Lead Qualification Chatbot: How to Build One Without Code

A lead qualification chatbot is a short, scripted conversation on your website that asks visitors a few questions (what they need, how big the job is, when they need it, how to reach them) and then decides what happens next: route them to a salesperson now, book a follow-up, or point them to self-serve resources. You can build one without code by planning three to five questions, saving each answer, and adding simple IF/ELSE rules that branch on those answers.

This guide walks through the whole process: deciding what "qualified" means for your business, writing the questions, setting up the branching logic, and handing off to a person at the right moment.

Why use a chatbot to qualify leads instead of a form?

A contact form collects information and then waits. A chatbot collects the same information one question at a time and can act on the answers while the visitor is still on the page.

That matters because speed changes outcomes. In a widely cited Harvard Business Review study of 2,241 US companies, firms that tried to contact a web lead within an hour were nearly seven times as likely to qualify that lead as those that waited even one hour longer. A qualifying chatbot gets the conversation started at second zero, then brings in a human when the answers say it's worth it.

Other practical advantages:

• One question at a time feels lighter than a long form with every field visible at once.
• Branching means visitors only see questions that apply to them.
• Context travels with the lead. When a person takes over, they see every answer instead of starting from "How can I help?"
• Unqualified visitors still get help. Instead of a dead end, they get a useful link or an answer.

Step 1: Define what "qualified" means for you

Before you write a single question, write down what a good lead looks like. Keep it to signals a visitor can actually answer in a chat.

You don't need all five. Most small businesses do well with a need question, one sizing question (team size, budget or project size) and an email. Classic sales frameworks like budget, authority, need and timeline are a useful checklist, but a chatbot is not a sales call. Ask only what changes what you'll do next.

A useful test: for each question, write down what you would do differently depending on the answer. If the answer wouldn't change anything, cut the question.

Step 2: Write questions people will actually answer

The wording and answer format decide whether visitors finish the conversation.

Use quick-reply buttons whenever the answer fits a list. Buttons are faster than typing and give you clean data to branch on. "1–5 / 6–20 / 21–100 / 100+" is easier for a visitor and easier for your rules than a free-text box.

Use free text for the one question where nuance matters, usually "Tell us a bit about your project."

Ask for an email with a proper email question, so the format is checked, and say why you're asking: "So we can send you the quote, what's your email?"

Use number questions for numeric thresholds you want to compare, such as team size or number of locations.

A sample question bank

Pick and adapt. Don't use them all.

• What brings you here today? (Pricing / Book a demo / Support / Just browsing)
• Which service are you interested in? (one button per service)
• How many people would use it? (number)
• What's your rough budget? (ranges as buttons)
• When do you need this in place? (This month / Next 1–3 months / Just researching)
• Are you choosing for yourself or for a team? (Myself / My team / A client)
• What's the best email to reach you? (email)
• Anything else we should know? (free text)

Collect only what you need

Every extra field is a reason to quit, and it's also data you're responsible for. Under the EU's GDPR, for example, personal data must be "adequate, relevant and limited to what is necessary" for the purpose (Article 5(1)(c), GDPR). Even if that law doesn't apply to you, it's a good design rule: ask for the minimum, and tell people what you'll do with it.

Step 3: Map the branches

Now turn your questions into a flow. Sketch it on paper first. A typical lead qualification flow looks like this:

1. Greeting that says what the chat can do: "Hi! I can get you pricing, set up a demo, or connect you with the team."
2. Intent question with buttons: Pricing, Demo, Support, Just browsing.
3. Branch on intent:
   • Support goes straight to a person (don't qualify customers who need help).
   • Just browsing gets a short message with useful links and an option to ask a question.
   • Pricing and Demo continue to qualification.
4. Sizing question, such as team size as a number.
5. Condition: IF team size is greater than your threshold, treat as a priority lead; ELSE continue on the standard path.
6. Email question, explained.
7. Handoff or end: priority leads go to a person now; standard leads get a confirmation and a promise of follow-up by email.

If you want a deeper walkthrough of planning branches and conditions, see our guide to building a chatbot decision tree.

Personalize with saved answers

When each answer is saved to a named variable (for example name, team_size, email), later messages can use it: "Thanks, {{name}}. For a team of {{team_size}}, the plan most people start with is..." It's a small touch that makes a scripted conversation feel attentive.

Step 4: Decide when a human takes over

Qualification is only useful if the right leads reach a person quickly.

• Hand off immediately when answers match your best-fit profile and someone is online.
• Always offer a way out. Some visitors will never answer a bot's questions. A visible "Talk to a person" option keeps them from leaving. Our chatbot to human handoff checklist covers this in depth.
• Pass the context. The person who picks up should see every answer so the visitor never repeats themselves.
• Plan for nights and weekends. If nobody is online, collect an email, confirm when the team will reply, and make sure the conversation lands in a shared inbox, not one person's mailbox.

Step 5: Write the messages around the questions

The words between questions do a lot of work.

• Open with a clear promise, not a generic "How can I help?" See our chatbot welcome message examples for templates.
• Say why you're asking before sensitive questions like budget.
• Keep messages short. One idea per message.
• Be honest that it's a bot. Say so in the greeting and make it obvious how to reach a person.
• Confirm the next step at the end: who will reply, roughly when, and where (chat or email).

Step 6: Test, then review real conversations

Before launch, go through every branch yourself, including the odd answers: someone who picks "Just browsing" then asks for pricing, someone who types a fake email, someone who clicks "Talk to a person" on the first message.

After launch, read real transcripts every week for the first month. Look for:

• Questions where many people drop off (shorten or remove them)
• Free-text answers that keep repeating (turn them into buttons)
• Leads that were routed wrongly (adjust thresholds)
• Visitors who asked for a human early (your greeting may be promising the wrong thing)

Common mistakes to avoid

• Asking too much up front. Get the intent first; ask for contact details once the visitor sees value.
• Gating support behind sales questions. Existing customers who need help should never be asked about budget.
• No exit to a person. A bot that traps people costs you the leads you most want.
• Qualifying with no follow-up plan. If no one acts on the answers within the day, the bot only made the delay feel organized.

How to do this in RetroChat

RetroChat's no-code flow builder is built for exactly this pattern. You add steps (message, question, condition, hand off to a person, end), and every answer is saved to a variable you name, such as team_size or email. Question steps support quick-reply buttons, free text, email and number answers. Condition steps branch with IF/ELSE rules using equals, contains, greater than, less than and is set.

You can start from the ready-made Lead capture template, which collects name, email and interest and qualifies quote requests by budget, then edit it to match your questions. The "Talk to a person" button stays visible in the widget throughout, agents see everything the bot collected, and when nobody is online the widget asks for the visitor's email and emails your team. On the Team plan and above you can export contacts and their answers as a CSV.

Ready to qualify leads while they're still on your site?

Sketch your three to five questions, decide what each answer changes, and build the flow. You can try it with the Lead capture template in a free 14-day RetroChat trial, no credit card required.

Frequently asked questions

How many questions should a lead qualification chatbot ask?

Most small businesses do best with three to five, including the email. Start with the intent question, add one or two that change how you'll respond, and ask for contact details last. If a question doesn't change your next step, remove it.

Is a lead qualification chatbot better than a contact form?

It depends on your visitors, but a chatbot has two practical advantages: it asks one question at a time and can act on answers immediately, including handing hot leads to a person while they're still on the page. Many businesses keep a simple form as well for people who prefer it.

Do I need AI to qualify leads with a chatbot?

No. Qualification is usually a set of known questions with known routing, which a rule-based flow handles reliably and predictably. AI is more useful for answering open-ended questions from your own content. Many teams use rules for qualification and add AI later for questions.

What should happen to leads that don't qualify?

Don't send them to a dead end. Give them a useful resource, answer their question, or offer to keep in touch by email. Today's small inquiry can become tomorrow's customer, and a good experience costs you nothing.

Where do the answers from a lead chatbot go?

That depends on the tool. In RetroChat, answers are saved on the contact's profile in the shared inbox alongside their past conversations, and on Team plans and above you can export contacts as a CSV to import into a spreadsheet or other system.