How long does it take to set up a white-label AI demo?
A first white-label AI demo takes about ten minutes, most of it spent choosing a subdomain and uploading a logo. Here is the setup order, what to have ready beforehand, and the limits to know before you start.
You can have a first white-label AI demo live in about ten minutes, and most of that is choosing a subdomain and uploading a logo. The demo itself is three decisions: which trade, which layout, and the prospect's business name. Everything else you do before sending the link takes longer than the build.
That is the short answer. Below is the order I would do it in, what to have open in other tabs before you start, and the part of the process that no tool speeds up, which is checking the demo before a real person opens it.
How long does it take to set up a white-label AI demo?
On Lumina Sales, about ten minutes for the first one and less for each one after. The Build it step is where that time goes, and the setup splits into two kinds of work.
The first kind happens once per agency. You choose a subdomain, upload a logo, set your colors, and paste in your booking link. Your subdomain lives under luminasales.io, so there are no DNS records to add and nobody to email for activation. Compare that with a tool like Smartlead, whose own white-label guide has you request activation from support, add TXT and CNAME records, then confirm so they can switch the subdomain on. That is a fair process for a custom domain. It is also waiting on someone else's inbox, which does not help when you promised a demo by Thursday.
The second kind happens once per prospect. You pick the trade from the 100+ packs, pick a layout, and type the business name and website. That is the part that takes a couple of minutes.
What do you need ready before you start?
Four things, and none of them are technical.
A logo file that looks fine on a dark background and a light one. A transparent PNG saves you a second upload. Your brand colors, as hex codes if you have them. A booking link that goes somewhere real, because the demo points prospects to it. And the prospect's business name and website address, spelled the way they spell it.
The booking link is the one people skip. A demo that ends at a dead link undoes everything the receptionist just proved. Open your calendar link in a private window first and make sure a stranger can book from it.
If you do not have a logo yet, make a plain wordmark in a free design tool and move on. A plain wordmark rarely costs a deal. A demo that never gets sent costs one every time.
What are the setup steps, in order?
Here is the sequence I follow.
- Pick your subdomain and upload your logo, colors, and booking link. This is the once-per-agency part.
- Start a new buildout and pick the prospect's trade. If the trade is on the list, the pack brings in services, price bands, objections, and booking language written for that business. If it is not on the list, the generic pack covers anything appointment-based and is written to read as industry-neutral.
- Type the prospect's business name and website. The agent talks as that business, and the website audit fetches their real domain.
- Arrange your nodes. A node is one working piece of the system, like the voice receptionist, the chat agent, or the website audit. Our post on the six demo nodes covers what each one proves and how to order them.
- Save it as a draft. Drafts are unlimited and cost nothing, so build the next five prospects' demos now if you want to.
- Test it, then publish it.
Steps one through five take the ten minutes. Step six is where I would spend another ten.
Why test before you publish?
Because the prospect knows their own business better than your demo does, and they will find the gap in the first thirty seconds.
Call the receptionist yourself. Ask what a first-time customer would ask: price, hours, whether there is room this week. Listen for the business name, which the agent should say correctly. Run the website audit on the prospect's domain and read the findings against the site, because the prospect will. Check the ROI figure to see if the numbers you seeded are ones you could defend out loud.
A voice-agent vendor's demo guide recommends testing with at least five sample calls before a prospect sees anything. I would not argue with that number. You spend a few voice minutes, and our post on the voice allowance does the arithmetic on how many testing calls still leave you room for real prospects.
Then publish, and note what the page says near the bottom. It states that the AI is a demo and that anything it books is simulated. That line stays on every published demo. Our post on what prospects actually see explains why we will not remove it. Tell the prospect about it before they find it.
Why does speed matter here?
Because the demo is only worth building if it gets sent while the conversation is warm, and the home service trades you sell into are often slow to answer their own leads.
Valve+Meter, a performance marketing company, tracked response times at 466 home services companies over five days. By the end, only 60% had responded to a service request at all. In the study's words, a company that had not responded within the first six hours had a 75% chance of never responding. That is a small sample from one firm, so treat it as a signal and not a law. The gap it points at is the one the receptionist in your demo fills.
The same logic applies to you. A prospect who says "send me something" on Tuesday and gets a working link in their own trade by Tuesday afternoon is in a different position than one who gets a deck on Friday.
Where does this setup have limits?
Three, and you should know them before you start.
First, one published demo at a time. You can hold unlimited drafts, but publishing a new one takes the previous link offline. If you build five demos on Monday, you cannot send five live links at once. You publish each the morning of its call, or you run one prospect at a time. The setup time is short because the structure is narrow. If your outbound sends dozens of links a week, this is the wrong tool.
Second, the caps. You get 100 live voice minutes, 500 chat messages, and 50 website audits a month, and a single call ends at five minutes. Testing eats into that. Build it into your plan.
Third, the page lives on a subdomain of ours. I am not claiming you can put it on your own root domain. If that matters to you, ask before you subscribe.
This is also not for someone who wants a finished, hand-tuned site for each prospect. The speed comes from the packs and the layouts. You get a strong default quickly. You do not get a blank canvas.
FAQ
Do I need to install anything or set up voice infrastructure?
No. The voice and chat agents run on our infrastructure, so there is no API key to find and no phone number to buy. Your prospect does not create an account or install anything either. They open a link.
Can I build demos for several prospects in advance?
Yes. Drafts are unlimited. The limit is on what is published, which is one demo at a time, so build ahead and publish the day you need it.
What if my prospect's trade is not in the pack list?
Use the generic pack. It covers anything appointment-based and is written as industry-neutral, not as a template with the trade left blank. If you want a dedicated pack, request it. Requests from subscribers get built first, and niche packs are written by hand.
Does the prospect see the Lumina name anywhere?
Not on a published demo. They see your logo, colors, subdomain, and booking link. The only thing always visible is the sandbox disclosure.
What does it cost?
The platform is $49 a month. Builds you order for clients are priced separately at wholesale on a 30-minute call, and you set the retail price. The pricing section has the full limits table.
Do this today
Sign up at app.luminasales.io/signup, pick your subdomain, and build a demo for the next prospect already on your calendar. Time yourself. Then run five test calls on it before you send the link, and send it the same day.