How to scrape beyond the 2,500-result limit with Split Search in Sales Navigator
Turn on Split Search so your Sales Navigator Leads Scraper automatically breaks a large search into facet sub-searches and collects leads beyond the 2,500-result limit.
Sales Navigator only ever shows 2,500 results per search — 100 pages of 25 profiles. No matter how many leads actually match your filters, you can't page past that ceiling. See scraping limits for how this fits with LinkedIn's daily search limits.
Split Search works around that cap. When a search has more than 2,500 results, it automatically breaks the search into smaller facet sub-searches — each one under 2,500 — and scrapes them all, so the crawler can reach leads it otherwise couldn't.
This setting is specific to the Sales Navigator Leads Scraper.

How to turn it on
- Open the batch's settings
Open your Sales Navigator Leads Scraper batch and click Settings. The option is in the Settings section; open Advanced settings at its bottom if you don't see it.
- Enable Split Search
Switch on the Split Search (beat the 2,500 cap) toggle. The change applies to the next launch.
Once it's on, splitting happens automatically — only when a search actually exceeds 2,500 results. Smaller searches run normally and are untouched.
How the split works
Split Search picks one filter dimension and divides your search along it. It does not combine dimensions or split recursively — it's a single split into one sub-search per value of the chosen dimension, plus one last sub-search for the leads that have none of those values (on LinkedIn, a lead can have no seniority at all).
It tries dimensions smallest-first, and uses the first one that (a) isn't already a filter in your search and (b) has enough distinct values to cover the result count:
| Order | Dimension | Distinct values |
|---|---|---|
| 1 | Seniority | 10 (In Training to Owner / Partner) |
| 2 | Industry | 20 (LinkedIn's top-level industries, each with its sub-industries) |
| 3 | Function | 26 |
The exact selection logic
- Estimate how many buckets are needed — roughly the total result count divided by ~2,000 (a margin below the 2,500 ceiling).
- Walk the order above and pick the first dimension that isn't already used as a filter in your search and has enough distinct values to cover what's needed.
- Split into one sub-search per value of that single dimension, then add one sub-search that excludes all of them.
Limits and edge cases
Split Search reaches most large searches, but it isn't unlimited:
- Works best up to ~50,000 results. Very large or heavily skewed searches may still miss some leads.
- No recursion. If an individual sub-search still exceeds 2,500 results, it is not split again. That bucket is paged up to the 2,500 cap and anything beyond it is missed. This is the main reason a very skewed search can leave leads uncollected; the run log tells you when it happens (see below).
- A few leads short. LinkedIn leaves a lead out of a page now and then (24 profiles instead of 25), so a sub-search can end a few leads under its total.
- Overlap. A lead can match several values (a CXO who is also an Owner / Partner) and come back in several sub-searches. It is kept once.
- Fallback dimension. If no single dimension has enough distinct values, it falls back to the largest unused dimension (Function) and uses it anyway.
- Nothing to split by. If all three dimensions are already used as filters in your query, Split Search does nothing — there's no unused facet left to divide on.
What Split Search applies to
Split Search only engages on a regular people-search URL (a query-based Sales Navigator search). It works by injecting a facet filter into the query itself, which it can't do to a saved reference or a static list — so it excludes:
- Saved searches
- Recent searches
- Shared searches
- Lead lists
For those, scraping is still capped at 2,500 results per search.
What you'll see
Splitting happens behind the scenes — you won't see extra tasks or extra runs. Every sub-search runs inside the same search task, and all the leads land under the same task and run.
When the split starts, the run log says:
🪓 Splitting search into N sub-searches to reach beyond the 2,500 cap.Three more lines show up only when the split cannot reach every lead:
⚠️ Sub-search 2/11 holds 3100 results: LinkedIn shows only the first 2,500 of them.
⚠️ Sub-search 2/11 ended with 150 of its 2383 leads: LinkedIn returned no more.
⚠️ Split Search could not reach every lead: ...The first one means a single value of the dimension (one seniority level, for example) still holds more than 2,500 leads, and LinkedIn shows only 2,500 of them. To collect them all, add a filter to your search (a region, a company headcount) and run the narrower searches as separate tasks.
The second one means LinkedIn stopped returning leads for that sub-search well before its total. Launch the run again later, and contact us if it keeps happening.
The split plan is saved across pause/resume, so it survives an interrupted run.
Does it respect limits and cost extra?
- It respects your safety limits. Split Search still obeys the daily search and ban-safety limits described in scraping limits — it doesn't push your account past them.
- No surcharge for splitting. There's no extra credit cost for the split itself. You're charged the normal per-lead rate only for the leads you actually collect — the same as any other run.