Last updated: November 19, 2026
The LinkedIn signals worth building around are the ones closest to a named person: someone viewing a rep’s profile, engaging with the team’s posts or a competitor’s, following the company page, or changing jobs. Company-level events such as hiring for a relevant role or announcing funding come next. Capture each signal with its source and timestamp, score it on fit, strength, relevance, recency and confidence, and route only the high scores to a rep the same day. Everything else feeds nurture or gets dropped.
This guide is written for GTM engineers and RevOps leads at B2B SaaS companies who are asked to turn LinkedIn activity into pipeline. It covers which signals to track, four ways to capture them, the fields to store, a scoring model you can build in the CRM, and the routing rules that decide whether a signal turns into a conversation or into noise.
Disclosure: This article was written with input from Valley, a signal-based outbound platform that is mentioned below as one capture option. The GTM Engineering Company builds signal, scoring and routing systems for B2B SaaS teams and does not resell Valley.
Key Takeaways (TL;DR)
Rank signals by distance from the buyer: person-level signals (profile views, post engagement, job changes) beat company-level ones (hiring, funding), because they tell you who as well as when.
Decay matters as much as the signal: engagement goes cold in days, job changes and funding in weeks to months. Every signal needs a timestamp.
Store every signal in one shape: type, source, date, person, account, evidence URL and a 30-day count. Scoring, routing and reporting all depend on it.
Score on five factors, 0 to 3 each: ICP fit, signal strength, contact relevance, recency and confidence. Fit is a gate, not just a factor.
Route fast, and only the top band: 12 to 15 goes to the account owner the same business day; 8 to 11 enters a sequence; the rest is nurture or nothing.
Why LinkedIn Signals Matter for GTM Engineers
Most B2B buyers do their research long before they talk to sales, and a lot of that research happens on LinkedIn: reading posts, checking who works at a vendor, looking at the profile of the rep who just emailed them. Those moments leave traces. A GTM engineer’s job is to turn the traces into a short, ranked list a rep can act on, instead of a firehose of notifications nobody reads.
The difference between signal-based outbound and cold outbound is timing. A cold sequence guesses when someone might care. A signal tells you that something just happened, and often who it happened to. Valley, for example, reports 15 to 45% reply rates on outreach triggered by these signals, against about 2% for cold outreach (vendor-reported). Your own numbers will differ, but the direction is consistent across teams: the closer the signal is to a named person and to today, the better the reply rate.
The catch is volume. A team with five reps posting regularly can generate hundreds of profile views and post reactions a week. Without capture, scoring and routing rules, those signals either sit unread in LinkedIn or get dumped into a sequence that treats a student and a VP the same way. The rest of this guide is the system that sits in between.
The LinkedIn Signals, Ranked by Distance from the Buyer
Signals closer to a named person convert better because the timing and the person are both known. Company-level signals tell you when, but not who.
Signal | Level | What it suggests | Typical decay |
|---|---|---|---|
Viewed a rep’s profile | Person | Curiosity about the seller, often after an ad, post or email | Days |
Engaged with the team’s posts | Person | Interest in the topic the team sells into | Days |
Engaged with a competitor’s posts | Person | Evaluating the category | Days to weeks |
Followed the company page | Person | Ongoing interest, lower intent than a view | Weeks |
Changed jobs (past champion or user) | Person | New budget, new stack, a known relationship | Weeks to months |
Hiring for a role tied to the problem | Company | A team being built or replaced around the problem | Weeks |
Announced funding | Company | New budget and pressure to grow | Weeks to months |
The decay column matters as much as the signal. A profile view from last week is warm; one from last quarter is noise.
Person-level signals
Profile views and post engagement are the warmest signals because the person chose to look at you or your topic. Competitor post engagement is slightly cooler but often more valuable for qualification: someone commenting on a competitor’s launch post is probably evaluating the category right now. Company page follows are the weakest of the person-level group, since people follow pages for many reasons, including job hunting.
Job changes deserve their own treatment. When a past champion or power user moves to a new company, you have a warm relationship, a buyer who already knows the product and a new team that may not have it yet. These signals last longer than engagement, often weeks to months, because new leaders usually review their stack in their first quarter.
Company-level signals
Hiring and funding are useful for timing and for account prioritization, but they do not tell you who to contact. A company hiring three SDRs probably needs outbound tooling; it does not tell you whether to message the VP of Sales, the new SDR manager or RevOps. Treat company-level signals as a reason to look harder at an account, then use person-level signals or role mapping to pick the contact.
For a broader view of these and other intent sources, Valley’s guide to LinkedIn buying signals walks through each signal type and how teams act on it.
How to Capture LinkedIn Signals
There are four ways to get these signals into a system, with different trade-offs.
Option | What it captures | Trade-off |
|---|---|---|
Manual review in LinkedIn | Profile viewers (with Premium, subject to the viewer’s privacy settings), post engagement, page followers | Free or cheap, but depends on someone checking daily; does not scale past a few reps |
Sales Navigator saved leads and accounts | Alerts on saved leads (for example, job changes) and company alerts such as hiring growth | Reliable for known accounts; only covers what has been saved |
Enrichment and workflow tools (for example, Clay) | Company-level events such as funding and hiring, joined to CRM records | Strong for company signals; person-level LinkedIn engagement usually needs another source |
Signal-based outbound platforms | Person-level LinkedIn engagement plus company events, qualified and turned into outreach | Fastest to run; less control over the raw data than a custom build |
Platforms in the last row, such as Valley, watch LinkedIn buying signals like profile viewers, post engagers, competitor post engagers and job changes, qualify the people behind them against the ICP and draft the outreach.
Choosing a capture method by team size
One founder or one or two reps: manual review once a day is enough. Copy the viewers and engagers who fit the ICP into the CRM with a signal type and date, and keep the habit for a month before buying anything.
Three to ten reps: manual review stops working because nobody checks consistently. Combine Sales Navigator alerts for named accounts, an enrichment tool for hiring and funding, and a platform for person-level engagement.
Larger teams with RevOps support: build the data model and scoring in the CRM and treat every capture source, bought or built, as an input that writes to the same fields.
Whichever route you pick, the capture layer should never write straight into a sequence. It should write to the CRM, where scoring and routing decide what happens next.
The Data Model: Fields Worth Storing per Signal
Whatever the capture method, write each signal to the CRM in the same shape. It makes scoring, routing and reporting possible later.
Field | Example | Why it matters |
|---|---|---|
signal_type | profile_view, post_engagement, competitor_engagement, job_change, hiring, funding | Drives scoring weights and routing rules |
signal_source | LinkedIn, Sales Navigator, enrichment tool, platform | Lets confidence be scored by source |
signal_date | 2026-11-10 | Recency decay needs a timestamp |
person / account | CRM contact and company IDs | Joins the signal to fit data |
evidence | URL of the post, job listing or announcement | Gives the rep something specific to reference |
signal_count_30d | 3 | Stacking repeated signals without double counting |
One rule saves a lot of noise: repeated events from the same source are not independent confirmation. Three views of one profile in a day is one signal, not three.
Store signals as their own records (a custom object in Salesforce or HubSpot, or a related table) rather than overwriting a single “last signal” field on the contact. History is what lets you see that an account has gone from one engaged person to four in two weeks, which is often a stronger sign than any single event. It also lets you report on which signal types actually lead to meetings, which you need before you can tune the weights. If the underlying records are messy, fix that first; CRM data hygiene covers the duplicates and missing fields that break signal matching.
A Scoring Model to Build in the CRM
Score each person on five factors, each from 0 to 3:
ICP fit. Does the company and role match the ideal customer profile? No fit means no outreach, whatever the signal.
Signal strength. Person-level engagement and job changes score higher than page follows.
Contact relevance. Is this person the buyer, a user, or someone with no stake?
Recency. Apply the decay windows from the first table.
Confidence. Is the signal observed directly, or inferred by a vendor?
Then route by total:
Score | Action |
|---|---|
12-15 | Rep reviews and reaches out within one business day |
8-11 | Enters a timed sequence within the week |
4-7 | Nurture: content, retargeting, a later check |
0-3 | No action |
Fit acts as a gate before the total: a perfect signal from a company outside the ICP should still score zero for outreach.
A worked example
A VP of Sales at a 120-person B2B SaaS company views your AE’s profile the day after a demo request from a colleague. ICP fit is 3 (right size, industry and role), signal strength is 3 (a direct profile view), contact relevance is 3 (the likely buyer), recency is 3 (yesterday) and confidence is 3 (observed directly on LinkedIn). The total is 15, so the account owner gets an alert today.
Compare a marketing coordinator at the same company who liked a competitor’s post three weeks ago. Fit is still 3 at the company level, but signal strength is 2, relevance is 1, recency is 1 and confidence is 3, for a total of 10. That person enters a slower sequence, or simply adds weight to the account while the rep focuses on the VP.
Scoring accounts, not just people
B2B deals close through buying committees, so roll person scores up to the account. A simple approach is to take the highest person score plus one point for each additional engaged person in the last 30 days, capped at 15. Three people from one account engaging in a week is a stronger signal than one person engaging three times. For more on combining fit and intent into one score, see our comparison of the best lead scoring software, and if you want to build the score yourself, the Clay lead scoring build walks through the formula step by step. The fit half depends on a clear ICP; if yours is still a guess, define your ICP with data first.
Routing: Getting the Signal to the Right Rep Fast
Owner first. If the account already has an owner in the CRM, the signal goes to that rep, not round-robin.
Same-day alerts for the top band. Post the signal, the evidence URL and the score to the owner’s Slack channel or task queue.
Dedupe across reps. One person engaging with three reps’ posts should create one task, not three.
Write back the outcome. Replied, booked or ignored. Without outcomes, the weights never improve.
Speed is the point of the top band. A profile view decays in days, so an alert that waits for a weekly review has already lost most of its value. Set a service level, such as same business day for 12 to 15, and report on it the way you would report on inbound response time. For unowned accounts, use your normal assignment rules by territory or segment; lead routing software compares the tools that handle this inside HubSpot and Salesforce.
What the alert should contain
A good alert lets a rep decide in under a minute. Include the person’s name, title and company, the signal type and date, the score with a one-line reason, the evidence link, and any open opportunity or recent activity on the account. Leave out anything the rep cannot act on. If reps start ignoring alerts, the threshold is too low or the alert is too noisy; raise the bar before adding more signals.
Guardrails
Respect platform limits. LinkedIn caps how many invitations and messages an account can send. Signal volume should never push a rep past those limits.
Privacy settings are real gaps. Some viewers browse privately and will never appear by name. Do not try to work around that.
Reference the signal carefully. Mentioning that someone viewed a profile can feel invasive. Lead with the problem the signal suggests, not the surveillance.
Keep the evidence. If a message references a hire or a post, the rep should be able to open the source in one click.
A practical test for the third guardrail: if the message would read the same without the signal, the signal was only used for timing, which is fine. If the first line is “I saw you viewed my profile”, rewrite it. A message that opens with the problem the person is likely working on, informed by the post they engaged with or the role their company is hiring for, gets the benefit of the signal without making the buyer feel watched.
Build or Buy
GTM engineers can build most of this in the CRM plus an enrichment tool, especially for company-level signals. Person-level LinkedIn engagement is the hardest part to capture reliably at scale, and it is where buying a platform such as Valley tends to save the most time. A common split: build the data model, scoring and routing in the CRM, and buy the LinkedIn capture and first-draft outreach.
Layer | Build | Buy |
|---|---|---|
Data model and fields | Usually build: it has to match your CRM | Rarely worth buying |
Company signals (hiring, funding) | Easy with an enrichment tool | Included in most signal platforms |
Person-level LinkedIn engagement | Hard to capture for a whole team | Where a platform saves the most time |
Scoring and routing | Build in the CRM so RevOps owns it | Platform scoring works if it writes back to the CRM |
First-draft outreach | Possible with AI steps in a workflow tool | Fastest with a platform that drafts from the signal |
Whatever you buy, check that it writes signals and outcomes back to your CRM in the shape described above. A tool that keeps its data in its own dashboard makes it impossible to score across sources or measure which signals pay off.
A 30-Day Rollout Plan
Week 1: data model. Create the signal object and fields, and agree the ICP fit criteria with sales.
Week 2: capture. Connect one person-level source and one company-level source. Do not add more until both write clean records.
Week 3: scoring and routing. Turn on the five-factor score, the four bands and same-day alerts for the top band, for one team first.
Week 4: review. Check reply and meeting rates by signal type and score band, then adjust the weights. Expand to the rest of the team only when the top band is producing meetings.
FAQs About LinkedIn Signals
What are LinkedIn buying signals?
LinkedIn buying signals are observable actions on LinkedIn that suggest a person or company is closer to buying: profile views, engagement with your posts or a competitor’s, company page follows, job changes, hiring for relevant roles and funding announcements. Person-level signals tell you who is interested; company-level signals tell you when a team may have budget.
Which LinkedIn signal converts best?
Person-level signals tied to your own content or profile usually convert best, such as a profile view or post engagement from someone who fits the ICP, because both the timing and the person are known. Job changes of past champions come close, since the relationship already exists. Company page follows and funding news convert less well on their own.
Can LinkedIn profile viewers be tracked automatically?
LinkedIn shows profile viewer insights to Premium subscribers, subject to each viewer’s privacy settings, so viewers who browse privately never appear by name. Turning the visible viewers into a scored, routed list needs either a daily manual export by each rep or a tool connected to the account.
How long does a LinkedIn signal stay useful?
Engagement signals such as profile views and post reactions decay in days. Competitor engagement lasts days to weeks, and job changes and funding stay relevant for weeks to months. Store a timestamp with every signal and reduce its score as it ages, so last quarter’s view does not outrank yesterday’s.
Should LinkedIn signals go straight into a sequence?
Only the top scoring band, and only after an ICP fit check. Mid scores belong in a slower sequence, lower scores in nurture, and anything outside the ICP should not be contacted at all, however strong the signal looks.
Do GTM engineers need a tool to capture LinkedIn signals?
Not for company-level signals: hiring and funding can be pulled into the CRM with an enrichment tool such as Clay. Person-level engagement, such as who viewed a profile or reacted to a competitor’s post, is the hardest part to capture reliably for a whole team, and that is where a signal-based platform tends to save the most time.
About the Author
Jorge Macias is the founder of The GTM Engineering Company, which builds signal, scoring and routing systems for B2B SaaS teams inside their own CRM. The company’s site presents him as a Y Combinator S18 alumnus who has built go-to-market systems for 50+ B2B companies.




