How to Recover Old Customers and Gain 4000$ —
Bulk Calls Case for a Health Food Delivery Service

Case

MF Kitchen (company’s name was changed because of NDA) — a healthy food delivery service. The company makes their food according to the nutrition program a client has chosen. A delivery guy brings a set of 6, 8 or 10 dishes every second day.

This is a popular thing among modern consumers, so the delivery service has a lot of actual clients and even more are old ones — those who make an order once and then never call again for some reason.

The company needed to recover these clients by informing them about new promotions and persuading them to make a second order.

Problem

MF Kitchen informed their clients by SMS. It was affordable and fast, yet the company cannot track if clients got a message or if they made an order after they got an SMS.

The company couldn’t calculate the tool effectiveness and a conversion in sales so they decided to find another service for tests.

The company’s indicators before they appealed to Effebot.

Solution

Effebot experts offered MF Kitchen to make a voice call with an IVR menu. It would have saved the employer from the necessity to send SMS manually, because the robot would have made it automatically after the client request.

Experts made a simple algorithm for the interactive menu. The employer uploaded the phone number base, an audio clip with an offer and launched the mailing.

The audio sample for a delivery service: Hello! I’m Jane from MF Kitchen. As our client you can participate in the campaign to win one week of free healthy dish delivery. Press 1 to get all the details in SMS.

What was automatized?

Implementation

Effebot called phone numbers from the loaded base and briefly informed contacts about a new promotion.

After that, the client was suggested to get the promotion details by an SMS:

If a client pressed 1

the robot sent him an SMS with detailed information and a link to the site with a tracking tag.

OR

If a client pressed 2

or other keys the robot finished the call.

The service has two options for answer recognition. In this case the employer decided to do some test — to check what works better: keys or voice?

They used 1500 contacts for tests: one half got a call with an answering through pressing a key and the other half got a call with an answering through voice.

Conversion in answer appeared to be better with a key pressing: 21,5% against 7% of the answering through voice. So in calls to remain clients the robot used a key pressing script.