Selling on WhatsApp when you don't have a website

A large share of sellers in South Asia, the Gulf, Nigeria and South-East Asia have no online store at all: they post on Instagram or Facebook, take orders in DMs or on WhatsApp, and collect payment by transfer or on delivery. Almost every guide to WhatsApp automation assumes a Shopify store behind it, which makes them useless to you.

The good news is that most of the automation still applies. What changes is that the flow is the storefront rather than a notification layer on top of one.

What you do not have

Missing	Consequence
Store events	Nothing fires automatically when an order is placed
An order record	No order number to reference, unless you create one
A cart	No abandoned-checkout recovery
A checkout	Payment happens by link or on delivery
A product catalogue in the flow	Products are described in messages, not selected

Every one of those is real. None of them stops the flows that matter most for a DM-first seller, which are about handling volume rather than notifying about orders.

What automation genuinely does for you

1. Answers the same five questions all day. Price, availability, delivery time, delivery cost, how to pay. For most sellers this is the majority of message volume and all of it is answerable automatically.
2. Qualifies before you spend time. Which item, which size, which city — collected before a human joins the conversation.
3. Captures the order details in a structured way, so you are not scrolling a chat to find an address at packing time.
4. Replies instantly at 11pm, which for a one-person business is the difference between a sale and a competitor.
5. Keeps a customer list you can message later, legitimately, because they contacted you.

The order-taking flow

This is the core of it. Rather than notifying about an order, the flow creates one.

The sheet is doing the job a store database would. It is not a compromise so much as the right tool at this scale — and it gives you something most DM sellers lack, which is a searchable record of what was ordered and where it went. Mechanics are in using a Google Sheet as your automation's database.

Everything here is free to run

Worth stating clearly, because price sensitivity is high in exactly the markets where this pattern dominates. Every message in that flow is a reply to a customer who messaged you first, which sits inside the 24-hour customer service window — and Meta's pricing documentation confirms service conversations have been free and unlimited since 1 November 2024.

You only start paying when you initiate a conversation with someone who has not written to you recently. An order-taking flow never does that.

Moving people off Instagram

If your traffic comes from Instagram, the sensible pattern is to handle the first exchange there and move to WhatsApp for the order, because WhatsApp is where the customer already keeps their conversations and where your notifications will actually be seen.

Two routes: a link in your bio that opens a chat with a pre-filled message, and a Click-to-WhatsApp ad if you advertise. The ad route has a real advantage — it opens a 72-hour free entry point window rather than 24, covered in the 72-hour free window.

Instagram and Messenger flows work the same way in the builder, so you can also run a qualifying flow directly in DMs and hand over to WhatsApp once you have the basics.

What you cannot do

• Take payment inside the chat. No in-chat payment step exists — the flow sends a link from your payment provider.
• Show a product catalogue in the flow. There is no catalogue step in the automation builder, so products are described in text and images.
• Recover an abandoned cart. There is no cart to abandon.
• Automatically notify about stock. Nothing watches inventory — a waitlist plus a manual broadcast is the honest pattern.

Messaging customers later

Once you have a customer list, the temptation is to broadcast to it. Two things to get right before you do:

1. Consent. Someone messaging you about an order has not agreed to receive offers. Ask explicitly, and record what they agreed to — opt-in rules.
2. Frequency. A one-person business with a loyal list can burn it quickly. One marketing message a week per customer is a sensible ceiling — how often you can message a customer.

When you outgrow this

The signals are practical rather than technical: the sheet gets slow, two people need to work the same orders, or you start losing track of what was promised to whom. At that point a real store — Shopify, WooCommerce or a local platform — starts paying for itself, and the automation you have built moves across mostly unchanged.

Until then, a flow plus a spreadsheet is a genuinely workable business, and it is how a large number of sellers in these markets already operate.

Answer the five questions before anything else

Before building an order flow, build the thing that removes most of your typing. Almost every DM-first seller answers the same five questions all day:

1. How much is it?
2. Is it available / in stock?
3. How long does delivery take?
4. How much is delivery to my city?
5. How do I pay?

Each is answerable from material you can write once. A knowledge base built from those five answers, plus keyword flows for the highest-volume two, removes more work than the order flow does — and it works at 2am, which for a one-person business is where a real share of sales are won.

Build it from your own messages rather than from imagination, and include the phrasings customers actually use. In these markets that means transliterated text — "kitna", "kab ayega", "available hai" — which no English keyword list will catch. See handling Roman Urdu and Hinglish.

Give the order a number

With no store there is no order number, and that absence causes real friction later: the customer cannot reference their order and neither can you. Fix it in the flow by generating one — a date plus a sequence is enough — writing it to the sheet and to a contact field, and telling the customer.

It takes one extra step and it makes every subsequent conversation shorter, because "your order 0826-14" is a reference you both share.

What it costs, and when that changes

Nothing, while customers are messaging you. Everything in an order-taking flow sits inside the 24-hour window that Meta's sending messages documentation describes, and that window resets each time the customer writes.

You start paying at exactly one point: when you message someone who has not written to you recently. A dispatch notification the next day, a payment reminder, a restock alert — those land on a cold chat and need an approved template.

So the practical rule for a seller at this scale is to finish what you can inside the conversation. If you need to tell them something tomorrow, ask a question today that keeps the thread alive.

Keeping the order book usable

1. One row per order, never one row per customer.
2. Consistent column headers, since the flow matches on the header text and a rename breaks it silently.
3. A status column you update by hand — taken, packed, sent, delivered.
4. The phone number in international format, so the same customer is not split across two rows.
5. A notes column for the landmark or delivery instruction, which is the detail most often lost in a chat.

That sheet then doubles as your order-status lookup: a customer asks where their order is, the flow finds the row and answers. Same data, second use.

Handling volume when you are one person

The failure mode for a DM-first seller is not lost sales; it is being unable to keep up during a good week. Three things help before hiring anyone:

• Answer the repeat questions automatically, so your attention goes to people who are actually ordering.
• Let the flow collect the order details rather than typing them into a chat and copying them out later.
• Say when you will reply outside your hours, which stops the follow-up message that doubles your queue.

Those three together typically halve the number of messages a person has to touch, which is the difference between a busy week and a missed one.

Growing without a platform

You do not need a store to build a repeat business. The sheet is a customer list, and customers who ordered once can be messaged again — with consent — using a broadcast built from it.

That is a legitimate, well-supported path, and it is how a lot of sellers in these markets operate for years before adopting a storefront. The constraint to respect is frequency: a small loyal list is easy to exhaust.