NudgeLink isn't a wrapper around a template engine. It's one system that finds the right people, writes something worth reading, sends it without burning your account, and hands you the replies. Here is all of it — including the parts that aren't finished yet.
Relevance decided before a message exists, not after it lands badly.
Every profile gets a 0–100 score with a written rationale: what matched your ICP, which signals pushed it up, what would make it a no. Leads below your threshold are skipped but stay visible as data — the default is 60 and you move it per run.
Describe your ideal client once and NudgeLink searches for people who match, scoring each before it reaches you. Every run shows its funnel — found, deduplicated, excluded, filtered, scored, suggested — plus the near-misses, so you can tell a strict threshold from a narrow ICP.
Group leads by event, segment or ICP. Each campaign carries its own context note woven into every message, its own scoring criteria, its own language and tone, and the specific accounts it's allowed to send from.
A Premium or Sales Navigator seat is detected automatically and lifts your discovery ceilings — the 120-a-day profile-view cap is waived and the daily run cap rises to 200. A free seat keeps the conservative limits.
One specific detail, built into one specific opener. No variables.
The model reads posts, role history, skills and recent activity, then builds the opener around one concrete detail it actually found. No {first_name} variables, no template with the nouns swapped.
Every draft you edit or approve is kept as a tone sample, and later messages are generated against your own writing rather than a generic professional register. The more you correct, the less correcting it needs.
Messages go out in a language you and the lead both speak at B1 or better — Polish, German, Spanish and more — picked from the lead's languages section, the text of their profile and their location, with English as the fallback. Set your languages per account or per campaign.
A restricted LinkedIn account is the expensive failure in outbound. Most of this section exists to make that not happen.
Connect several LinkedIn accounts and NudgeLink routes sending across them — each with its own warm-up ramp, its own timezone and its own daily ceiling — so volume never piles onto one identity.
A newly connected seat doesn't start at full speed. Its invite allowance climbs 19 → 21 → 22 → 23 a day across the first four weeks and then holds there — identical for every seat, with no plan-based boost.
Before the connection request, a lead may get a like or a contextual comment on a recent post. Deliberately not every lead and not always both — the mix is varied on purpose, because identical behaviour on every profile is exactly what looks like a script.
Actions are spread across the seat's own local 08:00–20:00 window, seven days a week, at jittered intervals — never a burst at the top of the hour.
A captcha or a LinkedIn warning auto-pauses the seat, and a paused seat stops receiving work instead of digging itself in deeper. Credentials are encrypted at rest. Sending runs on hosted infrastructure, with each account on its own fixed IP near your location — not a browser extension that has to stay open on your laptop.
The part most tools treat as someone else’s problem.
Drafts land in a review queue to approve, edit or reject — and your edits feed the voice model. Nothing goes out unattended unless you explicitly switch the account to auto-send.
Every inbound reply is tagged positive, neutral, question, rejection or spam, so the conversations worth your time surface first and a rejection stops the sequence instead of following up into a no.
When LinkedIn stays quiet and the profile exposes an email address, the sequence switches channel with a different angle rather than a copy-paste. If there's no address the sequence simply completes — we don't buy, guess or pattern-match email addresses.
Enter your average deal size and the dashboard turns qualified replies into an estimated pipeline figure. It assumes a 25% close rate — stated plainly, so you can discount it with your own number.
Scout takes your campaign's job titles, industry and location and searches the public web — LinkedIn profiles, company sites, news, directories — for the people in those roles. A person is campaign-ready only once their LinkedIn profile is confirmed; uncertain matches go to you for review, never an auto-pick. Every fact carries its source URL, every run shows what it searched, and you're billed only for campaign-ready people. Scout has no ability to contact anyone.
Included on the Scale plan — no waitlist, nothing to request.
A HubSpot integration and an AI-assistant connector are both in development. Neither ships today — here's exactly where both stand.
Score your own leads and read the messages NudgeLink writes — before you send a thing. Start free in minutes.
Start for free