Many businesses in Egypt and the Gulf sell almost entirely through chat: a post on Instagram or Facebook, a "DM for price", and the rest on WhatsApp. It works. Customers like talking to a person, the owner knows every order, and there is no website to build. Then one day the owner realises they have not had an evening off in months, messages from Thursday night are still unanswered on Saturday, and sales have stopped growing even though the page keeps growing.
Chat selling is not a phase to be ashamed of, and a website is not automatically the next step. The real question is whether chat, as it is run today, can carry the next stage of growth. This article gives the signs that it cannot, the stages most chat businesses pass through, and a worked example of what to change first.
In short: Selling through chat is enough while every message gets a quick answer, orders are tracked somewhere other than memory, and the business does not depend on one person or one account. It becomes a ceiling when growth means more messages than people can answer well, when customers who want to buy without talking leave, and when nobody can tell what is working. The fix is usually not a big website first. It is organising the chat, then adding self-serve steps only where the signs say so.
Why chat works, and where it stops working
Byron Sharp's How Brands Grow (2010) argues that brands grow by being easy to think of and easy to buy for as many category buyers as possible; he calls the second part physical availability. Chat helps with one kind of buyer and hurts with another. For a buyer who needs reassurance (a gift, a custom size, a first order from an unknown brand), a real person answering questions removes friction. For a buyer who already knows what they want, having to message, wait, ask the price and then wait again adds friction. As a business grows, the share of that second kind of buyer usually grows too.
Speed matters too. A widely cited study in the Harvard Business Review (Oldroyd, McElheran and Elkington, "The Short Life of Online Sales Leads", 2011) found that companies that responded to an online enquiry within an hour were far more likely to reach a real conversation with the buyer than those that responded later. Chat promises an instant answer. When the answer takes hours, chat stops being an advantage.
Four stages of a chat business
| Stage | What it looks like | Ceiling sign |
|---|---|---|
| 1. The founder's phone | One person answers everything, orders live in chats and memory | Replies slow at peaks; orders get lost |
| 2. Organised chat | Business account, catalogue, quick replies, labels, an order sheet, set hours | One person is still the bottleneck |
| 3. Chat plus self-serve | Shared inbox, a simple order page or payment link for common items | Repeat buyers still have to message to reorder |
| 4. Store plus chat for service | Most orders placed without talking; chat for questions, custom orders and after-sale | A different set of problems: traffic and conversion |
Most businesses that feel stuck are at stage 1 and think the only way out is stage 4. Stage 2 usually costs very little and removes much of the pain. Which build to choose when stage 4 is due is a separate question, covered in website, store or app.
The signs table
| What we see | What it means | First move |
|---|---|---|
| Reply time grows in the evening and at weekends | Demand is outgrowing the people answering | Cover peak hours; quick replies |
| The same few questions in most chats | Information is missing where buyers look first | Price, sizes, delivery in the catalogue and highlights |
| Orders tracked in screenshots and memory | Lost orders, no customer list | One shared order sheet |
| "Do you have a website?" or "Can I pay by card?" | Some buyers want to buy without talking | A payment link or simple order page |
| Nobody knows which post or ad brought sales | Spending blind | Ask and tag the source in every order |
| Everything is on one personal number or account | One lost phone or restricted account stops sales | Business account, backups, a second admin |
Finding your ceiling, step by step
- A chat-load count. For one normal week: messages received, conversations started, orders closed, and hours spent answering. Output: four numbers, and the hours the founder or team spent on chat.
- Fifty tagged conversations. For each: the first question asked, the reply time, where it stopped if it did not end in an order, and whether the buyer was new or returning. Output: the five most common questions and the three most common stopping points.
- A reply-time map. Average first reply by hour of the day and day of the week. Output: a simple grid showing the hours when buyers wait longest, usually exactly when most of them write.
- A dependence check. What happens if the main phone is lost, the account is restricted, or the person who answers is ill for a week? Output: a short list of single points of failure.
- One stage, one move. Place the business on the four-stage table and pick the single move that lifts the tightest limit. Output: the move, and the number that should improve in the next month, for example orders per hundred conversations or average first-reply time.
A worked example
This example is hypothetical, and its numbers are illustrative. A handmade leather bags brand in Cairo sells through Instagram and WhatsApp. The founder answers every message herself, often until after midnight. She is considering an online store and an app, because "the page is growing and sales are not".
The count shows around 400 messages and 45 orders a week, with about 25 hours spent answering. The tagged conversations show that most first questions are about price, available colours and delivery fees, and that about a third of conversations stop right after the delivery fee is mentioned. The reply-time map shows that messages sent between nine in the evening and midnight, the busiest window, wait the longest. The dependence check finds that everything runs on her personal number.
The stage is clearly 1. The move is not an app. First, stage 2: a business account with a catalogue showing prices and colours, quick replies for the common questions, and the delivery fee shown in the catalogue and in a pinned highlight so it is not a surprise at the end. A second person covers the evening window from a shared inbox, working from an order sheet. Only then, a simple order page with a payment link for the five best-selling models, aimed at returning buyers who already know what they want.
The number to watch over the next month is orders per hundred conversations, alongside the founder's hours on chat. If those move, the store can be planned later, on evidence, for the buyers who clearly want it.
Five mistakes we see most
- Jumping straight to a big store or an app. It happens because the pain is real and a website feels like the grown-up answer. But a store with no traffic plan and the same unclear prices just moves the problem. Organise the chat first.
- "Price in DM". Hiding the price is meant to start a conversation. For many buyers it ends one before it starts. Where the price is fixed, show it.
- A bot that blocks instead of helps. An automated menu that never reaches a person frustrates the buyers who chose chat to talk. Automate the repeated answers; keep a human one step away.
- No customer list. Every past buyer is a chat buried somewhere. Without a simple list of who bought what and when, there is no way to invite them back; see new or returning customers.
- Measuring followers instead of orders. A growing page with flat orders is a signal, not a success. Track orders per conversation and repeat orders.
The strongest objection
"Our customers want to talk. They will not buy from a website." Many of them, yes, and chat should stay for them. But not every question needs a human: the price, the sizes and the delivery fee can be answered before anyone asks. Removing those from the chat gives the person answering more time for the conversations that need them, and lets the buyers who do not want to talk buy anyway. It is not chat or a store; it is deciding which conversations deserve a person.
Honest limits
For some businesses, chat is the right model for good: high-value custom work, consultative B2B sales, small local services where every order is different. In those cases the goal is to run chat well, not to leave it. And sometimes the ceiling is not the channel at all: if production cannot make more bags, or delivery cannot cover more areas, more orders will only mean more apologies. We say so when that is the real limit.
How OG helps a chat business grow
We start with marketing strategy and planning: the counts and tagged conversations above, the stage the business is in, and the order of moves that will lift the ceiling. Organising the chat itself sits with customer communications and WhatsApp marketing: catalogues, quick replies, shared inboxes, and messages that bring past buyers back. When the signs call for self-serve, we build only what is needed, often a simple order page first and a full online store later, with tracking so every order shows where it came from. Scope follows the need; there are no fixed packages.
Built on: our work with chat-led businesses across Egypt and the Gulf since 2013; Byron Sharp, How Brands Grow (2010); J. Oldroyd, K. McElheran and D. Elkington, “The Short Life of Online Sales Leads”, Harvard Business Review (2011). The four stages are our working method, not a fixed rule. The leather bags brand is a hypothetical example.
Before closing this page
- How many hours a week go into answering chat, and by whom?
- What are the five questions asked most, and are the answers visible before anyone asks?
- At which hours do buyers wait longest for a first reply?
- What would happen to sales if the main phone or account stopped working tomorrow?
- Which of the four stages is the business at, and what is the one next move?
