i’ve been wrestling with this for a few weeks—how tight should my ideal customer profile actually be when i’m filtering prospects in liseller?
right now i’m doing what feels like pretty thorough filtering: company size (50-200 employees), industry (saas), revenue stage (somewhere between series a and c), plus a few other dimensions. and i’m getting pretty decent hit rates.
but i keep wondering if i’m actually being too broad, or if i should go the other direction and get even more specific. like, should i be filtering down to:
- specific job titles (only sales directors, not vps)
- specific department maturity level (only companies with a dedicated revenue operations team)
- specific compliance requirements they’re probably dealing with
- recent hiring patterns that signal they’re scaling sales
the tradeoff is obvious: tighter filters = fewer prospects, but theoretically higher conversion rates because they’re better fits. looser filters = more prospects to message, but lower conversion rates because some aren’t really a fit.
my real question is: at what point does filtering stop helping and start hurting? like, is there a filtering complexity ceiling where you’re just removing good prospects to artificially inflate your icp definition?
i’ve got enough outreach volume that i could run both—send to a broad-filtered list and a narrow-filtered list and measure conversion rate. but before i do that a/b test, i want to hear if anyone else has experienced a dropoff with over-filtering. did you get more conversions from 100 super-tight-fit prospects, or 500 decent-fit prospects with a lower conversion rate?
okay, so filtering is math. let’s say tight filter = 100 prospects at 5% reply rate. loose filter = 500 prospects at 2% reply rate.
tight: 5 replies
loose: 10 replies
loose wins on volume, but are those 10 replies the same quality as the 5 replies from the tight filter? probably not. some of those 10 are tire-kickers who aren’t a real fit. so your conversion to paid deal might be 20% from the tight filter (1 deal from 5 replies) vs. 10% from the loose filter (1 deal from 10 replies).
so the tighter filter actually wins on deal quality, even though it loses on volume.
the real answer: tight filter + excellent messaging beats loose filter + mediocre messaging every single time. so if you’ve nailed your messaging, tighten your filters. if your messaging is still weak, loosen your filters and get more volume to test with.
if you’re doing this at scale, you need to build the test infrastructure anyway. send 500 messages to broad-filtered list, 500 to narrow-filtered list, and track:
- reply rate
- qualified conversation rate (replies that turn into real discussion)
- meeting booked rate
- deal closed rate
then work backwards from your icp. which filtering approach generates better $/hour of your time?
like, if tight filtering takes 2 hours setup but generates 10 qualified conversations and 3 meetings, vs. loose filtering taking 30 min but generating 15 qualified conversations and 2 meetings—tight filtering wins because you’re closing more.
also integrate this with your crm so you can track which attribute (company size, revenue stage, title, etc.) actually correlates with deal closure. some attributes matter for response, others matter for conversion. they’re different.
from recruiting, i learned that over-filtering is absolutely possible. i used to filter for “directors at companies funded in the last 2 years.” that seemed super specific and valuable. but i was missing solid candidates who were at older, stable companies. dropped my applicant diversity way down.
when i loosened it to “any manager+ at growth-stage companies,” suddenly i got way more engagement because i wasn’t artificially excluding good prospects.
so the answer to your question: yes, over-filtering hurts. but it’s usually not because the filter itself is wrong—it’s because you’re using it as a crutch instead of relying on your message to do the heavy lifting.
if your message is good, you don’t need to filter as much. if your message is weak, no amount of filtering will save you.
my rule of thumb: use filters to eliminate disqualifiers (someone who obviously isn’t a fit), not to chase the perfect fit. if you’re a b2b saas tool, filter out freelancers and students. everything else is fair game. let your message do the qualification.
one more thing: aggressive filtering can actually hurt your account health. every filter operation you run consumes account juice. if you’re running super complex filters (7-8 dimensions), you’re putting more load on your account, and linkedin can flag that as unusual activity.
i’d recommend keeping filters simple and broad. company size + industry + basic seniority. then add behavioral filters (recent activity, profile changes) rather than static demographic filters. behavioral filters are cheaper for your account and more predictive anyway.
i’ve tested this exact thing. went super tight on filters (company size, revenue stage, hiring in sales last 6 months, and no competitors). ended up with 200 prospects. conversion rate was 4.5%.
then i loosened it to just company size + industry. 800 prospects. conversion rate was 2.2%.
200 * 0.045 = 9 conversations
800 * 0.022 = 17.6 conversations
loose filter won the conversation game. but when i tracked down the 9 from the tight filter, like 5 turned into actual sales conversations. from the loose filter, only 2 of the 17 turned into real conversations.
so the tight filter won the quality game, the loose filter won the volume game. i’ve since gone hybrid—loose filter to get volume, then secondary qualification in the follow-up sequence. filter hard on messaging, not on the prospect list itself.
biggest learnings: don’t over-filter on single attributes (like title). people have weird titles. do filter on behavior (recent changes, activity) because those are more predictive of actual buying intent.
in liseller, the filtering sweet spot is usually 3-4 dimensions. company size, industry, maybe job level, maybe recent activity. beyond that, you’re not really gaining signal—you’re just reducing your list.
what works best: use our smart lead filtering (which uses behavioral signals) rather than just demographic filters. behavioral signals like “profile updated,” “job change,” “company news” are way more predictive of engagement than static attributes like company size.
have you explored the behavioral filtering options in liseller, or are you mostly using demographic filters?
strategically, the right icp isn’t the one with the fewest prospects, it’s the one with the highest lifetime value per conversion. so you need to measure deal value, not just conversion rate.
if tight-filtered prospects close at double the deal value, then tight filtering wins even at lower volume. if loose-filtered prospects close at the same value, then loose filtering wins because you get more deal volume.
most teams never measure this. they just track “did they reply?” and miss the bigger picture. track all the way to deal closed, and your icp will sort itself out.
also, icp should change over time. your first 100 customers will teach you what actually looks like a good customer. use that data to inform your second generation of icp. don’t lock your thinking into an icp on day one.