Think Build Implement Repeat
London, UK +44 7367 067226
WhatsApp FOLLOW f in X
  1. Home
  2. Blog
  3. Why Can Nobody Tell Me How Much We Spend With Each Supplier or on Each Category?
Problems We Solve

Why Can Nobody Tell Me How Much We Spend With Each Supplier or on Each Category?

Supplier names recorded ten ways and spend coded to sundries hide the real picture. We use machine learning to clean and classify spend so you can see it.

Updated 3 min readBy SpiderHunts Technologies

Free estimateNo obligation

Get a free estimate

Tell us what you need. A senior engineer reads every enquiry.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →

Quick answer — TL;DR

Spend visibility breaks down because the same supplier is recorded under many names and lines are coded inconsistently or to catch-all accounts. Machine learning can match supplier name variants to one supplier and classify each line into a proper spend category from its description, with uncertain items sent to a person, giving a spend view you can negotiate from.

A simple question with no answer

Before a supplier review you ask a simple question: how much did we spend with them last year? Finance runs a report and gets a number. Then someone notices the same supplier appears as the full company name, an abbreviation, a trading name, a version with "Ltd" and without, and a separate account set up by a different site. The real number is somewhere across all of them.

The category view is worse. A large share of spend sits in "sundries", "general purchases" or whatever nominal code the person entering the invoice picked. You want to know what you spend on packaging, or IT, or cleaning supplies, across the business, and the answer is a week of someone's time in a spreadsheet.

Why spend data gets this messy

Supplier records are created by many people over many years, often under time pressure. Different sites, different systems after an acquisition, a purchase card here and an expense claim there. Nobody deduplicates, because each record works on its own.

Coding is similar. Nominal codes are chosen by whoever processes the invoice, and the chart of accounts is built for the accounts, not for buying decisions. "Office costs" might hold printer toner, a coffee machine and a software subscription. The information is in the invoice line descriptions, but nobody reads thousands of lines.

What poor spend visibility costs

Blind spotWhat follows
Split supplier recordsYou underestimate your spend and negotiate from weakness
Spend in catch-all codesCategories look smaller than they are and get no attention
Same item from many suppliersMissed chances to consolidate and agree better terms
Off-contract purchasesStaff buy from whoever is convenient
Manual analysisEvery spend review takes days and is out of date on arrival

Procurement decisions end up based on the few big suppliers everyone knows, while the long tail, where much of the waste usually sits, stays invisible.

How we clean and classify your spend

  1. We extract spend data from your accounting and purchasing systems, such as Xero, QuickBooks, Sage, NetSuite or your ERP, along with purchase card and expense data where relevant.
  2. We match supplier name variants to single suppliers using fuzzy matching on names, addresses, VAT numbers and bank details, with a person confirming the uncertain matches.
  3. We agree a spend category structure with you, either your own or a standard one such as UNSPSC, at the level of detail you will actually use.
  4. We train a classification model on lines you have already categorised correctly, then use it to classify the rest from invoice line descriptions, supplier and amount.
  5. Lines the model is unsure about go to a short review queue, and every correction is used to improve the model.
  6. We deliver a spend dashboard by supplier, category, site and period, and run new transactions through the same process automatically so the view stays current.

We can also write the clean supplier and category back into your systems if you want the fix at source, but many businesses start with a separate spend view and tidy the source records gradually.

What procurement and finance get

One figure per supplier, one figure per category, and the ability to drill down to the invoices behind it. You can see where the same thing is bought from several suppliers, where spend has grown, and which categories deserve a proper tender.

Spend reviews become a look at a dashboard rather than a week in spreadsheets, and they reflect last month rather than last year.

It also changes supplier meetings. Walking in with an accurate total across every name the supplier trades under, and a view of what else you buy that they could supply, puts the conversation on a different footing.

Is this your situation?

  • Nobody can quickly say how much you spend with a given supplier.
  • Suppliers appear under several names or accounts.
  • A large share of spend is coded to sundries or general categories.
  • Spend analysis happens rarely because it takes so long.
  • You suspect you buy the same things from too many suppliers.

FAQ

Frequently asked questions

The questions readers ask us after this guide.

Still have a question?

Ask us directly — a senior engineer will get back to you.

Ask about your project

How does machine learning spend classification work?

The model learns from lines already categorised, looking at the description, supplier and amount, and predicts categories for the rest. Uncertain predictions go to a person.

Do we need to recode our accounts?

Not necessarily. The spend view can sit alongside your accounts. Changing coding at source is optional and can come later.

What if our line descriptions are poor?

Classification then leans more on the supplier and amount, and more lines go to review. Improving what is captured at invoice entry helps over time.

What drives the cost?

The number of systems, the volume of transactions, and how many categories you want to track.

What do you need from us?

Access to or exports from your accounting and purchasing systems, and any category structure you already use.

Keep reading

More on Problems We Solve

Start here

Tell us which decision is running on instinct

Describe what you are trying to predict, what data you hold and who acts on it. We will tell you honestly whether machine learning fits, and if a simpler fix would do the job, we will say so.

  1. You tell us what you needTwo minutes on the form, or a message on WhatsApp.
  2. A senior engineer reviews itAnd comes back with questions, a realistic range and an honest view on fit.
  3. Free 30-minute scoping callWe talk through scope, options and a realistic estimate — with no obligation.
Free estimateNo obligation

Talk to someone who builds this

Send a short brief and we will come back with an honest view and a realistic range.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →