What is LinkedIn-sourced pipeline?

What is LinkedIn-sourced pipeline?

LinkedIn-sourced pipeline is the total value of open opportunities whose first qualifying touch happened on LinkedIn. The first touch might be a comment exchange, a connection that became a conversation, a profile view that led to a reply, or an inbound message sent after a rep engaged publicly.

The word sourced is doing a lot of work. It claims origin, which is to say that the opportunity would not exist without that first touch. That is a strong claim, and it is the reason the number is argued about more than almost any other in B2B reporting.

The alternative claim, influenced, is weaker and much easier to defend. It says only that a LinkedIn touch occurred before the opportunity opened. For most teams that is the number worth building, and the rest of this page explains why.

Sourced and influenced are different claims

Sourced attributes origin to one channel. Influenced attributes contribution and allows several. Choosing between them before you report anything prevents most of the arguments that follow.

Sourced is single-channel by definition. Only one touch can be first, so every sourced number implicitly takes credit away from another team. That is why the sourced conversation becomes political in any company with more than one demand channel, and why the debate rarely resolves on evidence.

Influenced is additive. An opportunity can be influenced by an event, a webinar and LinkedIn engagement simultaneously, and that is usually a truer description of how a considered B2B purchase actually came together. Gartner B2B buying research has consistently found buying groups spending only a small share of the purchase cycle with supplier representatives at all, split across every vendor on the shortlist, which is a portrait of a fragmented journey rather than a single decisive touch.

The recommendation is simple. Report influenced as the headline number, keep sourced for the subset where the first touch is genuinely documented, and never present sourced as though it were the larger figure.

Why the CRM source field will mislead you

The standard source field records the last conversion event, not the first contact, which systematically overstates paid and search while hiding social. This is the single most important thing to understand before reporting on this.

The typical path looks like this. A buyer sees a rep comment usefully under three posts over six weeks. They look at the company. Weeks later they have a need, search the brand name directly, land on the site and book a demo. The CRM records direct or organic search, because that is what the form captured.

Every step of that is working as designed. The form knows how the visit arrived, not why the person came looking. The LinkedIn touches were causally decisive and are invisible to the field that assigns the label.

There is a second distortion working in the same direction. Channels that end in a click are recorded automatically and completely, while channels that end in a human conversation are recorded only if somebody remembers. Paid advertising is therefore measured with near-perfect fidelity and social engagement with almost none, which means the reporting gap between them is a data-capture artefact rather than a performance difference. Teams reallocate budget on that artefact every year.

The practical consequence is that any audit of where deals came from must not start from the source field. Ask the reps who closed them, and expect disagreement with the CRM on roughly a third. Do not exclude referrals either, because a referral from somebody who has watched your team post for a year is LinkedIn-originated wearing a different label.

What has to exist before you can measure anything

Nothing is attributable unless engagement is written back to CRM contact records automatically. This is a data-capture problem before it is a reporting problem, and no dashboard recovers data that was never collected.

Manual logging does not work at the scale this motion runs at. A rep engaging five to eight buyers a day generates around thirty records a week, and Salesforce State of Sales research has repeatedly found representatives already spending a minority of their week actually selling. Any process adding administration to that load stops being followed within about a month, usually silently.

So the write-back has to be automatic, it has to match the engaged profile to a CRM contact, and it has to land as a dated activity on that contact record. Once it does, no special reporting layer is needed: normal pipeline reporting can filter on whether a contact had prior engagement, which is why influenced becomes easy the moment capture exists.

Build this before the first comment goes out. Retrofitting attribution onto a quarter of past engagement is theoretically possible and, in practice, nobody does it. The full sequence, with what to measure at each stage, is in the LinkedIn engagement-to-pipeline playbook for B2B GTM teams.

The four ways this number gets inflated

Every inflated attribution number this side of fraud comes from one of four ordinary mistakes, and all four are easy to make honestly. Checking for them before you present is cheaper than being caught by them afterwards.

The first is counting passive impressions as touches. A buyer scrolling past a post is not an interaction, and including view counts turns the number into reach wearing a pipeline label. Require a two-way action: a comment, a reply, a reaction, a profile view that led somewhere.

The second is an unbounded lookback window. If any touch ever counts, then a reaction from two years ago attributes a deal closed last month. Pick a window, ninety days is defensible for most sales cycles, and apply it consistently rather than choosing per deal.

The third is counting the wrong person. Engaging a company VP of Marketing does not warm its Head of Demand Generation, and attributing an opportunity because somebody at that account was touched inflates the number considerably in enterprise deals with large buying groups. Match on the contact, not the account.

The fourth is quietly switching between sourced and influenced depending on which is larger this quarter. That is the one that destroys trust permanently, because it is indistinguishable from the honest version until someone checks two quarters side by side.

How to define the number so it survives a finance review

Agree the definition with finance before you report it, not after they question it. A number defined in advance is a disagreement about strategy. The same number defined afterwards is a disagreement about your credibility.

Fix four things in writing. The qualifying touch, meaning which interactions count, since a passive impression should not and a two-way exchange should. The window, meaning how long before opportunity creation a touch still counts, where ninety days is a common and defensible choice. The unit, meaning opportunity value at creation rather than a moving forecast number. And the treatment of multi-touch deals, which for influenced reporting means the opportunity appears in every channel that touched it, with the overlap stated openly.

That last point is where credibility is won or lost. If your influenced numbers across channels sum to more than total pipeline, say so explicitly and explain why, in the same slide. An overlap acknowledged is a methodology. An overlap discovered by the CFO is a problem.

Write those four lines down once and reuse them verbatim every quarter. Redefining the metric between reports, even for good reasons, makes the series uncomparable and hands any sceptic an easy objection. If the definition genuinely has to change, restate the prior quarter under the new definition alongside the new one, so the trend stays readable.

Report the cost alongside it from the first quarter. Fifteen minutes per rep per day is a real number and a small one, and volunteering it makes the return credible in a way that presenting the return alone never does.

Who should own the number

Attribution reported by the team being measured is discounted by everyone who reads it, so the number should be produced by RevOps or finance rather than by the programme owner. This is not about honesty. It is about how the number is received.

The programme owner should define the methodology, argue for it, and be accountable to it. They should not be the person who runs the query, because a number that is both produced and benefited from by the same team invites exactly the scrutiny you were trying to avoid.

Where no RevOps function exists, the workaround is to hand over the definition and let somebody outside the programme run the report once a quarter. It takes an hour and it changes how the result is heard.

Publish the methodology alongside the number, every time, even when nobody asks. Four lines covering the qualifying touch, the window, the unit and the multi-touch treatment. A number with its method attached invites questions about the method. A number without one invites questions about the person.

What good looks like after a quarter

Expect the first two months to look worse than reality, because engagement precedes opportunity creation by weeks. Knowing the shape of the curve in advance stops teams killing the programme during the lag.

Coverage moves first, within two weeks. It is the share of your named buyers that anyone on the team has engaged in the last ninety days, and it exposes a programme that is busy without being aimed at anyone specific.

Reciprocity moves next, usually around week six. It counts how many of those buyers engaged back, and it is the best available proxy for familiarity.

Influenced pipeline moves last, in the first full quarter, and it is the number the executive team actually asked for. There is a reason it takes that long, and it is not the tooling. GTM Brigade runs the State of LinkedIn, built from 56,845 unique posts across 11,020 active creators on a rolling 60-day window, where the top 1% of posts capture 40% of all engagement, with a Gini coefficient of 0.841. Attention is concentrated enough that presence has to be accumulated deliberately, and accumulation takes time.

Report all three, in that order, every month. A single number reported late looks like a claim. Three numbers reported in sequence look like a programme, and they let a reviewer see progress before the revenue arrives. The wider practice this measures is described in what is social selling in 2026.

Last updated: September 2026