Automate for recruiters: decide three things before you pay

Three things sit in front of you before any payment: which tier suits your search volume, whether your model usage stays with the platform or moves to your own key, and which of your existing tools you are willing to connect. Take them in that order, because each answer narrows the next one and stops you buying reach you will not use in a narrow specialism.

What you are paying for is a single connection screen where the apps you already run sit side by side, plus a place for automated steps to work inside those accounts. Nothing is copied into a shared pool of candidate records, so what you search reflects your own account state on the day you search it rather than a list that has drifted since it was gathered.

A guided setup takes you through the first connections and shows an activity log of what each run touched. That log is the part that matters commercially, because it shows whether the steps you built are doing the sourcing work you wanted or sitting idle. Read it before you commit to a larger volume tier.

Starter lifetime gives you a first tier of volume and the core connection and logging setup. For a recruiter in one specialism it is enough to test a short sequence: pull a set of roles and locations, check the matches by hand, and see whether the steps hold up across a week of real searching before you widen anything.

Growth lifetime raises the volume you can run, which matters once several searches a day become normal and you want the same sequence working across more roles or locations. The screen, the accounts and the activity log stay the same; what changes is how much work you can push through them before you reach the tier ceiling.

BYOK unlimited lifetime removes the volume ceiling and puts model usage on your own provider key. That suits a recruiter who already holds a key, wants no cap on runs, and is content to keep model billing separate. It is a different purchase from the volume tiers, not simply a bigger version of them.

The fear is a database that has gone stale, so the role you need is missing or the person listed has moved on. That fear assumes a shared pool. Here the steps act on your own connected accounts, so the state you read is the state those accounts hold, not a snapshot refreshed on someone else's schedule.

The other half of the doubt is control. Automated steps that would delete or overwrite records pause and wait for your confirmation, so a sequence cannot quietly remove a candidate from a list you maintain. You see each pause in the activity log and decide whether to continue, which keeps a human in the loop on the steps that carry risk.

What remains is honest: what a run can touch follows the permissions on each connected account, so an app your administrator has not authorised will stall that step until access is granted. Plan your first sequence around the tools you already control, and add the restricted ones once the permissions are in place.

Move up when the log shows the lower tier regularly hitting its ceiling, not when a feature list looks fuller. If your searches queue against the volume limit, or you find yourself repeating the same role and location sequence by hand because the tier will not take another run, that is a volume signal and Growth lifetime answers it.

Choose the key-owned option only if you already hold a model key and want no cap on runs. There, your provider bills each model request separately from the lifetime price, so the comparison is between a fixed ceiling and a bill that moves with how much you search. Neither is wrong; they suit different working patterns.

If you are unsure, stay where you are for another week and watch the activity log. A recruiter in a narrow specialism often finds the lower volume is enough once the sequence is tuned, and the cost of upgrading early is simply volume you paid for and never used.