The Real Cost of Running Procurement on WhatsApp

WhatsApp is good at enquiry and negotiation and bad at record-keeping, approval trails and reconciliation. The measurable costs are duplicate orders, disputes with no retrievable evidence, and hours spent scrolling for a number somebody sent three weeks ago. Move the quotation and the purchase order off it first. The conversation can stay.

We run a WhatsApp number ourselves and we are not about to tell anybody to stop using it. What we are going to do is draw the line between a conversation and a record, because almost every expensive problem in this trade comes from treating the second like the first.

What WhatsApp is genuinely good at

It deserves this section before the criticism, because the criticism is often overstated by people selling software.

It is instant, and everybody already has it. There is no onboarding, no login, no training. A dealer with two employees and a buyer with a purchase department can use the same tool with no negotiation about which system to adopt.

It handles ambiguity beautifully. “Something like the one we took last March, but 4mm” is a sentence that works on WhatsApp and works nowhere else. The back-and-forth of figuring out what somebody actually needs is a conversation, and WhatsApp is a very good conversation tool.

It carries photographs, which matters enormously in this trade. A picture of a nameplate, a damaged carton or a terminal that will not accept a lug resolves in one message what three paragraphs cannot.

And it is where your customers already are. Any argument that ignores this is an argument from a slide deck.

The four failure modes

1. Retrieval. The information exists but you cannot find it. A rate quoted six weeks ago, a delivery commitment, a photograph of a damaged consignment — all present, all buried under four thousand messages, and searchable only if you remember the exact word somebody used. The cost is small each time and continuous.

2. Duplication. Two people in the same firm order the same item because the conversation happened in two different threads. This is common enough that most purchase departments have a story about it, and it is a direct cash cost.

3. No approval trail. A purchase order exists because somebody said yes in a chat. There is no record of who approved what against which budget, which is fine until an auditor, a partner or a new finance person asks.

4. Reconciliation. Your accounts team has an invoice and needs to match it to what was agreed. What was agreed is in a chat thread on somebody else’s phone. This is where the time actually goes, and it goes every single month.

The dispute problem

The failure everybody learns the hard way involves a person rather than a phone.

The purchase executive who placed the order leaves. His number goes with him, and so does every rate he negotiated, every commitment a supplier made, and the photograph of the damaged consignment that was going to settle the debit note. Nothing was hidden. Nobody did anything wrong. The record simply lived on a device that walked out of the building, and the firm finds out about three weeks later when a supplier disputes something and there is nothing to point at.

There is a second version of the same problem, quieter and more common. Messages get deleted — sometimes deliberately, sometimes because somebody was clearing space, sometimes because a phone was replaced and the backup was partial. A thread with a hole in it is worse than no thread, because you can see that something was agreed and cannot show what.

And a third, which matters if a dispute ever goes anywhere formal: chat records are admissible in India, but proving them is a procedural exercise with requirements around electronic records, and it is considerably more work than producing a signed purchase order. Nobody thinks about this until they need it, which is precisely the problem.

GST reconciliation against a chat thread

This is the one that costs real money quietly.

Your input tax credit depends on invoices matching what appears in your GSTR-2B, which depends on the invoice being correct, which depends on the place of supply, the HSN and the tax split being right. If the underlying agreement lives in a chat thread, every one of those fields was transcribed by hand at least once, and every transcription is an opportunity for an error that surfaces three months later.

The specific failure we see most: the delivery address discussed on WhatsApp was a site, the invoice was raised against the billing address out of habit, and the place of supply is now wrong. Nobody catches it because nobody is comparing the invoice against the conversation. By the time reconciliation catches it, getting a revised invoice out of a supplier you have no open business with is a project.

A structured document flow does not make you immune to this. What it does is make the source of truth retrievable, so the error is a correction rather than an archaeology exercise.

What to move off first, in order

You do not need to migrate everything, and firms that try usually fail because they fight the habit rather than working with it.

1. The purchase order. It is the commitment. It should exist as a document with a number, a date, line items and a reference to the quotation. This single change removes most of the dispute risk.

2. The quotation. Once quotes exist as documents, comparison becomes possible and the rate you were given six weeks ago is retrievable.

3. The invoice and payment record. Usually these are already in your accounting software. The gap is the link back to the PO and the quotation, and closing it is what ends the monthly reconciliation exercise.

4. Stock and delivery confirmations. Last, because these change often and are genuinely conversational.

Notice what stays: the enquiry, the negotiation, the photographs, the “can you check if you have this” messages. That is most of the daily traffic and it should stay exactly where it is.

The cheapest version of all of this is a numbered PO template and a shared folder. It is not software, it costs nothing, and it removes more risk than anything else on the list.

What we still do on WhatsApp

We take enquiries on WhatsApp and we always will. Our number is on every page of our site and there is a chat route on the marketplace.

The distinction we draw is between a conversation and a record. Somebody asking whether we have a 4 sq mm push-in terminal block in stock is a conversation and belongs there. A purchase order for a panel’s worth of switchgear is a record, and a record that exists only in a chat thread is one lost phone away from not existing at all.

Keep the conversation. Move the document. That is the whole argument, and anybody telling you to abandon WhatsApp entirely has not worked in this trade.

A test you can run this week

Pick an order your firm completed four months ago. Try to produce, in under ten minutes: the rate you were quoted, who approved the purchase, the delivery commitment you were given, and the invoice.

If you can do it, your current system is working and you should ignore most of this article.

If you cannot, you have just measured the problem. And the cost is not the ten minutes — it is that same ten minutes, or an hour, every time a question comes up, plus the one dispute a year you cannot win.

Related reading: Where the days go in electrical procurement · How to write an RFQ that gets comparable quotes back · Input tax credit on electrical purchases