose money after delivery and order handling?Contribution per customerGross margin less rebates and cost to serve, per account, with the drops and keyed lines that drive it.
Where are we giving price away?Price realisationInvoiced against list or agreed price by customer and line, with the unapproved discounts totalled.
Which customers buy a category elsewhere that they could buy from us?Category share against similar customersEach account's category spend against the median for its segment, ranked by the gap in dollars.
How much did stock-outs cost us last month?Fill rate and lost linesLines cancelled for no stock, valued, by branch and account, against the fill rate.
Are we speaking to the accounts that matter?Value coverage at cadenceRevenue of accounts touched within their tier cadence over revenue assigned, by rep.

A conversation, with the figures

Every figure below was computed by a tool from the rows and checked before it was shown. The lines sum; the percentages match; each line opens to the rows that make it.

Where is revenue going quiet?

Dormant prior-year revenue is $4.2m across 3 branches, measured against each account's own order gap rather than a fixed ninety days. North branch holds $1.9m, 45% of it.

North branch$1.9m45% of the total
South branch$1.3m31% of the total
Central branch$1.0m24% of the total
The 3 lines sum to$4.2m0 unexplained

Which North accounts are behind that?

Of $1.9m dormant at North branch, 3 accounts hold $1.4m, 72% of it. Each is past three times its usual gap and none has a logged call in the period.

Harbour Foods$620kusual gap 9 days, last order 41 days ago
Meridian Catering$410kusual gap 14 days, last order 58 days ago
Ashby Stores$330kusual gap 30 days, last order 104 days ago
These 3 are$1.4m72% of North branch

What should happen this week?

The list is worked, not glanced at. Each account gets an owner, a call and a reason code, and the same question next Monday shows what came back.

North branch managerCall Harbour Foods and Meridian Catering; log the reasonThis week
Sales directorReview the dormant list by value with each branchMonday
PurchasingCheck the shorted lines on Harbour Foods' last three ordersBefore the call
Which customers have stopped ordering, and how much were they worth?Which accounts lose money after delivery and order handling?Where are we giving price away?Which customers buy a category elsewhere that they could buy from us?

The measures behind the answers

Each measure has one formula, one source and one meaning. They are computed per rep and branch and in total, and every one carries an identity that must hold before it is shown.

MeasureFormulaFromWhat it tells you
Value coverage at cadenceRevenue of accounts touched within their tier cadence ÷ revenue of assigned accountsCRM activity; assignment file; ledgerWhether the accounts that matter are being spoken to
Dormant accounts, by prior valuePrior-year revenue of accounts past k × their own typical order gap ÷ prior-year revenueInvoice ledgerRevenue that has stopped without anyone deciding it
Category share against similar customersAccount spend in category ÷ account total, against the median for its segmentLedger with category; account master with segmentWhere a customer buys elsewhere what it could buy from you
Contribution per customerGross margin − rebates − cost to serveLedger; delivery, order, visit and returns filesWhich accounts earn money after what they cost
Price realisationInvoiced price ÷ list or agreed price, by customer and lineLedger; price fileDiscounts given without approval; agreed prices not applied
Average drop valueRevenue ÷ delivery drops, per customerLedger; delivery fileAccounts that could take half the deliveries
Order channel mixOrder lines by portal, EDI, phone, rep ÷ all linesOrder fileCost of order handling; accounts to move to the portal
Fill rate and lost linesLines shipped complete ÷ lines ordered; value of lines cancelled for no stockOrder and shipment filesSales turned away by stock-outs, by branch and account
Customer concentrationTop ten share; largest customer share; effective number of customersLedger, rolled up to parentDependence, and whether it is rising
New account activationAccounts opened in period with a second order within 90 days ÷ accounts openedAccount master; ledgerWhether opening accounts produces customers

Each measure is worked through, with the export it comes from and what to drop, in Sales KPIs for wholesale distributors.

What the AI model does, and what our tools do

The AI model chooses the measure

From your question and the measures declared for distributors, the model picks the one that answers it, and the period and comparison the question implies.

Our tools do the arithmetic

Deterministic code reads the rows, computes the measure, and checks the identities below. The same question on the same data gives the same answer, every time.

The AI model explains, and cites

The AI model writes the sentence around the result, naming the rep or account behind it. It states no figure that is not in the result, and every figure links to its rows.

The identities that must hold

CoverageAssigned accounts = covered + uncovered; no account in two books
DormancyCustomers = dormant + active + too few orders to judge
ContributionAllocated cost to serve = sum of the cost pools
Lost linesLines ordered = shipped complete + shipped short + cancelled + open
ConcentrationSum of customer revenue = ledger total for the period

What it reads

The exports distributors already produce. Column names are mapped once and the mapping is reused. A file is the way in; scheduled delivery and connections to your systems come with the plan, and every source is listed here.

  • CRM activity
  • Assignment file
  • Ledger
  • Invoice ledger
  • Account master with segment
  • Delivery, order, visit and returns files
  • Price file
  • Delivery file
  • Order file
  • Order and shipment files
  • Ledger, rolled up to parent

Who owns each answer

An answer is a list with an owner and a cadence, or it is a chart nobody works.

MeasuresOwnerCadence
Coverage; dormancySales director; branch and sales managersWeekly
Lost lines; fill rateOperations and purchasing, with salesWeekly
Price realisation; drop value; channel mixCommercial directorMonthly
Contribution; category share; concentrationCommercial director; financeQuarterly

Questions distributors ask about AI analytics

Does the AI calculate the dormant list itself?

No. A deterministic tool reads the invoice ledger, works out each account's typical order gap and lists those past it, with the revenue at stake. The AI model chooses that tool from your question and puts the result into words. The arithmetic is done by our code and checked, and the AI model never adds, divides or estimates a figure, which is why the numbers can be trusted in a meeting.

What data does a distributor need to start?

The invoice ledger alone answers dormancy, concentration and price realisation. Add the delivery and order files and you get cost to serve and contribution per customer. Add the CRM activity export for coverage. Every file is one you already produce, and the data map matches your column names once.

Can it tell us why an account went quiet?

It tells you what the data shows: the shorted lines before the orders stopped, the price change on the lines it bought, the drop in categories, the missing calls. Whether that is the cause is the rep's judgement after the call, and the reason code they log becomes data the next time you ask.

How is this different from our BI dashboard?

A dashboard shows the figure; this answers the question and hands over the list. Ask in plain words, get the measure that fits, the rows behind it and the owner, and ask the next question. The figures are the same ones your finance team would compute, because they are computed the same way, and every one reconciles to a control total.

See it on your data

Bring a few thousand rows. The data map opens next, every column mapped once, and the first question is answered in minutes. Free, in your browser, no account.