Everything Linkubit actually does
Every one of these is shipped and running today. They alternate: the things any tool like this has to get right, and the features that set Linkubit apart.
What Linkubit does, feature by feature
It sets itself up from your website
Type your business name, paste your website, drop in a brochure or price list. Linkubit reads the pages and writes the whole setup itself — persona, questions, rules, scoring, follow-ups, even your first campaign. Minutes, not a project. Nothing to build.
Four channels, one customer
WhatsApp, email, Instagram DM and Messenger in one inbox. Someone who emails Monday and WhatsApps Thursday is one contact with one history.
Rules enforced in code, not the prompt
Write a rule in plain words. It’s checked outside the AI, so it isn’t a suggestion: “never give a price before you know their budget” is refused and stopped before sending.
Take over, and hand back on purpose
A rep steps in and the assistant stops, mid-sentence if it has to. It only resumes when handed back deliberately — never on its own because a timer ran out during a negotiation.
Take the payment inside WhatsApp
Customer’s ready, a UPI card with your catalogue price appears in the chat, they pay, and the deal is marked won with the revenue recorded. Money goes straight to your account.
The 24-hour window, before you type
WhatsApp only allows a free-form reply inside the day after a customer writes. Outside it the composer offers your approved templates and nothing else — you find out before you type, not after Meta rejects the send.
It can only quote prices from your catalogue
The assistant cannot invent a number. A price it states has to come from a product it just looked up. Same for photos: only a product it actually read this turn.
A score you can read line by line
Every lead’s score shows its working: budget, timeline and intent, each with the points it added or took away. Not a mystery number.
Follow-ups written from the real conversation
At 1 hour, 6 hours and 24 hours it sends a nudge written from what was discussed — not “just checking in!”. A reply at any point stops the rest.
A pipeline that refuses illegal moves
Stages carry entry requirements. Drag a deal somewhere it cannot go and it is refused with the requirement named — so the board tells you what to capture next instead of snapping back in silence.
Hindi voice notes, answered in Hindi
Send a voice note in Hindi, get a reply in Hindi. Hinglish is fine. Built for India first, not translated after.
Pre-flight before the first send
A campaign is checked across the whole audience before anything goes out: who has consented, who unsubscribed, who is missing a number. You see who would be refused and why, because the first send is where deliverability damage happens.
The right photo, only for a product it read
The first time a reply names a product, its photo, video or PDF goes with it — the one you uploaded, captioned. A picture is a claim about what the thing is, so it is only ever attached for a product the assistant just looked up.
Sequences stop on reply
A reply stops the remaining steps, opens a conversation and starts qualification. Nobody gets step three after they already said yes.
Walk any deal back to the ad
Revenue to deal to lead to conversation to campaign to audience. Nothing is stamped on the lead — attribution is worked out from the touchpoints, so switching first-touch to last-touch really recomputes, and closed revenue stays where it was.
Meta and Google spend, joined to revenue
Ad accounts connect in a click. Spend and leads come in daily and sit beside the deals they became, so the number you see is what an ad earned, not what it cost. New campaigns are built here too — always paused until you switch them on.
Tell the networks which leads were real
Meta and Google optimise toward whatever you report back. Report form fills and you buy form fills. Linkubit sends them the leads that actually qualified and closed, so the next campaign hunts for those.
Anything that spends waits for a person
A discount past its limit, a new ad campaign, a change to its own rules — the assistant can draft it and never enact it. A person decides, with the number in front of them.
It will not deny being an AI
It doesn’t volunteer it, and it never lies about it. A reply that claims to be a person is caught by a check that sits outside the model and cannot be configured away.
Audiences from filters or a spreadsheet
Filter on anything you know about a contact — your own qualification fields included — and watch the count update. Or import a sheet. Either way it is deduplicated before it lands, not after.
Deals open themselves, priced from the catalogue
When a lead qualifies, the products they showed interest in become one open deal at your list price. The value is never a figure the assistant said in a chat — those are unverified — it is what the catalogue says.
Answers that show their source
Upload your brochures, policies and price lists. Every answer drawn from them points at the document and page it came from, so a claim can be checked rather than taken on trust.
Suggestions drawn from your own conversations
Every day it reads what happened across your chats and proposes a change with the evidence attached: a question to reorder, a field to add, an objection that keeps coming up. Accept and the setting changes; nothing changes on its own.
Roles decide who sees what
Every screen and every action needs a permission, and a role decides which records it covers — own leads, the team’s, everyone’s. A screen somebody may not open simply isn’t in their sidebar.
Scores from what a customer did, not just what they said
Opened the price list, replied within the hour, went quiet for a week — behaviour is scored beside the answers they gave, worked out fresh each time rather than stored, so a score is never stale.
Tasks, meetings and the next best action
Follow-ups the assistant creates for you, meetings with outcomes, and for every open deal a suggested next move — ask the missing question, answer the objection, schedule, escalate. Surfaced for a person, not done silently.
Two reps on one customer, caught
Open a conversation and colleagues see you are in it. Two people answering the same person with two different prices is the mistake this exists to prevent.
WhatsApp templates, synced with Meta
Write a template, submit it, and its status arrives from Meta as it changes — approved, rejected, paused. A campaign can only pick one that is approved, so a send never fails on a template you thought was live.
An out-of-office is never answered
Bounces, mailing lists and auto-replies are recognised before they become a conversation. An out-of-office is filed where a person can see it and never replied to — the loop where two robots email each other all night cannot start.
A welcome message with buttons to tap
A new customer gets a greeting you wrote, with quick-reply buttons for the things people usually want. One tap and the assistant is already on the right track.
Your WhatsApp bill, before Meta sends it
Meta bills per conversation, days late. The dashboard shows what it has charged and what the days it hasn’t billed yet will cost, drawn dashed so you can tell the two apart — beside ad spend and your credits.
Consent recorded, unsubscribe honoured everywhere
Every contact carries where and when they said yes. STOP on WhatsApp or one click in an email suppresses them across every channel at once, and nothing — not a campaign, not the assistant — can reach them after.
Assistant against your reps, honestly
Win rate, time to first reply and deal size, assistant beside human, counted only on conversations that actually closed. You see where it is better than the team and where it isn’t.
Erase a person, or the whole workspace
A customer asks to be forgotten and one action removes them from every table, every file and every export. Leave Linkubit and the same runs for the whole workspace. Nothing readable is kept anywhere the erasure can’t reach.
A figure without a currency is asked about, once
A customer who says “around 5,000” gets asked whether that is rupees or dollars — once, not every turn. What they said is kept as they said it; the deal is written in your currency so every forecast adds up.
Messages can’t be edited, and everything is logged
What was said stays exactly as it was said; a correction is a new message. Every change to a deal, a rule or a role is written down with who made it and when, so a dispute has a record to go to.
It works towards a goal, not through a script
Say what a qualified enquiry needs — what they want, quantity, budget, commitment. The assistant asks only for what is still missing, one question a message, never twice for something the customer already said. Code decides when the goal is met, not the model, and a met goal opens the deal.
In depth, by what you are looking for
See it answer a real enquiry
The demo is the live product with a business already loaded into it.
Ready to put your sales on autopilot?
Start freeNo coding required. No credit card required.