Post by ruhaimaromana22 on Nov 5, 2024 6:57:19 GMT
The previous article discussed routing conversations to responsible employees, limits on conversations processed by an employee at the same time, and the need to warn the client if it is not possible to respond immediately.
Today, we will consider three approaches to distributing conversations – uniform distribution, random distribution within the limit, and manual analysis. And let us also recall two almost social media marketing service obvious facts. Fact : clients expect different response speeds in different channels, so it is worth setting channel priorities relative to each other. Fact the speed of the first response is especially important – that is why in our platform, requests that are not processed for a long time are highlighted. Details are below.
3. Managing the request queue: distribution algorithm
Large companies most often choose uniform distribution, when new requests are sent to the operator with the lowest load. Requests from the queue are distributed evenly among all operators in the network.
Example:
A limit of 5 requests is set for all employees. Natalia has one dialogue in progress, Tatyana has three, and Maxim has five. Natalia will receive a new request from the queue. Everyone is happy.
Some online stores choose random distribution within the limit - with this order, requests are sent to an employee randomly selected from those whose limit allows them to take this dialogue. This approach is good with a small load, affects the motivation of employees to be online, and if there is a KPI, operators are on equal terms.
Example:
A limit of 5 requests is set for all employees. Natalia has one dialogue in progress, Tatyana has three, and Maxim has five. A new request from the queue can be received by Natalia or Tatyana. The system will not give the consultant Maxim a sixth dialogue, since it exceeds the limit. Fair.
There is also manual parsing - with this approach, operators independently select requests from the general queue and take them into work.
Example:
Operators Natalia, Tatyana and Maxim see a request from Ivan Petrovich in the queue. Ivan Petrovich is an important client, everyone knows him. Each of the operators is ready to consult such a client. "Mine!" — Natalia, Tatyana and Maxim say in chorus. Ivan Petrovich is pleased.
Another example:
There is an angry message in the queue. The client is angry, swearing and wants to solve the problem as soon as possible. The consultant Natalia is busy, and Tatyana decides that Maxim can handle it better. Maxim skips the negative dialogue in the hope of returning to it someday. The client is dissatisfied.
And another one:
"Sorry, he doesn't speak Russian. I need help or consultation with delivery service" - the operators see the request in the queue and don't understand anything. There is no time to figure it out, because there is a dialogue from Ivan Petrovich in the queue. The client remains misunderstood
Obviously, this approach is only suitable for small firms, where all consultants are enthusiasts. In a typical case, operators will try to ignore unclear and negative messages and those dialogues that threaten to drag on for a long time.
Almost always, the best solution will be uniform or random distribution within the limit.
4. Set priorities
Email is usually perceived as a slower way of correspondence, but there is a gradation among other text channels. Many LiveTex clients have noticed that those who contacted in some channels are willing to wait longer than in others.
And some companies are contacted through certain channels by higher-priority clients.
It is precisely in order to respond first in the “fastest” and/or “targeted” channels that the LiveTex platform has a priority setting. A request from a channel with the first priority will be distributed to the first available operator, even if it was received later than requests from other channels.
Today, we will consider three approaches to distributing conversations – uniform distribution, random distribution within the limit, and manual analysis. And let us also recall two almost social media marketing service obvious facts. Fact : clients expect different response speeds in different channels, so it is worth setting channel priorities relative to each other. Fact the speed of the first response is especially important – that is why in our platform, requests that are not processed for a long time are highlighted. Details are below.
3. Managing the request queue: distribution algorithm
Large companies most often choose uniform distribution, when new requests are sent to the operator with the lowest load. Requests from the queue are distributed evenly among all operators in the network.
Example:
A limit of 5 requests is set for all employees. Natalia has one dialogue in progress, Tatyana has three, and Maxim has five. Natalia will receive a new request from the queue. Everyone is happy.
Some online stores choose random distribution within the limit - with this order, requests are sent to an employee randomly selected from those whose limit allows them to take this dialogue. This approach is good with a small load, affects the motivation of employees to be online, and if there is a KPI, operators are on equal terms.
Example:
A limit of 5 requests is set for all employees. Natalia has one dialogue in progress, Tatyana has three, and Maxim has five. A new request from the queue can be received by Natalia or Tatyana. The system will not give the consultant Maxim a sixth dialogue, since it exceeds the limit. Fair.
There is also manual parsing - with this approach, operators independently select requests from the general queue and take them into work.
Example:
Operators Natalia, Tatyana and Maxim see a request from Ivan Petrovich in the queue. Ivan Petrovich is an important client, everyone knows him. Each of the operators is ready to consult such a client. "Mine!" — Natalia, Tatyana and Maxim say in chorus. Ivan Petrovich is pleased.
Another example:
There is an angry message in the queue. The client is angry, swearing and wants to solve the problem as soon as possible. The consultant Natalia is busy, and Tatyana decides that Maxim can handle it better. Maxim skips the negative dialogue in the hope of returning to it someday. The client is dissatisfied.
And another one:
"Sorry, he doesn't speak Russian. I need help or consultation with delivery service" - the operators see the request in the queue and don't understand anything. There is no time to figure it out, because there is a dialogue from Ivan Petrovich in the queue. The client remains misunderstood
Obviously, this approach is only suitable for small firms, where all consultants are enthusiasts. In a typical case, operators will try to ignore unclear and negative messages and those dialogues that threaten to drag on for a long time.
Almost always, the best solution will be uniform or random distribution within the limit.
4. Set priorities
Email is usually perceived as a slower way of correspondence, but there is a gradation among other text channels. Many LiveTex clients have noticed that those who contacted in some channels are willing to wait longer than in others.
And some companies are contacted through certain channels by higher-priority clients.
It is precisely in order to respond first in the “fastest” and/or “targeted” channels that the LiveTex platform has a priority setting. A request from a channel with the first priority will be distributed to the first available operator, even if it was received later than requests from other channels.
