Gig & Side IncomeIntermediate5 min read

Ratings, acceptance rates, and completion: which gig metrics actually matter

The apps track everything and imply everything matters. Mostly, it doesn't. Which numbers protect your account, which unlock perks, and which to ignore.

Every gig app shows you a dashboard of metrics engineered to feel like a report card: acceptance rate, completion rate, customer rating, on-time percentage. New drivers treat them all as sacred and accept garbage offers to protect numbers that were never in danger. Veterans know the truth: a couple of metrics can end your account, a couple can earn you perks, and the rest are motivational wallpaper.

The metrics that can end your account

Two numbers carry real deactivation risk on most platforms. Completion rate — the share of accepted jobs you actually finish — is the big one: canceling after accepting strands a customer and platforms enforce minimums (often around 80–95%, varying by app). Customer rating has a floor too: fall well below your market's average (commonly cited around 4.6 or so for rideshare) and you get warnings, then a review. These two deserve genuine protection.

  • Completion rate: keep cancellations rare and pre-pickup when unavoidable — canceling before pickup is treated far more gently than after.
  • Customer rating: driven mostly by basics — cleanliness, following delivery instructions, communication when things go wrong.
  • On-time/delivery metrics: enforced on delivery apps, and the main way sloppy multi-apping shows up.
  • Fraud-style flags (GPS anomalies, undelivered orders): rare but instant — document deliveries with photos and never mark 'delivered' early.

Acceptance rate: the metric designed to scare you

On most major platforms, acceptance rate alone cannot get you deactivated — the apps need that flexibility to be true for contractor-status reasons. What declining does cost you, on some platforms, is perks: priority tiers, better offer visibility, scheduling advantages. That's a real trade, but it's a math problem, not a compliance problem — and the apps benefit enormously when drivers don't notice the difference.

Pricing the perk tier
Trey's delivery app gives priority offers at 70%+ acceptance. Holding 70% means taking roughly 15 bad offers a week that net about $3 each after mileage costs — call it $45 of low-value work consuming five hours. His alternative: accept 45%, cherry-pick, and earn about $70 more per week in the same hours based on his two-week log. The 'perk' costs him $25/week plus wear on the car. On another driver's market and app, priority access might genuinely win — the point is he priced it instead of obeying it.
Run the two-week experiment
Log one week playing for the perk tier (high acceptance) and one week cherry-picking with a per-mile floor, tracking net earnings and miles for each. Your market's answer beats any forum's. Repeat the test when the app changes its incentive structure — and they change it often.

Protecting the numbers that matter, cheaply

  1. Decline instead of cancel: the moment before accepting is free; the moment after costs completion rate.
  2. Photograph every delivery and follow dropoff instructions literally — most bad ratings and 'never arrived' claims die against a photo.
  3. Message early when something breaks (long restaurant wait, closed gate) — customers rate communication, not just outcomes.
  4. Check your rating breakdown monthly; a sudden dip usually traces to one fixable habit, not bad luck.
  5. Know each app's published minimums and appeal process before you need them — screenshots of your metrics history help in disputes.
Don't buy ratings with your own money
Some drivers tip-chase with free candy, hand-warmers, and five-star begging signs. Small touches are fine, but spending real money to defend a 4.9 you'd keep anyway at 4.8 is negative-margin theater. Ratings need to stay above the floor, not reach the ceiling — there is no bonus for a perfect score, only costs.

The bottom line

Treat gig metrics like a triage list: completion rate and customer rating protect your account — guard them with pre-pickup declines, photos, and communication. Acceptance rate is a negotiation between you and the app's perk structure — test what the tier is actually worth in dollars. Everything else on the dashboard is decoration. Selective drivers with clean completion records earn more and get deactivated less; that combination is the entire game.

A worked example: what declining low offers actually costs

A delivery driver is offered 30 orders in a shift and accepts all of them, grossing $148 across eight hours with several $3 orders dragging the average down — about $18.50 an hour. The next week she declines everything under her floor, accepts 19 of 30 offers, and grosses $139 in six and a half active hours: roughly $21.40 an hour, plus ninety minutes of her evening back. Her acceptance rate drops to 63 percent. On most delivery platforms, nothing bad happens — acceptance rate is not a deactivation metric there, though it may cost access to priority-order programs. The lesson is double-edged: selectivity usually raises hourly earnings, but the price varies by platform, and you need to know which metrics are safety-critical before optimizing any of them.

MetricDeactivation riskWhat it usually affects
Acceptance rateRarelyPriority programs, offer quality
Completion rateYes, highAccount standing directly
Customer ratingYes, below thresholdAccount standing, order access
On-time rateSometimesFlags, program eligibility
Which metrics typically matter where (general patterns; check your platform's current policy)

Protecting the metrics that can end your account

The operational rule is simple: be ruthless before accepting and impeccable afterward. Almost every ratings disaster comes from inverting that — accepting everything, then canceling or delivering badly under pressure.

  • Never accept an order you might cancel; completion rate is the metric with the least forgiveness built in.
  • Decide in the first five seconds using distance and payout, not restaurant wait-time optimism.
  • Communicate early on delays, since a proactive message routinely converts a one-star mood into a five-star review.
  • Contest unfair ratings and violations immediately with photos and timestamps while the platform still has the data.
  • Track your own metrics weekly rather than discovering a threshold problem in a deactivation email.

Ratings systems reward consistency far more than heroics. A driver who is merely reliable — on time, communicative, careful with the last hundred feet of every delivery — will outlast and out-earn the one with occasional brilliance and a completion-rate problem, because the algorithm never meets you at your best. It averages you.

Check your understanding

1 of 3
Which two metrics carry the real deactivation risk on most platforms?

Not quite — try again.

The Worth letter

Get smarter about money every week

One email, no spam — practical guides and Worth updates. Unsubscribe anytime.

Put this into practice

Worth tracks your accounts, budgets, and goals — so the concepts in this article aren't just theory.

Start free trial