People
You have met them before. Now you will remember.
Every person who comments, messages or joins through your channels becomes one record — their accounts on each platform, your whole history together, the tags and notes your team added, the leads and the things they asked for. Built entirely from your own conversations.
No follower lists. No imports. No enrichment. No broadcasts.
One record
The same person, however they reach you
Somebody comments on Instagram in March, sends a WhatsApp message in June and joins your Telegram group in September. Today that is three strangers. Here it is one person, with one timeline.
Their accounts, as chips
Each platform account you know them by sits on the record, with a link to open that conversation again.
A timeline you can read
Messages, comments, mentions, tags, notes, stage changes, group joins and what your agents did — in order, in plain language.
Linking is your decision
Nimply suggests a possible match and waits. Two accounts on different platforms are never merged on a guess.
Right there in the inbox
Open any conversation and the person's tags, stage, interaction count and last few events are in the sidebar before you type.
Leads
From a comment to a customer, without a spreadsheet
A lead captured by an agent, by your team or by the browser extension lands on a board attached to the person it came from — so you always know what was already said.
A board with stages
Drag a lead through your stages, set an owner, a value and the date you plan to follow up.
Follow-ups that find you
When the date arrives, the reminder reaches you in the app, on your phone and by email.
Push to HubSpot
Leads that carry an email or a phone number sync one way to HubSpot contacts, with the lifecycle stage mapped. Never the other way, never without contact details.
Or to anything else
Lead created and stage changed fire as outbound webhooks, so Zapier, n8n and Make can take it from there.
Requests
The things people actually asked for
A table for four on Friday. An order for two of the blue ones. A quote for a job next month. Your support or sales agent writes it down while it answers, and it joins a queue you can work through.
Captured in the conversation
Bookings, orders, quotes and appointments, with the details the person gave and a link back to the thread.
A queue, not a memory
Everything open in one list, moved along as you confirm or fulfil it. Log one by hand from anyone's profile.
Insight
Who keeps coming back
Counted from your own conversations, over seven, thirty or ninety days. No follower counts, no reach estimates, no inferred age or income — none of which we collect.
First-time against returning
How many people got in touch, and how many of them had spoken to you before.
Comments that became leads
The share of your commenters who turned into a lead over the period.
Superfans, badged on their own
Five or more interactions in ninety days earns a Superfan badge; anyone new this week is marked New. Both are rules on your own counts, not a guess about the person.
Your AI workforce
Agents that know who they are talking to
An agent answering a regular should not open with “hi, and welcome”. People is what lets it know the difference — and the rails are the same ones everything else here runs on.
Look up, tag, note, record
Four tools: find the person, tag them, leave a note, log what they asked for. Write tools follow the agent's action mode.
Allow and block lists by tag
Your Personal DM agent's allow list survives a display-name change, because it points at a person rather than a string.
Route by who it is
Assignment rules and automations can branch on a person's tags or lead stage: VIPs to a named teammate, new leads to sales.
Never sends on its own
Agents draft; you approve. Every run is recorded against the person so you can see exactly what was done for them.
What we refuse to do
A record of a relationship, not a profile
Every rule below is enforced in the product, not promised in a policy. They are the reason this can exist without turning your audience into a dataset.
Only people who spoke to you
No follower lists, no audience imports, no bought or scraped data, and no enrichment filling in details nobody gave you.
LinkedIn and Pinterest excluded
Their developer terms do not allow storing people this way, so those identities are refused at ingest rather than quietly kept.
Retention you choose
Six to thirty-six months, defaulting to twenty-four, and a platform's own shorter cap always wins. A nightly sweep enforces it.
Deletion that travels
One deletion removes the record, the notes, the leads, what your agents remembered, and the contact pushed to your CRM.
No inferred attributes
Nothing here guesses age, gender, income or interests. Automatic tags are counts of your own interactions, and AI suggestions stay suggestions until you turn them on.
Exports are accountable
Only owners and admins can export, and every export is written to the audit log.
More detail in the People section of our privacy policy and our data processing agreement.
People — frequently asked questions
Where do People records come from?
From your own conversations. When someone comments on your post, sends you a message, joins your Telegram group or is captured as a lead, they get a record. Nothing is imported: there is no follower list, no audience upload, no bought data, and no enrichment service filling in an email you were never given.
Does it link someone's accounts across platforms automatically?
No. Two accounts on different platforms are never merged on a guess, however similar the names look. Nimply can suggest a possible match, and a person with the same Telegram id across your channels is recognised as themselves, but every cross-platform link is a decision you make and confirm.
Which platforms are included?
Facebook, Instagram, Threads, WhatsApp, X, YouTube, TikTok and Telegram, wherever the inbox is switched on for that channel. LinkedIn and Pinterest are deliberately excluded: their developer terms do not permit storing people this way, so those identities are refused at the door rather than quietly kept.
How long is someone's data kept?
You set the window, from six to thirty-six months, and it defaults to twenty-four. Some platforms cap it lower and their cap wins: YouTube and TikTok identities are held for thirty days unless the person interacts again. A nightly sweep anonymises what is past its window and deletes records nothing keeps alive.
What happens when someone asks to be deleted?
Deletion propagates. It removes the record, their accounts, notes, tags, timeline, consents and leads; it deletes the memories your agents formed about them; it redacts the agent runs that quoted their messages while keeping the billing record; and it deletes the contact pushed to HubSpot through HubSpot's own erasure endpoint. Platform deletion callbacks do the same automatically, and an X account that disappears is picked up by a nightly check.
Can People send messages or newsletters?
No. There is no broadcast, no bulk send, no email or SMS channel, and no marketing list. People is a record of who you have spoken to and what was said. Replying still happens in the inbox, one conversation at a time.
Can my AI agents use it?
Yes, on rails. An agent can look someone up, add a tag, leave a note and log a request such as a booking or a quote. Write tools obey the same action mode as the rest of the agent, and an agent never messages anyone on its own — drafts go to you for approval.
Is it a CRM?
It is the social half of one. There is a lead pipeline with stages, owners, values, follow-up dates and reminders, and a one-way sync to HubSpot for leads that carry an email or phone number. There are no deals beyond that board, no invoicing and no email sequences — for those, push to your CRM and keep it as the system of record.
Stop meeting the same people for the first time
Connect a channel, switch on the inbox, and the records build themselves from the conversations you were already having.