Trust and Safety Operations at Smaller Scale

Most writing about trust and safety operations, the day-to-day work of confirming who is on a platform and keeping everyone involved safe, describes companies with operations centers and specialist teams. Many platforms have an operator, some tooling, and decisions to make. This piece draws on a conversation on Know Your People between Spencer Chee, Gig Vertical Leader at First Advantage, and Chad Dennis, Head of Driver and Fleet Operations at Fetii, a group rideshare company in Austin. Dennis spent the previous decade in trust, safety and operations at some of the largest delivery and rideshare platforms, and the conversation sets out how the real operating decisions change at smaller scale.

What stays the same in trust and safety operations at any size

Two things carry over whole from the biggest platforms. The first is the foundations. Identity verification, the background check and the motor vehicle record anchor onboarding whether a platform runs hundreds of drivers or millions, and modern tooling means a small team can run them at the standard of the largest players. The second is the threat. Phishing, account takeover, in which a bad-faith actor seizes control of a legitimate worker’s account, and attempts to drain worker earnings reach platforms of every size. An edge case in a verification flow deserves the same attention too, because the gap is the same gap no matter how many workers pass through it.

Read about The four layers of workforce trust: a way of looking at gig economy background checks and identity verification

Can a smaller platform verify workers more deeply?

Often, yes, and this is the finding most operators do not expect. The giants make depth decisions under a constraint smaller platforms do not share. Across millions of workers, deep continuous verification of everybody fails the budget, so depth gets rationed by activity. A platform with fewer, more active workers can invert that. It can choose the most comprehensive verification available per worker and ask what will genuinely catch the most bad actors, at a per-worker cost a giant could never carry. The depth the biggest platforms ration becomes affordable precisely because the platform is smaller.

How often should a platform re-check its workers?

The practical question after onboarding is how often to look again, and the honest answer is that activity decides. This is re-verification cadence, the rhythm at which a platform confirms a worker is still who they say they are and that nothing new has appeared on their record. A worker on the platform daily may justify continuous monitoring. One who appears rarely may justify periodic re-runs instead. Each platform’s own risk profile turns that principle into a policy, which is why two well-run programs at different platforms can legitimately look different.

Where judgment replaces process in adjudication

Adjudication is the decision stage after a background check returns, when a platform decides what to do about something on an applicant’s record. At the largest scale it has to be process, because volume may make individual attention difficult. At smaller scale it can include more customization in making a decision. The strong applicant with something on their record gets weighed in full, local market rules are respected, and safety is put first without apology. Speed is a variable here too. Without tens of thousands of applications a week, an operation can afford to slow a decision down to get it right.

Designing friction for the day awareness fails

The better fraud scripts now study how a platform actually works and what a real employee would say. The operational answer pairs awareness, meaning everyone knows what the company will never ask for, with designed friction: behavioral checkpoints, so that an unusual sign-in followed by an attempt to redirect earnings triggers a fresh identity verification before funds move. The design premise is that somebody will eventually be fooled, and the system should hold even then.

What to ask a background screening partner

Two questions separate a partner from a supplier at this size. Can an operator change a verification flow without an engineering ticket? And does the pricing reflect the growth ahead rather than the volume behind, from a partner who has handled fast-scaling platforms before? A platform that expects to double should not be priced on last quarter.

Takeaway: a smaller platform’s program isn’t the giants’ program with pieces missing. It’s a different set of decisions, and several of them might be better.

The full conversation lives on the episode page. Listen or watch here

Frequently Asked Questions