Should you split LinkedIn outreach by persona
or run one broad campaign?

When should you split LinkedIn outreach by persona?

Split the campaign when role changes the reason someone should care. In practice, that usually means the buyer sees a different problem, owns a different metric, or needs different proof before replying.

A founder, a sales leader, and an operations leader may all sit inside the same account list, but they do not read the same message the same way. Founders tend to care about speed, risk, and whether the offer is worth attention at all. Sales leaders care about pipeline quality, rep usage, and whether your process creates extra admin. Ops leaders care about workflow friction, handoff quality, and whether the motion is controllable.

If you send one broad message to all three, you normally remove the detail that makes any of them respond. That is the real cost of broad campaigns. They look efficient in setup, but they often force generic copy.

• Split by persona if the problem statement changes by role
• Split by persona if the proof point that builds trust changes by role
• Split by persona if the call to action should change by role
• Keep one campaign only if the message still reads as specific to every prospect in it

Why do broad LinkedIn campaigns usually underperform?

Because LinkedIn punishes vague relevance twice. First on the connection request, then in the DM after acceptance.

The connection request is a relevance filter. People accept when your profile and implied context feel close enough to their world. Then the DM is another filter. If the message sounds like it could have gone to anyone, reply intent drops.

On one white label programme across advisor workspaces, we saw 59% connection request acceptance and about 9% LinkedIn DM reply rate, versus about 1.5% email reply rate on the same accounts in the same window. I would not misuse that as a promise for your campaign. I use it to make one narrower point. When LinkedIn works, it works because the context is close to the buyer. Persona splits are one of the easiest ways to keep that context intact.

A broad campaign also hides diagnosis. If acceptance falls, you do not know whether the issue is role fit, account fit, the profile, or the message. If replies fall after decent acceptance, you do not know which persona rejected the framing. Segmentation makes failure easier to read, and that matters more than people admit.

How many personas are actually useful?

Usually fewer than teams expect. Most outreach programs do not fail because they used too few personas. They fail because they invented too many weak distinctions and could not maintain message quality.

My default is to split only when the outbound logic changes. Not just the wording, the logic. If two personas can receive the same opening, same proof, same offer framing, and same next step with only light edits, keep them together. If you need a different angle to be credible, split them.

The mistake is treating job titles as personas by default. Title alone is not enough. The split should follow decision context, not just org chart labels.

What should change between persona campaigns?

Not everything. If you rewrite every line for every segment, execution gets messy and reporting becomes noise. Change only the parts that carry relevance.

• The opener, if each persona notices a different trigger
• The problem framing, if the cost of the problem is role specific
• The proof, if one persona trusts operator detail and another trusts commercial outcomes
• The CTA, if one role will take a call and another prefers a lighter next step

Leave the rest stable where possible. That gives you cleaner reads on what matters. It also keeps your operation maintainable when campaigns scale across multiple accounts.

This is the same discipline we apply to DM writing. Specific enough to feel intentional, simple enough to execute consistently. If your team keeps changing offer, opener, proof, and CTA all at once, you are not testing. You are just rewriting.

Can persona splits improve benchmark performance?

They can, but only if the underlying list is already sane. Segmentation does not rescue a bad market, a weak offer, or poor profile credibility.

As a working benchmark, 0.5 to 1% positive on sends is workable, 1% and above is strong, and under 0.5% should usually be killed. Persona splitting helps you move toward the stronger end when role relevance is the main blocker. It does not magically create demand.

A useful example is the follower sourced segment we tracked at 52,786 sends, with 0.14% positive, despite being 2.85x the fleet baseline. That is still not a segment I would celebrate. It simply shows that a source can outperform the rest of a poor fleet and still be too weak to keep. The lesson for persona work is obvious. Relative improvement is not enough. You still need absolute performance that clears the bar.

So if you split by persona and one segment improves from bad to slightly less bad, do not keep it alive for emotional reasons. Kill weak segments early.

When is one broad LinkedIn campaign the better choice?

When segmentation overhead will cost more than the extra relevance is likely to return.

That is common in narrow niches where multiple titles behave similarly, in very small total addressable markets, or when you are still validating whether anyone cares about the core proposition. Early on, one broad campaign can be the fastest way to learn if the market is alive at all.

Broad also makes sense when your differentiation is simple and the buying trigger is shared. For example, if every prospect is reacting to the same operational event and the same downstream pain, a broader campaign can stay crisp.

But be honest here. Many teams say the pain is shared when what they really mean is that they do not want to build multiple campaigns. Operational laziness is not segmentation strategy.

Where does this advice fail?

It fails when the sample is too small to read, when the offer is still changing every week, or when your list quality is so inconsistent that no campaign structure can produce a clean signal.

It also fails if the operator running outreach cannot maintain the extra complexity. Two clean persona campaigns beat six half maintained ones every time.

And this advice is not for teams trying to solve cross channel sequencing inside one framework. That belongs elsewhere. If you are deciding how LinkedIn should interact with email or other channels, handle the sequencing logic separately and do not force that question into this campaign design decision.

Who should not follow this too aggressively? Early founders with tiny TAM, agencies with unstable client positioning, and teams with no discipline in reporting. If that is you, keep segmentation lighter, get to signal faster, then split once a pattern is real.

What is the practical way to decide?

Use a simple operator test. Ask four questions before you launch.

• Would each persona describe the problem differently in their own words?
• Would each persona need different proof before taking the next step?
• Would the same CTA feel natural to every persona in the segment?
• If one segment underperforms, do I need to know that separately to make the right next decision?

If the answer is yes to most of those, split the campaign. If the answer is no, keep it broad for now.

That is the operator version of segmentation. No brand workshop language. No fantasy buyer avatars. Just whether role differences are material enough to change outreach decisions.