Not every bot is the same product
For an Israeli business in 2026: if customers already write on WhatsApp, start there. Discord is for a community that already lives in a server. Telegram is for a channel and updates. A website is for people who search. An app is only for daily use that does not fit a chat.
The line "build me a bot" shows up in almost every call. Then it turns out one person meant an auto-reply on WhatsApp, one meant commands in a Discord server with thousands of members, and one meant a Telegram bot that posts updates to subscribers. All three are called a bot. They cost differently, break differently, and serve a different person.
I am Naor Cohen. NaorX is the name I build this under: WhatsApp bots, Discord and Telegram bots, websites, apps, and automations that connect them. This post is not here to sell all of them. It is here to stop the wrong build before any code.
WhatsApp: the default in Israel
If the customer already messages the business number, a WhatsApp bot is usually the first step. Not because it is trendy. Because the conversation is already there: opening hours, order status, booking, handoff to a person.
What matters before the excitement: business WhatsApp goes through Meta. There is account approval, there are templates for messages you start, and there are rules about messaging people who did not write first. A bot inside a chat the customer opened is one product. A bot that blasts marketing to every contact is another, and it usually gets blocked.
I wrote separately about why the same word "bot" can cost ten times more: what a WhatsApp bot costs.
Discord: when the community is already in a server
Discord does not replace the WhatsApp number of a local shop. It fits when a community already lives there: a course, a gaming server, a product whose users talk to each other, support that needs roles, logs, and commands.
A useful Discord bot does the work people in the server already do by hand. Assign a role, open a ticket, check a subscription against the site, ping when something breaks. A bot built because a competitor has one, with no command anyone actually runs, is dead in two weeks.
Telegram: a channel and updates, not a front desk
Telegram is a good fit when the audience is already there, or when you need a channel: a price update, a system alert, a closed group, a bot that answers a command without waiting on Meta. The API is stable, and setup is usually shorter than business WhatsApp.
The common mistake in Israel is to put all customer service on Telegram because it looks cheaper, while customers write on WhatsApp and will not switch apps. If you have no audience on Telegram, the bot will be finished and empty.
From the work:
Before you pick a platform, open the phone of a typical customer. Where they already write to you is where the bot goes. Not the platform the developer likes.
When it is a website, not a bot
A bot does not show up in Google instead of the business. Someone searching for a local service lands on a page, not inside a chat. You need a site when there is something to show people who do not know you yet: services, a plain price range, a phone number, an address, a form.
A bot without a page is a number that replies. A page with no way to continue into a conversation is a site people read and close. For a lot of small businesses in Israel the sane mix is a short site plus WhatsApp, not a choice of one.
When it is an app
An app earns its cost when there is repeated use: login, history, notifications, camera, work that should not drown between family messages. If the action is "ask for a status" once a week, an app is a heavy project on top of a message.
A lot of "we need an app" requests end more honestly as a site that works on a phone, or as a bot. That is not cutting corners. It is not asking people to install something for one action.
Short table
| The need | What to build |
|---|---|
| Customers already message the business number | WhatsApp bot |
| Community, roles, support inside a server | Discord bot |
| An update channel, or an audience already on Telegram | Telegram bot or channel |
| People search for you on Google | Website |
| Daily use with an account and history | App |
The part people skip: the connection
A bot that stays inside the chat and never touches orders, a calendar, or a sheet becomes another box someone copies from by hand. The moment there are two tools, there is automation: an order on the site opens a message, a form lands in a CRM, an expired membership removes a Discord role.
That is not a button. It is logs, and an error when the other side is down. If nobody looks at the errors, the connection "works" until the day it does not.
Before you send "build me a bot"
- Where does the customer already write today?
- What is the first action they should finish without a person?
- Who steps in when the bot does not know the answer?
- Is there a website, or is the bot also supposed to be the business card?
- Who maintains it after the first month?
Five answers like that save a quote for the wrong product.
Not sure if it is WhatsApp, Discord, a site, or an app?
Write what the business does today and where customers already reach you, via the form on NaorX. We can cut it down to a first phase that is actually buildable.
Need help with your project?
Whether it's a website, bot, automation or something else - I'm here to help you build a solution that works
