What happens when you press Share

When a support agent starts a consultation, our server creates a support session at once and gives the agent a six-digit code. The agent reads or sends it to the customer. The customer types it on the code-entry page, and that is the whole sign-in. There is no account for the customer to create and nothing to pay.

The code also carries a decision. It opens the direction the agent prepared. If the agent prepared a consultation in which the customer shows their screen, the customer is asked for consent and then invited to share. If the agent prepared one in which they show their own, the customer simply waits and watches. The customer does not choose the direction by typing something different, and cannot be pushed into publishing their screen by an agent who prepared the other direction.

A code can be used by one customer to join, and it expires three minutes after the agent created it. Both limits are deliberate. After a customer has joined using the code, it has done its work. After three minutes without anyone joining, it stops working, and the agent can simply create a fresh one.

If a customer sees a message that the code was refused or has expired, nothing is broken. It means more than three minutes passed, or somebody already used it. Ask the agent for a new code. That is all.

A six-digit code has only a million possible values. That is convenient to read over the phone, and it is also small enough that guessing is a real concern. Three things keep it safe in practice.

First, the code is short-lived. An unused code stops working after three minutes, so a guess has a very small window. Second, it is single use. Once a customer has joined with it, nobody else can take that place. Third, only codes that are currently live need to be unique. Once an old one has expired, the same digits can be issued again to a later consultation, which is how six digits can serve many sessions over a day.

There is also a human reason. A code that lives forever tends to be left in chat logs, written on sticky notes and forwarded to the wrong person. A code that dies in three minutes cannot be found next week.

The three-minute limit applies to the code and only to the code. Once a customer has joined, the live consultation is separate and is not cut off by a timer. A customer who joined at minute two is not thrown out at minute three. The consultation goes on until somebody ends it.

What does end it is a definite action. If the customer presses Stop sharing, or closes or reloads the customer page, the consultation ends and a new code is needed to rejoin. Switching to another browser tab does not end it.

When a customer is asked to share, the order is the following. The customer agrees. The customer clicks the control to choose a screen. The browser opens its own picker, and the customer chooses and presses Share. The two pages then exchange their connection notes through our server, as described in how browser screen sharing works, and the video begins to flow once the browsers have connected.

When it is the agent who shares, the agent sees the six-digit code first and then starts the share. We chose that order on purpose, and the stories article explains why. The code is shown beside the share button before the browser's picker opens, so the agent has it in hand before the browser's attention moves elsewhere.

It is a fair question, and the honest answer has layers. A guess has to land on a code that is live at that moment, which narrows the field to the few codes agents have created in the last three minutes, not all million. If a guess does hit one, it can only open the direction the agent prepared, and the customer's screen is still never published without consent and a click on the browser's picker.

Guessing attempts are also limited on the server side, so that trying many codes quickly does not work. None of this makes a code a secret to be guarded like a password. It is a short-lived ticket for one consultation, which is why we ask agents to hand it only to the person they are helping.

On the agent's side the code is an everyday object. The agent creates one when a customer is ready, reads it out or types it into a message, and watches the consultation page wait for the customer. If three minutes pass with nobody joining, the agent is not stuck with a dead page. They simply create another code.

That is why a code is cheap to replace and safe to lose. The code is only a ticket for one consultation, and getting a new one does not require the customer to do anything except read six new digits.

Enter the code soon after receiving it. If you are not ready, ask the agent to wait and create it when you are. Digits only: the page keeps only digits, and the Connect button waits until there are six.

If the code does not work, do not keep trying variations. Ask for a new one. For the rest of the customer's journey, the screen-sharing guide has the steps with screenshots, and teams who want to see how this fits into pricing can read the company overview.