Social Selling: From Reaction to List

What social selling actually measures

When reach and reactions grow while pipeline stays flat, the content budget goes back up for debate every quarter. The cause is almost never the post. It is that nobody owns the decision about what happens to the people who reacted. At that point social selling is list work, and list work rarely gets assigned to anyone.

Our position: reach is the raw material, the list is the product. Optimising reach without building the list produces attention and hands it to nobody.

How far reach and result can sit apart shows up more clearly in one of our own campaigns than in any benchmark.

The second number is not a reward for the first. It is the outcome of the work in between: the people who reacted became a list a human signed off. Everything that follows describes that work.

Someone who reacted is not yet a lead

Every reaction looks like interest, which is why it lands in the list unfiltered. The cost is paid by whoever works that list: they approach people who can never become customers, and they lose trust in the list itself. From then on every further row gets read with suspicion.

Our position: anyone who reacts is checked against the target picture on two levels before entering the list — person level and company level. The right person at the wrong company is not a lead, and the right company with the wrong role at the table is not one either.

In practice: the LinkedIn export is only the intake. Every row gets a decision, and anyone who does not fit is listed with a reason. That sounds like bookkeeping and it is the part that pays off. Without a recorded reason the next pass checks the same person again, and by the third pass nobody checks properly at all.

The message that must never be automated

An automated first message can land in the middle of a conversation that is already running. The person spoke to your founder three months ago, met them at a conference, or got a personal note. To them the sequence reads as if nobody had been listening, and that costs more than a message you never sent.

Our position: you may only automate outreach to people you newly won through your own connection request. Before that request the two of you were not connected, so no earlier conversation can exist.

In the list this becomes its own field: every row carries a send mode. Contacts from your own connection campaign may enter the sequence, for example in Lemlist. Anyone who was already connected goes into a weekly batch the sender reviews and releases person by person. That costs half an hour and replaces the apology otherwise due three months later.

Empty means never checked, not clean

An empty field looks like an all-clear. When matching against suppression and customer lists it is the opposite: a pre-computed column stays empty when the computation never ran, and it does not recompute when the suppression list has grown since. Reading it means mistaking an unasked question for an answer.

Our position: the suppression list is pulled fresh from the CRM on every pass and matched in code. No stored column, no intermediate state, no exception.

The comparison rule is the real decision here, and it usually gets made in passing. Three rules that make the difference in practice:

Each of these three rules can be changed in one line, and each shifts who receives a message. That is why it belongs written down, in the same place as the reason.

Enriched fields age, your target picture does not

Job title, company and location usually sit in the list exactly as they were on the day of enrichment. The target picture behind them is current; the data is not. You find out in the conversation: the role no longer exists, the person moved, the company now belongs to someone else.

Our position: verification happens immediately before handover. Enrichment tools such as Clay keep the value they found on the first pass; they do not refresh it on their own, and they are not built to.

In practice: before a batch reaches the sender, the job title, company and location of everyone in it get checked live. Anything unverified does not travel. That single rule costs time and protects the scarcest resource in the whole process — the attention of the person who releases the batch. Show them three stale rows and the fourth review is no longer a decision, it is a nod.

Finished people leave the list

When the same names reappear every week, sooner or later nobody reads closely. A list without an exit grows, and with it the time each review costs, until the review is skipped and the list goes out unchecked.

Our position: an active list only carries people something is still meant to happen with. Everyone else moves out, with outcome, reason and source.

Three cases leave the list: spoken to, not relevant, deliberately not pursued. Each takes its reason along so the decision stays traceable and does not have to be made twice. Anyone who should still get a follow-up stays, and that is the only criterion.

How you know it is working

Reach metrics go up almost regardless, including when nothing is happening. They never answer the pipeline question. The question that counts is simpler and less comfortable: do deals with LinkedIn in them close differently from deals without?

Our position: measure the deal, not the post. The number that decides whether social selling survives sits in the CRM, not in the LinkedIn analytics tab.

The bottom two rows come from our own pipeline across twelve months. They are why we consider the list work justified: the touchpoint has to be documented in the CRM to count at all, and it only gets documented when the message came from a list somebody owned.