Skip to content
Academy
Marketing Academy · Field Work●AI in Marketing
CoreBuild the Asset· 45 minutes

Build the Flow: A Qualification Script That Doesn't Feel Like a Form

RateGain Travel Technologies

Objective: Design a complete qualification chat flow, trigger through handoff, for a company's pricing page, applying the two-to-three-question cap and the lesson's explicit escalation triggers.

You're the growth marketer at RateGain Travel Technologies, the travel-tech SaaS company, and you've been asked to script the chatbot that will sit on the new hotel-revenue-management pricing page.

Script the qualification flow first (earn-then-ask, capped questions, branch to the right next step), then define the handoff rules that pull a human in before the bot starts guessing.

Before you start

What you'll need

Free path (everything below is enough to finish)

FreeDraft the turn-by-turn script and the handoff-rules table

No account friction, plain rows are enough to script and review a flow

Paid upgrades (optional, faster/deeper)

HubSpot Marketing Hub(optional)
FreemiumBuild and publish the live qualification chatflow once the script is approved

Native chatflow builder with branching logic and CRM handoff routing

No access? Any chatbot platform with branching logic and a live-agent handoff works; the script is platform-agnostic

The process

2 steps

Step 01 of 02

Earning information before asking, capped at two to three questions

A good qualification flow answers a real question first, then asks one soft qualifier at a time, capped at two to three questions, and requests email only after delivering value.

A visitor lingers 70 seconds on the pricing page and asks 'does this integrate with our existing PMS?' Script the next four bot turns so the flow earns information instead of demanding it, and stays under the question cap.

Google Sheets— A turn-by-turn flow script, one row per bot/visitor exchange, with a 'question count' column.

Procedure

  1. Turn 1: bot answers the PMS-integration question directly, no gate
  2. Turn 2: bot asks ONE soft qualifier tied to the answer just given (question count: 1)
  3. Turn 3: branch on the answer, route toward either a resource or a next question (question count: 2 max)
  4. Turn 4: request contact info only after a specific answer or resource has been delivered
Sample output
RateGain pricing-page flow script (4 turns)

Bot: Yes, we integrate with 40+ PMS platforms including Opera and Protel. Which one are you on?
Visitor: Opera.
Bot: Good, that's a certified integration. Out of curiosity, are you evaluating this for one property or a portfolio? (Q1)
Visitor: A portfolio, about 12 properties.
Bot: For a 12-property portfolio, I'll have our revenue-management specialist send over a group pricing sheet, what's the best email? (Q2, value delivered first)

Healthy

Two questions total, both tied to the answer just given, email requested only after a specific resource was promised.

Unhealthy

Bot opens with 'What's your email so we can help you?' before answering the PMS question at all.

What this means

Every question needs a payoff attached to it; a flow that asks before it gives reads as a form wearing a costume.

So what do I do about it?

SymptomActionEffort
Flow asks for email in the first or second bot turnMove the email ask after at least one specific answer or resource has been delivered30 min
YouYou can do this yourself, no engineering access required.

Step 02 of 02

Escalating to a human on explicit triggers, not on a bot's best guess

Every flow needs explicit escalation triggers: repeated confusion, high-value signals, emotional or complaint language, anything outside the knowledge base, and any explicit request for a human, honored on the first ask.

The same pricing-page flow above needs handoff rules written before launch. List the five trigger conditions and the exact bot response for each, including what happens if the visitor types 'talk to a person' mid-flow.

Google Sheets— A five-row handoff-rules table: trigger, bot action, escalation target.

Procedure

  1. List all five trigger types from the lesson (repeated confusion, high-value signal, emotional language, out-of-scope topic, explicit request)
  2. Write the exact bot response for each, including a disclosure that it's a bot
  3. Confirm the explicit-request rule fires on the FIRST ask, not the third
  4. Route each trigger to a named escalation target (live rep, ticket queue, callback form)
Sample output
RateGain handoff rules (5 rows)

Trigger                        Bot action                                          Escalation target
Repeated confusion (2x)        'Let me connect you with a specialist.'             Live chat rep if available
High-value signal               Route immediately, no further questions            Live rep, real time
Emotional/complaint language    'I'll get a person on this right away.'             Live rep, priority queue
Out-of-scope (contract terms)   'That needs a specialist, here's the fastest way to reach one.' Ticket queue
Explicit 'talk to a person'     Honor on first ask, no further bot questions        Live rep or callback form

Healthy

All five triggers are covered and 'talk to a person' is honored on the first ask, not delayed.

Unhealthy

The bot asks two more clarifying questions after a visitor types 'agent', trying to resolve it first.

What this means

An explicit request overrides every other flow logic instantly; delaying it is one of the fastest ways to lose a visitor's trust.

So what do I do about it?

SymptomActionEffort
Bot has no rule for 'talk to a human' typed mid-flowAdd it as the highest-priority trigger, checked before any other flow logic30 min
YouYou can do this yourself, no engineering access required.

Final deliverable

A four-turn qualification flow script plus a five-row handoff-rules table for the pricing-page chatbot, question count capped at two.

See a reference example
Sample output
MapmyIndia enterprise-page flow script (excerpt)

Bot: Yes, our SDK covers both Android and iOS with the same location APIs. Which platform are you building first?
Visitor: Android.
Bot: Good, that's our most-used SDK. Is this for an in-house app or a client project? (Q1)
Visitor: Client project, a logistics company.
Bot: For a logistics client, I'll have our SDK specialist send over the fleet-tracking integration guide, what's the best email? (Q2, value delivered first)

Success criteria

You're done when you can:

  • Scripts a complete 4-turn qualification flow under the 2-3 question cap
  • Requests contact info only after value delivered
  • Covers all five handoff-rule trigger types
  • Honors an explicit 'talk to a person' request on the first ask