SMBs

How to run a segmented WhatsApp campaign

A segmented campaign is built in this order: set the goal, get consent, build the list with its data, define the segment, choose the template, preview and send. Afterwards you read the delivery status and handle the replies. At each step we say where Textra does it.

Summary

  • Goal and audience first, tool second.
  • No consent, no marketing campaign.
  • Every piece of data you want to segment or personalise with has to be in your list.
  • Preview the audience before you send.

1. Set the goal and who you write to

Write in one sentence what you want to happen after the message: they renew, they book, they pick up an order. Then decide what the people this message is useful for have in common, such as a plan, an area or the date of their last purchase.

Before writing to someone with a promotion, it is best that the person has agreed to receive it. According to Meta's documentation, the awareness objective is aimed at customers subscribed to the business's WhatsApp messages.

Keep a record of when and how each person agreed, and offer a simple way to stop receiving messages.

Textra stores a consent record for every contact, which you enter by hand or which the person gives by replying, and it holds back marketing for contacts lacking it.

3. Build the contact list

Gather in one table the phone number of each person and the data you will use: name, plan, area, whatever helps you segment. Export your base as a CSV, with one column per piece of data and one row per person.

Textra needs a field with each row's phone number or email; the remaining columns become custom attributes. For the name, the file must carry name or full_name as a column header.

4. Create the attributes

An attribute is a piece of data about the person that you later use to filter or to fill in the message. Decide what each one stores: a text, a quantity or a yes-or-no answer.

Textra supports custom attributes that hold a text, a number or a yes-or-no value.

5. Define the segment

A segment is the group that meets one or more conditions on those attributes. They are combined with "and" (both hold), "or" (one is enough) and "not" (excluded). Start with one condition and add more only if needed: a very narrow segment can end up empty.

Textra lets you build segments with subsegments. It joins the conditions using "and", "or" and "not" and supports the operators =, !=, <, >, <=, >= and "in". To preview the outcome, it reports how many contacts qualify, along with an example of them.

6. Choose the template and its category

According to Meta's documentation, messages to someone outside the customer service window are sent as templates. Pick the category that fits the content; for a promotion, that is marketing. How to choose between them is covered in our guide to categories.

In Textra, each template has a category.

7. Personalise with a field per recipient

Replace the variable parts of the text, such as the name, with a piece of data from each person. That way each recipient reads text written with their data instead of a generic one.

Textra fills each variable with the value of one of that contact's attributes. In a WhatsApp template the variable can be a text, the header picture or each card of a carousel.

8. Preview, then schedule or send

Check the count, look at some of the contacts included and read the message as they will receive it. Only then decide whether it goes out now or later.

Textra sends right now or books the send for a specific future moment (date and time); before it goes out you can edit or cancel the booking.

9. Read the result and handle the replies

Look at how many messages were delivered and to whom they did not arrive, and note what people replied. Replies are the most neglected part of a campaign and the one that pays off most.

Textra reports, recipient by recipient, whether the message was delivered, and builds campaign reports. When someone replies, an AI agent can handle that reply on WhatsApp and transfer the chat to a staff member, who can also take it over.

Illustrative example

One case, step by step

  1. A language school wants to announce that enrolment is open to people who finished the intermediate level in a given city and agreed to receive news.
  2. It exports its base as a CSV with phone, name, level and city, and loads it. Each person's consent stays in Textra's consent record, which is filled in separately.
  3. It defines the segment: level equal to intermediate and city equal to Rosario.
  4. It picks a marketing template that greets people by name, looks at how many there are and who appears in the preview, and schedules the send.
  5. The next day it reviews who received the message and answers the questions.

Frequently asked questions

What data do I need before starting?

The phone number of every contact, their consent and the data you will use to segment and personalise.

How many conditions should I use?

The fewest that describe the group you care about. With many conditions the segment can end up empty.

Can I write to any contact?

It is best that the person agreed to receive them. For marketing meant to raise awareness, Meta speaks of subscribed customers.

How do I know it went well?

By checking, contact by contact, whether the message arrived, by reading the campaign report and by reading the replies.

Sources

Keep reading

Want to see it with your case?

Tell us what you want to solve and we will go through it together in a demo.

Request a demo