Role
Retention
How quickly do you find out that a client intends to leave? If it is at the moment of notice, you have one conversation and one argument - usually price, because there is no time left for anything else.
Intro call
Four questions we would start with
How quickly do you find out about an intention to leave?
Notice is the end of the process, not its beginning. The decision is made weeks earlier, and then it can still be influenced by something other than a discount.
What do you not know about a client until they call to terminate?
Usually: whether they used what they pay for at all, how many times they raised a problem and how fast they got an answer. Those are the things that explain the decision, and you have them.
When do you decide what to offer?
If it is during the conversation, a discount usually comes out, because it is the simplest. A prepared proposal is often cheaper and more effective.
What is left after a conversation that did not save the client?
If it is nothing, you lose the most valuable information in the whole process - the reason the client left, given by them.
We take over the detecting and the preparing, you keep the conversation
Catching signs of leaving, preparing the picture of the client and laying out the possible proposals can be taken off the desk. The conversation and the decision what to offer stay with a person.
- A signal is material, not a verdict. - A drop in usage can mean the client is in trouble or that it is their low season. The person who knows that company tells the difference.
- The proposal is prepared, not invented in the conversation. - A set of options with the cost of each, so that a discount is not the first reflex.
- The leaving conversation is recorded. - The reason the client gives is the best product information you will get - provided it goes further than a note.
Scope
What can be taken off retention work
Eighteen example tasks. Each one has a human approval point. Below is a selection, and the scope we fit to your process.
Signs of leaving
- A drop in usage against the client's own history (assessment: the specialist)
- An unresolved case open longer than usual for that client
- No contact for longer than is typical for the segment
- An approaching notice or renewal date
- An enquiry about termination terms
Preparing the conversation
- The picture of the client: what they bought, what they use, what they do not
- The history of tickets and response times
- Earlier promises and how far they were delivered
- The client's value and the cost of losing them (assessment: the specialist)
- Similar cases and what worked in them
The proposal and the conversation
- A set of possible proposals with the cost of each (choice: the specialist)
- Pointing out features and services the client has not used
- Preparing the terms for the manager's approval
- Recording what was agreed in the conversation (approved by: the specialist)
Learning from departures
- A register of the reasons clients give for leaving
- Grouping the reasons by theme (conclusions: the specialist)
- How effective the proposals were
- Material for product and service about why clients leave
A day's work
Today
- You learn about a departure from the notice
- Before the conversation you assemble the picture of the client from several systems
- You invent the proposal during the conversation
- A discount is the first and often the only argument
- You record the reason for leaving in a field nobody reads
- You do not know which proposal actually works
With AI assistance
- The signal comes weeks before notice
- The picture of the client is waiting, ready, before the conversation
- You have a set of proposals with the cost of each
- A discount is one of the options, not a reflex
- The reasons for leaving add up to themes and reach product
- You can see what worked in similar cases
- Early detection
the conversation happens before the client's decision, not after
the signals are monitored against the client's history
- A proposal that fits
the proposal answers the real reason
the picture of usage is available before the conversation
- The cost of keeping a client
fewer discounts given by reflex
the options have their cost calculated, so a cheaper one can be chosen
- Knowing why clients leave
product learns what pushes clients away
the reasons are grouped rather than only recorded
We start with the picture of the client, not with the signals.
- Step 1
The picture of the client before the conversation
What they bought, what they use, what they raised. It changes the character of the conversation before any signal appears.
- Step 2
A register of reasons for leaving
Collecting and grouping. It gives material that usually surprises product.
- Step 3
Signs of leaving
One, the least ambiguous - usually a drop in usage. False alarms we fix or switch off.
- Step 4
Proposals and how well they work
A set of options with costs and measurement of what works. It needs several dozen conversations.
Common questions
Questions that come up at the first meeting
Will this role be replaced by an automated campaign?
An automated retention campaign is the simplest way to lose a client faster - and that is a risk we take seriously. A signal about a client drifting away should reach a person, not a sending mechanism. A specialist who is not hunting for those signals by hand has time for the conversations that actually change something.
Does this fall under the AI Act?
Monitoring how business clients use a service is not a high-risk use. With individual clients, care is needed with profiling that affects the terms offered to a particular person - we do not propose differentiating prices on the basis of predicted churn.
Still have a question that is not here? Book a call
Next step
When did you last talk to a client about leaving before they gave notice?
We will show you what retention looks like when the signal comes early and the proposal is prepared before the conversation.