AI-powered flow creation. No drag-and-drop required.

What is WhatsApp automation?

WhatsApp automation is a set of rules that send a message when something happens, without anyone in your team pressing send. Something happens (an order is placed, a cart is abandoned, a form is submitted, a customer replies with a keyword), a condition is checked, and a message goes out.

That is the whole idea. The complexity in most automation tools is not in the concept, it is in the interface, which is why Weflux lets you describe the flow in plain English and builds it for you rather than making you assemble it from boxes and arrows.

The useful distinction is between automation that replaces a conversation and automation that starts one. Order updates and OTPs replace a conversation nobody wanted to have. A cart-recovery message starts one you very much want. The second kind needs more care with timing and tone, and it is where the returns are.

What can trigger a flow

• An inbound message. A customer messages you, or replies with a specific keyword.
• A button tap. Someone taps a quick-reply or call-to-action button on a template.
• A CRM change. A contact moves to a new lead stage, or gains a tag.
• An external event. Your store, CRM or backend calls a Weflux webhook: order placed, payment failed, shipment dispatched.
• A schedule. A fixed delay after a previous step, or a set time before an appointment.

Five flows worth building first

Rather than a feature list, here is what these actually look like in practice.

1. Abandoned cart recovery

Trigger: cart abandoned webhook from your store. Wait: 15 to 60 minutes. Condition: order still not placed. Send: a Marketing template naming the item, with a button back to the cart.

The wait matters more than the copy. Fire it in two minutes and you interrupt someone who is still deciding. Fire it the next day and they have bought elsewhere.

2. Cash-on-delivery confirmation

Trigger: COD order placed. Send: a Utility template asking the customer to confirm with a button. Then: tag confirmed orders, and route the silent ones to a human before dispatch.

For Indian D2C this is usually the highest-return flow on the list, because every unconfirmed COD parcel that ships is a return-to-origin cost you pay both ways.

3. Appointment reminders

Trigger: a booking exists. Send: a Utility reminder ahead of the slot, with the time and, for virtual bookings, the meeting link. Weflux generates a Google Meet link automatically for virtual appointments and includes it in the reminder.

4. Post-purchase sequence

Trigger: order delivered. Send: a thank-you and care instructions, then after a sensible gap a review request, then a replenishment nudge timed to the product's actual life. Three messages spread over weeks, not three messages on Tuesday.

5. Lead qualification

Trigger: new inbound message from an unknown number. Send: two or three quick-reply questions, tag by the answers, set the lead stage, and assign to the right agent in the shared inbox. The point is not to close the lead, it is to make sure a human opens the right conversation first.

Building a flow

1. Describe it. Write what should happen in ordinary language, including the trigger, the wait and the exit condition.
2. Check the templates. Any step that reaches someone outside the 24-hour window needs an approved template. The flow will tell you which steps those are.
3. Set the exit. Decide what stops the flow: a purchase, a reply, a stage change. A flow with no exit condition is how customers end up getting a cart reminder for something they already bought.
4. Test on yourself. Run it against your own number before it touches a list.
5. Turn it on and watch the opt-outs. Opt-out rate is the honest measure of whether the automation is welcome.

What WhatsApp automation cannot do

Worth being direct about, because these are platform rules rather than product gaps, and every provider has them.

• You cannot message someone first with free-form text. Outside the 24-hour window, every automated message must be an approved template. If a flow needs to say something unpredictable, it cannot reach a cold contact.
• Automation cannot manufacture opt-in. A flow can only message people who have consented, and no amount of clever triggering changes that.
• Templates are approved before use, not on the fly. A flow that needs new wording needs a new template and Meta's review first.
• Facebook Messenger and Instagram behave differently. On those channels you cannot start a conversation at all, and there are no templates. Automation there can only respond to someone who messaged you.
• Automation is not a substitute for the inbox. Anything that draws a reply hands a human a live conversation, which is the point.

Connecting your own systems

Flows are more useful when the trigger comes from where the event actually happens. Weflux takes inbound webhooks so your store, CRM or backend can start a flow, sends outbound webhooks so a flow can update your systems, and offers a REST API for anything bespoke. Google Sheets works for the many real cases where a spreadsheet is genuinely the system of record. Details in the API documentation.

Is this a chatbot or an automation?

The words get used interchangeably and it causes real confusion when comparing platforms, so worth separating them.

An automation is proactive. Something happens in your business and a message goes out: an order ships, a cart is abandoned, an appointment is tomorrow. The customer did not ask for it, and outside the 24-hour window it has to be a template.

A chatbot is reactive. A customer messages you and something answers, either with a menu of options or by interpreting what they wrote. It runs inside an open 24-hour window, so it can use ordinary free-form messages.

Most useful setups are both. A broadcast goes out, some people reply, a bot handles the first exchange and qualifies them, and anything it cannot resolve is assigned to a human in the shared inbox with the context already attached. Selling those as separate products is a pricing decision, not a technical one.

Timing is most of the work

The single biggest difference between a flow that earns money and one that generates opt-outs is when it fires, not what it says.

• Cart recovery: long enough that they have actually left, short enough that they have not bought elsewhere. Fifteen to sixty minutes for considered purchases, longer for high-value ones.
• COD confirmation: immediately. The order is fresh in their mind and you want the answer before the parcel is packed.
• Review requests: after they have used the thing, not after it was delivered. For a consumable, days. For something durable, weeks.
• Renewal and replenishment: work backwards from when they actually run out, which your order data already tells you.
• Anything promotional: business hours in the recipient's timezone. A marketing message at 11pm buys a block, not a sale.

A useful discipline: for every flow, write down what the recipient is doing when it arrives. If the answer is "sleeping" or "at work and irritated", change the timing before you change the copy.

Knowing whether a flow is working

Automation makes it easy to send a great deal without anyone reading it, so pick the measure carefully. Delivery rate tells you almost nothing, since it is near total on a clean list. Read rate is high on WhatsApp for everything, including messages people resent.

The pair that tells the truth is reply rate against opt-out rate. A flow generating replies is starting conversations, which is the point. A flow generating opt-outs is spending your list to make a short-term number look good, and a list is not a renewable resource. Watch both after every change, and be willing to switch a flow off.

For revenue flows such as cart recovery and COD confirmation, attribute properly: compare against a holdout that receives nothing, rather than crediting the automation with every sale that happened to follow it. The uplift is usually real and usually smaller than the raw number suggests.

Related

• WhatsApp broadcasts for the campaigns that feed people into these flows
• Shared team inbox for the human handover when a flow draws a reply
• WhatsApp CRM for the tags and stages flows read and write
• API and webhooks for triggering flows from your own systems
• Industry playbooks for the flows that work in your sector