Know what keeps a customer past contract end.
Tell Simulatte the problem. It works out what to ask, then interviews customers who left from one link, or gets an early read from simulated customers on the plans and in the markets you describe.
Illustrative example. Numbers and quotes are invented to show the format. It plays both outcomes.
Where retention plans get stuck.
The churn score names who, not why
The model flags the accounts at risk. It can’t tell the retention team what to say when they call.
Offers get tested on live customers
Every retention trial gives discounts to people who were staying anyway, and you only learn which worked a quarter later.
Plan names are decided in a meeting room
Plus, Max or Unlimited Pro: the team argues for a week, and nobody asks a customer what they think they’re buying.
Four calls telecom teams make every quarter.
The loyal, the wavering and the ones who left.
Simulated customers
MinutesSimulated customers, modelled on real people, from ready-made audiences in 12 countries, including the UK, the US, Germany, Spain, the UAE, Saudi Arabia and Qatar. Or describe your base and build one. Try an offer on them before it reaches a single real bill.
Your own customers
As people respondSend an interview link to customers who ported out or are near contract end, or tell us who you need and we’ll source them, quoted per study. Interviews run in English, Hindi, Tamil, Arabic and Spanish, and you read every answer in English with the original alongside.
Priced in credits. A quick ask on plan names starts at 300 credits; a survey of 100 people answering 5 questions runs 2,500–4,000. Interviews with your customers are 500 credits each, charged only when completed. Survey links are free.
See pricing →Asked by telecom teams.
We already have a churn model. What does this add?
The why, and what to offer. Your model says who is at risk; simulated customers and interviews tell you what would keep them, so the retention call has something to say.
How well do simulated customers pick who will leave?
In a blind test on a telecom operator’s customers, simulated customers scored 0.75 at ranking who was most likely to cancel, where 0.5 is a coin toss. A model trained on that company’s own data still did better, and we publish that too. See the audits →
Can we compare retention offers without giving them away?
Yes. Run the offers past simulated customers by market and plan first, then check the winner with customers near contract end before it reaches anyone’s bill.
Can we interview customers in Arabic?
Yes. Interviews run in English, Hindi, Tamil, Arabic and Spanish. Each customer picks their language and hears a native voice. You read the results in English with the original quote alongside.
Are simulated customers real people?
No. They are simulated customers, modelled on real people. Their answers are directional. Confirm the decision with customers who actually left.
Hear from the customers who left last quarter.
Test the retention offer today. Hear the reasons from customers who ported out this week.