How to automate your WhatsApp customer intake in 5 steps

Most businesses we talk to are handling WhatsApp enquiries the same way they were five years ago: one person reads every message, decides what it is, replies manually, and then copies the details somewhere else. This works when you get five enquiries a day. It stops working at twenty.

Here's the exact five-step intake flow we've built for service businesses that want to stop managing this manually — using Make or n8n, no custom code required.

The problem with manual intake

Manual WhatsApp intake breaks in three specific ways:

• Messages arrive at unpredictable times. Staff can't always respond in under an hour, which is when leads go cold.
• There's no consistent information collected. One customer gives a full brief; another sends a single sentence. Both end up in the same queue.
• Nothing is tracked. There's no record of who enquired, what they wanted, or whether they were followed up.

Automating intake doesn't remove humans from the process — it makes sure every enquiry arrives at a human with the information already collected, qualified, and organised.

Step 1: Capture the trigger

Every automation needs a trigger. For WhatsApp intake, this is the first inbound message from a new contact.

In Make or n8n, you configure a webhook that receives messages from your WhatsApp Business API provider whenever a new message arrives. The webhook fires immediately — no delay, no polling.

What to do: Set up a webhook trigger that captures the sender's phone number, their message content, and the timestamp. Store these in your automation run context — you'll need them in every subsequent step.

One thing most people miss

Check whether the number has messaged before. If it has, you're dealing with a returning customer, not a new enquiry. Route them differently — either to a status update flow or straight to a human.

Step 2: Collect the right fields

Before you can qualify or route an enquiry, you need consistent information. The fastest way to get it: a conversational form, delivered as a WhatsApp message sequence.

Send a structured reply that asks for exactly the fields you need. For most service businesses, this is:

1. Full name and business/organisation
2. The specific service they're enquiring about
3. A brief description of what they need
4. Their preferred timeline

Keep it to four fields maximum. Every additional field increases drop-off. Use numbered options where possible ("Reply 1 for X, 2 for Y") — it's faster for the customer and easier to parse in your automation.

In our experience, businesses that use structured intake see response completion rates of 70–85%, compared to 20–30% for open-ended "tell us more about your project" prompts.

Step 3: Qualify automatically

Once you have the collected fields, run them through a qualification step before routing to a human. The goal is to identify: is this a real lead, or is it noise?

A basic qualification filter checks:

• Is the service requested one you actually offer?
• Is the timeline realistic (not "by tomorrow" for a six-week project)?
• Does the description suggest a budget that fits your minimum?

For more sophisticated qualification, you can run the message content through a simple AI classification step using OpenAI's API. Pass the enquiry text and a prompt that scores it against your ideal customer profile. It costs fractions of a cent per enquiry and takes under a second.

Step 4: Route intelligently

After qualification, you have three possible routing outcomes:

1. High-quality lead: Notify the right team member immediately with a structured summary of the enquiry. Include the collected fields, a qualification score if you're using one, and a one-tap link to reply.
2. Out-of-scope: Reply to the customer with a polite, specific message explaining you don't cover that area, and optionally suggest an alternative. This is automated — no human needed.
3. Incomplete information: Send a follow-up message asking for the specific missing field. Give them one more chance before escalating to a human review queue.

Routing notifications to your team work best via WhatsApp itself (a dedicated ops group), Slack, or email — whichever your team actually checks. Don't route to a dashboard nobody opens.

Step 5: Confirm and close the loop

Send the customer an automated confirmation the moment their enquiry is received and processed. Include:

• Acknowledgement that you received their message
• A clear statement of when they'll hear back (be specific: "by end of day Tuesday" is better than "within 48 hours")
• A reference number or summary of what they submitted

This one step alone eliminates most "did you get my message?" follow-ups. It also sets a firm expectation — which means if you miss it, the customer has something to point to.

Update your dashboard or CRM automatically at this point. The enquiry status should move from "received" to "in review" so your team sees it without having to check the WhatsApp thread.

Tools we use for this

For most service businesses, Make is the right starting point — it has a visual workflow builder, handles WhatsApp API webhooks cleanly, and doesn't require programming knowledge. For higher volume or more complex branching logic, n8n is faster and more flexible.

For the WhatsApp API layer, we use 360dialog for Ghana-based businesses (competitive pricing, good WABA approval rates) and Twilio for UK businesses (better documentation, faster support).

If you want AI classification in step 3, you'll need an OpenAI API key. The cost is negligible — a hundred thousand enquiries would cost under $5 at current pricing.

What to do next

If you want to implement this yourself, start with step 1 and 2 only — just capture inbound messages and send a structured intake form. Run that for a week and see what data you're actually getting before building out the routing and qualification layers.

If you'd rather have someone build it for you, we do exactly this. A basic intake automation usually takes two to three weeks from brief to live, and costs under $1,000 for most service businesses.