Selling chatbot services to a stranger is hard. Selling them to a business whose website you already manage is one conversation. Most guides skip this and send you cold-DMing restaurant owners; start where you already have trust instead.
This is the process that works for freelancers I talk to, step by step.
Pitch your existing clients first
If you build websites, do SEO, or run social media for clients, you have a list of businesses that already pay you and already believe you know the internet better than they do. They are also the easiest accounts to deliver, because you have their content.
Rank your client list by one question: who gets website traffic but complains about response time, missed inquiries, or repeating themselves to customers? Those complaints are the chatbot pitch writing itself.
Build the demo before the meeting
Nothing else in this post converts as well, and it costs nothing. Take the client's existing website content, build a bot trained on it (a free plan covers this), and walk into the meeting with their chatbot already working.
Do not present slides about AI. Open the widget and tell the owner: "Ask it something a customer would ask." They will type their most annoying repeated question, in Roman Urdu half the time, and watch their own answer come back. That moment closes more deals than any pitch deck.
If they ask how long this took, tell the truth: an hour, because you already manage their content. That is the service.
The pitch is one question
"What happens to the people who visit your website at 1am?"
Silence, usually. Then walk through it: a third or more of their visitors come outside business hours, the website cannot answer them, and they do not come back. The bot answers the routine questions and writes down the names and numbers of buyers. You hand the owner a list every morning.
Sell the morning list. The AI is plumbing.
The objections and the honest answers
"What if it says something wrong?" Reasonable fear. Show strict mode: ask the demo bot something outside its knowledge and let the owner watch it decline and offer to take contact details instead of inventing an answer. Fear handled by demonstration beats fear handled by promises.
"My customers only use WhatsApp." Agree, then show the Continue on WhatsApp button. The bot catches the website visitors and routes the serious ones to the WhatsApp number they already check.
"How much?" Have the number ready and say it without flinching: setup fee plus monthly retainer. I wrote the full pricing breakdown with PKR and USD ranges; anchor it against what a receptionist costs, not against software.
"I need to think about it." That is what the pilot is for.
Close with a two-week pilot
"It is already built. Let it run on your site for two weeks. If the lead list is not worth the price, I take it off and we are done."
The pilot works because the leads table does the selling. After two weeks, send the portal login and let them count the captured numbers themselves. Convert at the retainer price with the pilot data as evidence.
One discipline note: offer the pilot to two or three businesses at a time, not ten. Each pilot needs its knowledge base tuned in the first weeks, and a sloppy pilot sells nothing.
After the first yes
Ask the new client one question a month later: "Do you know another shop owner who misses customer messages?" A referral from a happy local business converts faster than anything you can write. Then take the leads-table screenshot (ask first) and use it as the opener for the next pitch.
Frequently Asked Questions
Try it on your next client project
Free plan — no credit card required.