LinkedIn job titles: exact match or seniority plus function for a finance software ad?

I'm running a LinkedIn campaign for a B2B software company that sells to finance teams, and I keep going back and forth on targeting. Exact job titles like "Financial Controller" look precise, but LinkedIn's title field is messy. People write "Controller, Finance Ops", "FC" or "Head of Finance", so an exact match misses a lot of the people who actually sign off on purchases. Seniority plus function reaches more of them, but you end up paying for impressions on people with no budget authority. My plan is to start broad on seniority and function, then let the first month of lead quality decide whether to narrow. I'd judge it by the sales team's opportunity rate rather than cost per lead, because a cheap lead from the wrong title is the expensive mistake. Has anyone seen exact titles work at a smaller company where the title is the only signal you have?

Sequencing makes sense to me. Go broad on seniority and function, then let opportunity rate decide. Exact titles can shrink a LinkedIn audience faster than people expect once you add country and company size, so you probably won't have the volume to learn much from them anyway.

What I'd add is one qualifying question on the Lead Gen Form, something like "Do you take part in approving software purchases?" with a few answer options. Profile titles are a loose signal, but a self-reported answer tells you about budget authority directly, and it screens out some of the cheap leads before sales ever sees them. Then tag each lead in your CRM with both the title and the form answer. After the first month you can split opportunity rate by each one, so you're not guessing which actually drove the results.

Honestly, I'm not sure that form question tells you much. People who want the gated asset will often tick "yes, I approve purchases" whether it's true or not, so the answer shows you who's keen more than who actually holds budget. I'd still ask it, but compare its opportunity rate against the title data before you use it as a filter.

Agreed, the yes answers are inflated. I'd use the question as a negative filter only: anyone who says no is unlikely to sign, so drop them, then rank the yes group by title.

I wouldn't drop the no group yet. Plenty of people who don't sign off on purchases still run the evaluation. They shortlist vendors, sit in the demo and tell the approver what to buy, so someone who says "no, I don't approve software" could be the finance ops lead doing the real work. And if you already think the yes answers are inflated, the no answers may be the more honest signal. I'd keep that group as its own segment and compare its opportunity rate with everyone else's. If sales says those leads never move, cut them then. Until you have that split, dropping them is just another guess.

Agreed, cutting the no group on a guess would be a mistake. A small segment gives a noisy opportunity rate, so one month of sales feedback can mislead you in either direction. I'd keep any self-reported segment in its own bucket until it has roughly 30 leads, then cut it only if its opp rate is clearly below the rest, not just a few points behind. It also helps to have sales tag each opportunity as "evaluator" or "approver," since that shows whether the no group is doing the real work.

If the Lead Gen Form pre-fills job title from the member's profile, the title in your export is the profile value, not something they typed for this form. So title and the form answer aren't fully independent, and a cross-tab can look cleaner than it really is. Only the qualifying question is truly self-reported.

If you want the title field to tell you anything, check which fields are pre-filled before you build the CRM split, and see whether the title values actually vary across leads.

If the title is pre-filled, the split is mostly re-measuring the targeting you already chose, so it's less independent than it looks. You're not learning something new, you're confirming your own assumption. How often do people actually overwrite that pre-filled title before they submit?

I don't think "just re-measuring the targeting" covers it. The audience was chosen on seniority and function, not on title wording, so splitting by title text can still show whether "Controller" outperforms "Head of Finance" inside that same group. The bigger problem is that a pre-filled value may not match what people would have typed themselves, so check how much the titles vary before you trust the split.

Iris, fair correction. Splitting title wording inside a seniority and function audience can tell you something real. My worry is narrower: if almost every lead has the same pre-filled value, there's nothing to compare. So the rule I'd use is simple. If fewer than about a fifth of leads change the title field, treat title as a non-signal and look at the form question and sales tags instead. That way the title split can't mislead you by accident.