Written by Claudia Sin

‍

Imagine you are a small business owner just getting your new business off the ground.

You add your business phone number to business directories and online platforms, hoping to get the attention of your potential customers.

The phone starts buzzing and the enquiries start coming in. You feel excited! You start flexing your best sales pitch — converting these first customers is crucial to the early success of your business. You answer each call one by one, with care and lots of attention.

And yet, no matter how much effort you put into the communication, most of them don’t end up buying anything.

You know your pitch is amazing, and you know your product is even better. So, what’s going wrong?

Why qualifying leads matters

After some time, you start noticing that the people who contact you aren’t always a potential customer. Some of them are window shopping. Some of them are just scoping your prices out so they can compare with others. Some are so top of the funnel that they might end up converting, but not for many weeks or even months. 

But for every enquiry, you and your team have to spend a considerable time replying, nurturing, and trying to make sales.

After a few months, you come to realize that there’s one question that helps to qualify leads quickly. Asking “Are you having X troubles right now, and are you looking for service Y to solve your problem?” cuts straight to the point and saves you time.

You share this insight with your team, who also start asking the question early on in lead-qualifying calls.

As a result, the time spent on customer enquiries has significantly dropped. Your team has become more efficient in determining whether an enquirer is a potential customer.

Job done, right? 

Not quite! Handling incoming enquiries manually is pretty repetitive. Despite your team members becoming efficient, they still spend a lot of time talking to potential customers. There’s more work to be done to further improve this process.

Using the power of automation and AI, you can remove repetitive tasks from your team’s workload, freeing them up to spend time and effort on more important, manual-based tasks.

Let’s look at how that’s done.

How a virtual AI assistant works to qualify your leads

In this post I’m going to show you how to build a virtual AI assistant that can handle qualifying leads and record keeping for you.

The virtual assistant’s workflow will look like this:

Save the enquirer phone number and personal info (name, company name, email) into a Customer Relationship Management (CRM) software↓

Ask a lead-qualifying question↓

If user passes the lead qualifying question, schedule a call with sales representative↓

If user does not pass the lead qualifying question, send a closing message

How to build a virtual AI assistant without code

These are the tools we’ll use to build our virtual AI assistant:

Zapier – Controls the whole interaction of the virtual AI assistant

Twilio – Receives and sends SMS messages

Airtable – Marks down user info and previous messages into the CRM

ChatParse.AI – Extracts user information from messages, and analyzes if they mean “yes” or “no” in messages

Calendly – Used to schedule calls with sales rep

‍

Now we’ll look at each tool individually and walk through how to set them up.

Airtable setup

1. Create your account at https://airtable.com/.
2. From there, create a new base and table which has these fields:

How we’ll use Airtable

We will use Airtable like a database or CRM. All users’ incoming messages, phone numbers and names (if name has been mentioned by the enquirer) will be stored in your Airtable table.

Every time a new phone number is given, a new row will be created in the table. The incoming messages made by that number will be saved separately in the columns titled Message 0*, Message 1, Message 2 and so forth.

*In programming, similar information that is stored in the same container is called an array, and the array starts with item 0, item 1 and so on. We follow the pattern and start our message with an index 0, i.e. Message 0 is the first message received, Message 1 is the second message received.

Twilio setup

‍

‍

1. Create your account at https://www.twilio.com/.
2. Buy a number which has SMS functionality.
3. Configure Messaging Webhook, use TwiMLBin in the dropdown, and create a new template like below. This stops Twilio’s default SMS auto-reply.

How we’ll use Twilio

We will use Twilio like a mobile phone with SMS, except the mobile phone is managed by cloud computing.

When a customer messages this phone number through SMS, Twilio will help to forward the incoming message to your preset destination, which is Zapier in our case.

ChatParse.AI setup

‍

‍

1. Create your account at https://www.chatparse.ai/.
2. Follow the Zapier plugin installation instructions on their website.

How we’ll use ChatParse.AI

We will use ChatParse.AI like a translator — it will, in essence, translate human language into computer language.

When a user sends a message about their intention, ChatParse.AI will interpret its meaning and make it understandable by computer logic.

Calendly setup

‍

‍

1. Create your account at https://calendly.com/.
2. Configure any availability and time slot settings that you need depending on when your team is available.

How we’ll use Calendly

We will use Calendly like a secretary. Calendly will access your calendar and let visitors choose from available time slots and make an appointment with you.

You can set up Calendly so that customers leave their mobile number or email, or create a Zoom meeting right away when they choose the time slot. Once the customer confirms the appointment, a calendar event will be marked down for all participants.

Zapier setup

‍

‍

‍

1. Create your account at https://zapier.com/.
2. Click on ‘Make a zap’.

How we’ll use Zapier

Zapier will be the ‘brain’ of our virtual AI assistant. We will build a zap following the flow design shown below and the no-code tools listed above.

Our zap will use custom branching logic (i.e. paths), which is available in the professional pricing tier of Zapier.

Flow Design

Here’s a flow diagram to show the logic our zap will follow:

Overview of Zapier setup

Below is a series of screenshots that demonstrate the main steps involved in our zap. Scroll down below the screenshots for a tutorial video covering all the steps in detail.

‍

For a detailed tutorial on how to set up this zap in Zapier, watch this video:

‍

‍

Your virtual AI assistant is ready!

Once you’ve followed all the steps mentioned above, your virtual AI assistant will be ready to start helping you qualify leads.

When a potential customer messages the virtual AI assistant’s phone number, it will first greet the user and then ask a lead-qualifying question.

After the customer responds, the assistant will either schedule a meeting or close the conversation nicely based on the customer’s answer. It will also record its interactions with customers in Airtable.

‍

‍

Next steps: upgrade your virtual AI assistant

If you want to take your assistant a step further, you can upgrade it to have the following capabilities:

• Add more lead qualifying questions and have a more complex interaction — this will help the interaction with potential customers feel more authentic and less automated, and can turn more enquiries into leads through nuanced conversation.‍
• Handle WhatsApp messages instead of SMS — ideal if your customer base prefers WhatsApp.‍
• Outreach a prospect once a new row is added to Airtable (using form submit), and trigger the chatbot flow once they reply — further qualifying your leads will save your team even more time.‍
• Use your virtual AI assistant in other scenarios, for example asking for feedback (through a survey) from attendees after an event.

‍