Every campaign carries its own targeting — who NudgeLink looks for, how each lead is scored, and what your outreach is grounded in. It's the highest-leverage fifteen minutes in the product, and it lives on the campaign, not on your account: two campaigns can chase completely different audiences without interfering with each other.
What goes in
You answer these while creating the campaign, in your own words:
- What your company does — the factual statement of what you offer. This is the anchor of every message written for this campaign: drafts are forbidden from inventing a different company, product, city, or role.
- Your ideal customer — who you want to reach (roles, industries, geography, any signal you look for).
- Voice and languages — how the messages should sound, and which languages they may be written in.
From those, NudgeLink derives the structured targeting: a summary, scoring criteria, search keywords, and the location / industry / skill facets discovery searches with. You see all of it on the review screen before anything runs, and you can edit any of it there.
Be precise about location
If your offer is tied to a city or country — an on-site role, a local market — say so explicitly in your description and make sure the campaign's locations reflect it. Discovery filters candidates by location and the scorer treats out-of-market profiles as materially weaker fits, but only when the targeting actually carries the constraint.
Changing it later
Open the campaign and press Edit. You get the same screens you saw when you created it, pre-filled with your answers — change the targeting, the voice, the angle, or just the keywords, and save. No AI regeneration happens unless you ask for it, so a small correction never disturbs a voice you were happy with.
Saved changes apply to leads scored and messages written from that point on. Leads already in flight keep the sequence they started.
Suggested keywords
After a discovery run, NudgeLink learns from the leads you accepted and rejected. When that history points somewhere your current keywords don't cover, a Suggested keywords card appears in the campaign's Settings tab with the proposed terms and the reasoning behind them. Add to campaign merges them into that campaign's search keywords (duplicates are skipped); Dismiss declines. Only one suggestion is pending per campaign at a time, and a new campaign won't see one until there's enough accept/reject history to learn from.
"Don't target" — exclusions
The targeting editor has a Don't target section: titles, companies, industries, keywords, and locations you never want to see. Exclusions are enforced, not suggested:
- matching candidates are dropped before scoring — they never cost you a discovery credit, and each run reports how many were filtered;
- search queries are built to avoid excluded directions;
- anything that slips through is hard-capped by the scorer and marked as a poor fit with the matched exclusion named.
Which LinkedIn account a campaign uses
A campaign's Settings tab shows the connected account it sends from, and lets you point it at a different one. Sequences already running stay on the account they started with, so switching never re-attributes outreach that is already in flight.
Several campaigns
Nothing is shared between campaigns except your connected LinkedIn accounts and your default voice. Each campaign owns its targeting, its angle, and its languages — so a campaign for founders in Poland and one for ops leads in Germany can run side by side without either one blurring the other.