A customer comes back after six months and asks “so what did we agree?” Whoever answers was not in that conversation, starts from nothing, and makes them tell it all over again.
The customer side: an answer that does not start from nothing
The file on a customer writes itself
Everything that already happened with a customer collects in one place: what they wrote, what they were told, what they bought, what they complained about. Whoever answers next — a person or an agent — sees it before writing a word.
- Everything said to them, in every channel, including a year ago
- What they bought, what they paid, and what is still open
- Which of you spoke to them, and what was said in that conversation
What the customer never told you, filled in from the web
A lead leaves a name and a phone number and stays a line in a table. The system fills in what is missing from a search on the web, so whoever calls already knows who they are talking to.
- A fuller picture than the form they filled in
- Which group of customers they fall into, so you know what to offer
- Something that fits them that nobody has offered yet
- Each filled-in detail says where it came from. A detail with no source is a guess
The employee side: the answer comes from someone the customer knows
The agent answers in an employee's name, and in their style
The customer wrote to Ronit in sales. The answer comes back from Ronit — the name they know, written the way Ronit writes. In whichever channel they used.
- The name the customer knows, not “the support team”
- That employee's own style: how they open, how much they spell out
- The customer's file standing behind every line of it
The employee only gets what actually needs them
Most of the messages in a day do not need Ronit. They need an answer. What has an answer gets one, and what does not waits for her, with the conversation and the file already open beside it.
- Questions that keep coming back are answered without reaching her
- What does reach her arrives with the background, not as a lone message
- She approves with one click, or writes it herself
Who decides what goes out in your name
Putting a person's name on an answer is a serious thing. So you set what goes out on its own and what waits for someone to approve it.
On its own, or on one click
The agent sends by itself, or it writes the answer and nothing goes out until the employee clicks. You can start with approval and change the setting later.
A person steps in and carries on
Every conversation is open in the conversations screen, with what was answered and where it came from. Anyone can step in and carry on from there, and the agent hands over when a person is needed.
No answer beats an invented one
When the file does not know, the agent says so. An answer going out in an employee's name has to be right, because they are the one who will have to explain it.
What gets better on its own
The file does not stay as it is. The system reads the channels through the day, so what a customer was told this morning is already behind the answer they get this afternoon.
The agent's instructions do not stay as they are either. The system looks at where an employee had to step in and answer instead of the agent, and recommends what to add or change — it does not rewrite the instructions itself, and the call is yours. Ronit is needed fewer times, and the answers get more accurate as time goes on.
This knowledge builds out of the conversations you are already having.
Who this is for
- A business where a customer has one person they deal with, and expects to hear from them
- A sales team where every rep runs their own customers
- Anyone whose customers explain themselves from scratch every time they get in touch