How interpreter ranking works
Learn how Fluent ranks interpreters and partner agencies on the Assignment step — what factors go into the score, and how to read the candidate list.
At a glance
- Every qualified candidate gets a score from 0 to 1. Higher scores rank higher in the list.
- The score is built from four things: completions with this customer, completions with your agency overall, acceptance rate, and proximity (for in-person appointments only).
- Recent activity counts more than old history. An interpreter who hasn’t worked with you in over a year gets their score discounted automatically.
- Partner agencies are ranked in the same list as interpreters, with a built-in discount so direct interpreters generally rank higher.
- New interpreters get a small temporary boost so they aren’t buried at the bottom of every list during their first 90 days.
- Each row shows the inputs that produced the ranking. There’s no hidden math.
What the ranking is trying to do
When you open the Assignment step, Fluent’s job is to surface your most likely candidates first — so the people who already know this customer, accept reliably, and are practical for the appointment land at the top of the list. The ranking is a starting point, not the final word. You can always assign someone further down the list when your judgment says otherwise.
The four factors
Every interpreter gets a score built from four components. The weights differ slightly depending on whether the appointment is in-person, video, or phone — proximity only matters when someone has to physically show up.
Completions with this customer
The heaviest factor. An interpreter who has done twelve appointments for this hospital or school district before is almost always a stronger choice than a stranger. They know the people, the workflows, and the subject matter.
This component caps at 25 completions — after that, additional appointments don’t push the score higher. The signal is “they’re established with this customer,” not “they’ve done the most appointments.”
Completions with your agency
A secondary signal. Even when an interpreter hasn’t worked with this specific customer before, a long history of completed appointments with your agency is a strong signal of overall reliability and quality. It carries less weight than customer-specific history, but it counts in both in-person and remote rankings.
This component caps at 100 completions.
Acceptance rate
How often the interpreter accepts the offers you send them. An interpreter who routinely declines wastes scheduler time, even if their other stats are strong. The system rewards consistency.
Acceptance rate is calculated from the last 12 months of resolved offers. New interpreters with fewer than 10 offers in that window are scored against the platform average instead of their own thin data — so they aren’t unfairly boosted or punished while you’re still getting to know them.
Proximity (in-person only)
How close the interpreter lives to the appointment location. Distance is a factor, not a hard limit — a great interpreter 80 miles away is often still better than a mediocre one across the street.
For video and phone appointments, proximity isn’t used at all. The remaining factors are weighted slightly higher to compensate.
The weights
For in-person appointments, the four factors are blended in fixed proportions. Customer history carries the most weight, with proximity as a meaningful secondary factor:
- Customer history: 40%
- Agency history: 15%
- Acceptance rate: 20%
- Proximity: 25%
For video and phone appointments, proximity disappears and its weight is redistributed:
- Customer history: 55%
- Agency history: 20%
- Acceptance rate: 25%
The weights are fixed across all agencies and aren’t currently configurable.
Adjustments on top of the score
A few situations adjust an interpreter’s score after the basic math:
Recent activity counts more than old history
An interpreter with 30 completed appointments three years ago isn’t really equivalent to one with 8 completions in the last six months. To keep the rankings honest, the system scales completion scores by how recently the interpreter last worked with your agency. Completions in the last 6 months count fully; older ones gradually count for less. Anything beyond two years has only marginal weight.
Recency is measured against the interpreter’s last appointment with your agency overall — not their last appointment with this specific customer. An interpreter who’s still actively working with you, just not with this customer in a while, doesn’t get penalized.
New interpreters get a small boost
Brand-new interpreters obviously have zero customer history and few accepted offers, which would otherwise bury them at the bottom of every list. To give them a fair shot at their first appointments, interpreters in their first 90 days get a small score bonus — until they’ve completed 25 appointments, at which point they have enough real data to stand on their own. The boost gradually decreases as they accumulate completions, so there’s no abrupt drop when it ends.
Unavailable interpreters are dimmed, not hidden
If an interpreter’s calendar shows a conflict with the appointment time — either outside their stated availability hours or because they’re already scheduled for another appointment during that window — their score is cut in half, but they stay in the list.
This is intentional. If your normal first-pick interpreter has a conflict today, you still want to see them in the list so you can decide whether to reach out anyway and rearrange your schedule. Hiding them entirely would mean you’d never know they were an option.
Interpreters who haven’t set up their availability yet aren’t penalized — they’re treated as potentially available, with a small note on the row so you know their availability isn’t on file.
Where partner agencies fit in
Partner agencies are scored on a similar formula and surfaced in the same ranked list as your direct interpreters. The goal is to show you your best option for the appointment, whether that’s an interpreter or an agency, without making you toggle between two views.
That said, partner agencies get a substantial built-in discount. All else equal, a direct interpreter beats a partner agency — fewer hops, more control over the assignment, faster confirmation. Agencies need a meaningfully stronger underlying score to outrank a comparable interpreter, which means they tend to surface when your direct roster is weak for a given appointment, not when it’s strong.
The discount also has a useful side effect for dormant relationships. Partner agencies you haven’t worked with in over a year get their history discounted by the same recency rule, and the agency discount on top of that drives their score low enough that they naturally fall toward the bottom of the list. They’re still there if you need them — they just stop cluttering your top results.
Each row in the list shows whether it’s an interpreter or a partner agency, so you always know which type of candidate you’re looking at.
How to read the list
The candidate list is one continuous ranked list, paginated 10 at a time. There are no section headers and no toggles between views. As you click “View more,” you’ll typically see:
- Top of list: your direct interpreters and active partner agencies, interleaved by score. These are your strongest candidates.
- Middle: distant interpreters (more than 250 miles away for in-person appointments) — they’re still ranked, just pushed lower by the proximity penalty.
- Bottom: dormant partner agencies — old relationships that don’t currently fit the appointment well.
The order emerges from the score itself, not from hard partitions. This means a strong interpreter who happens to be 200 miles away can still rank above a closer one with weak history — which is usually what you want for in-person appointments where customer familiarity matters more than miles.
Why we show the math
Every row in the list shows the inputs that produced its ranking — completions with the customer, completions with your agency, acceptance rate, distance, and any active adjustments like the new-interpreter boost or recency discount.
There’s no hidden number. If interpreter A ranks above interpreter B, you can see exactly why. Schedulers trust rankings they can audit; they fight rankings that look like a black box.
The ranking is a starting point. Your judgment is the final word.