How Do You Get Set Up on Dubb in One Session?
You get set up on Dubb in one session by doing five things in order: understand which side of the platform you are on, connect the account you already send from, record one video and send it to yourself, connect your contacts, and set one task running in the agent. Everything else can wait until you have used those five.
First, learn the two sides
The platform has a recording side and an AI side, and a first session goes badly when both are treated as one screen. The recording side, Dubb Studio, is where videos are made, sent and tracked. The AI side, Dubb Agent, is where you type or speak an instruction and a task happens.
The dropdown between them is the main navigation. Learn that movement first and the rest of the interface stops being a list of unfamiliar menu items.
Connect the account you already send from
A video that arrives from an unfamiliar address is a video from a stranger. Connect the mailbox you already use, so the recipient sees the name and address they recognise. This is the step most likely to be postponed and it is the one that determines whether the first campaign feels like outreach or spam.
While you are there, connect the contacts source rather than starting a separate list. If you already run a CRM, connect it. If not, the built in contacts are enough to begin.
Record one video and send it to yourself
Do this before anything clever. Record a plain webcam video, thirty seconds, one idea, and send it to an inbox you control. Open it on a phone. Watch what the first three seconds look like as a looping preview, because that preview is the entire click decision for a recipient.
That test teaches three things at once: what the recipient sees, whether the tracking works, and how long a video really needs to be. Almost everyone shortens their second video after watching their first.
Set up the channel you already use
The common failure in a first session is choosing a channel you do not currently use. If your existing approach is a social platform or an inbox, keep it and remove the manual steps around it. Adding a new channel and a new tool in the same week produces neither.
Give the agent one task, not ten
Pick the task you do manually and resent most: finding prospects, writing the first message, chasing replies, or preparing a follow up. Ask the agent to do that one task completely, and watch how it handles the pieces.
Only after that task works end to end do you add the second. A person with one working routine uses it every day. A person with six half configured routines uses none of them.
Connecting a scheduling tool is usually the second task people want, and connecting Buffer to your agent explains that setup end to end.
A first session checklist
| Step | What done looks like | Why it matters |
|---|---|---|
| Find both sides | You can switch between agent and studio without hunting | Every other step assumes it |
| Connect your mailbox | A test video arrives from your own address | Recognition decides whether it is opened |
| Connect contacts or CRM | Your real contacts are visible | No parallel list to maintain |
| Record and send one video | You have watched it on a phone | You see what the recipient sees |
| Set one routine | It ran once without you | Proves the loop before you scale it |
| Pick one channel | It is one you already get replies on | New channel plus new tool is two experiments |
What to leave for week two
Landing pages, custom integrations, bulk sends and automations are all worth doing and none of them belong in a first session. They are the parts that make a working setup faster, not the parts that make it work.
The test for whether you are ready for week two is simple. Have you sent a video, watched a view register, and followed up because of it? If yes, the loop works and everything else is optimisation. If no, adding tools will not fix the missing loop.
What to measure in week one
Three numbers are worth looking at and the rest are decoration. How many videos you sent, how many were opened, and how many replies you got. Send volume tells you whether a habit is forming. Opens tell you whether the preview is earning the click. Replies tell you whether the video was worth sending.
Ignore view duration in week one. It is interesting later, when you have enough videos to compare styles. In the first week it mostly reflects how long your video is, which you already know.
The number that matters most is the one most people skip: replies. A video that gets opened and never answered is a video that was watched politely. If opens are healthy and replies are not, the problem is in what the video asked for, not in whether it played.
- Sent. The only number you fully control, and the only one that predicts anything later.
- Opened. Low opens on a healthy send count point at the preview, not the list.
- Replied. The one that justifies the channel. Track it from day one or you will argue about it later.
- Booked. If a reply says yes, note it. Ten of those in a quarter changes how the team treats video.
Write the three numbers down weekly. A four week record of sends, opens and replies is enough to tell whether video is working for your audience, and it is more evidence than any opinion about whether people prefer email or video.
Common mistakes
- Recording ten videos before sending one. Send first. The reply teaches you more than the recording.
- Connecting a mailbox you never check. Replies arrive there and stall.
- Starting a new list. Connect the one you have.
- Setting up six routines. One working routine beats six configured ones.
- Adding a channel you do not use. Two experiments at once produces no information.
- Skipping the phone test. Most recipients open on a phone, so review on one.
Frequently asked questions
How long does setup actually take?
The five steps above fit in an hour if the mailbox connection goes smoothly. The variable is usually access to the mailbox or CRM, not the platform itself.
Do I need to connect a CRM?
No. The platform holds contacts and supports bulk sending, which is enough to start. Connect a CRM when you already run one, so that the work does not live in two places.
What if I have never recorded a video of myself?
Record the first one on webcam with no props and no script, at thirty seconds, and send it to yourself. The discomfort is in the first three, not the first thirty, and the only way past it is to send one.
Should I use the AI side first or the recording side first?
Recording. A video with no agent is already useful. An agent with no videos has nothing to send. Learn the send loop first and let the agent remove the work around it afterwards.
How many videos should I send in week one?
Between five and ten, to people you would have emailed anyway. The point is not volume. It is finding out which of your existing messages land better as a video.
If you want to run the five step setup with your own accounts, start a free 7-day Dubb trial.
One session, both products working
Dubb Studio records and sends the video, Dubb Agent runs the follow up. Set both up while the trial is still free.
