The loop, in a picture
Six steps. Each one nudges the next. The loop reinforces itself.What standing changes about your experience
A few concrete examples.Higher tier = higher rank in employer searches
A site contractor in Bodija posts a job for bricklayers. Our matching model ranks every plausible candidate. All else equal, a Trusted-tier bricklayer ranks above an Emerging-tier one. The employer sees you sooner; you get the application earlier; you get hired more often.Higher tier = different jobs in the marketplace
The marketplace atguild.com.ng/jobs shows different jobs to different users. Premium jobs (higher pay, larger employers, longer projects) are gated by tier. Building-tier Job Seekers see entry-level work; Trusted-tier sees the full marketplace; Established sees the highest-paying roles first.
This isn’t gatekeeping for its own sake. Employers who post premium work specifically want reliable, high-standing candidates. We surface their jobs to candidates who match.
Higher tier = credit access
Building and Emerging tiers don’t typically qualify for credit. Trusted opens the door. Established gets the best terms. This isn’t a Guild policy; it’s the partner lenders’ underwriting policy. They’ve calibrated their risk models against our tier classifications. A Trusted-tier borrower has a default rate they’re willing to underwrite at; a Building-tier borrower doesn’t yet have the signal for that risk model.Higher tier = different language in your interactions
Subtle but real. A Trusted-tier Job Seeker is greeted with “Welcome back” in a familiar voice. A first-time caller gets the onboarding flow. The system knows where you are in your journey.What changes your standing
Six categories of action.Completing a job
The biggest single signal. EveryCOMPLETED application adds to your score. The boost is larger for higher-paying jobs (proxy for skill) and for jobs that lasted multiple shifts (proxy for sustained reliability).
Getting a review
After a job ends, the employer rates you. A 5-star review nudges your score up. A 4-star is neutral-positive. A 3-star is neutral. A 2-star or below is negative. A genuinely bad review can be flagged for moderation if you believe it’s retaliatory or false. Resolved disputes neutralise the review.Showing up on time
Attendance is marked when you arrive. On-time rate is a calculated signal: shifts where you arrived on time / total shifts. The model treats this as a strong signal of reliability.Working consistently
A user who does one job a week for six months ranks higher than one who did six jobs in one week and then went quiet. The model rewards consistency.Repaying credit (if applicable)
If you’ve taken a loan through a partner lender, every on-time repayment adds to arepayment_streak counter. A clean streak is one of the strongest positive signals in the model. A missed payment resets the streak and adds a negative signal.
Diversifying your employer base
Working for many distinct employers (over time) is a small positive signal. It suggests you’re not over-reliant on one relationship — which matters for resilience if that one employer disappears.What pulls your standing down
The list is short.
You can always recover. The model reads the trailing 90 days; three months of consistent good behaviour erases most of a bad month.
The other half: how your behaviour ripples
This is where the loop becomes interesting. When you do something the model considers positive, the score updates and the matching model re-ranks you. The next job that posts will surface you earlier to its employer. The employer is more likely to approve you. You complete it, get a good review, your score goes up further, you rank higher again. Conversely, a missed shift drops you in ranking. Fewer matches. Fewer alerts. The loop dampens — until you do something to restart it. This isn’t gamification. It’s the same dynamic as a real workforce: reliable workers get the next call; unreliable ones don’t. The platform is making that loop visible and traversable.How the model improves
The model itself learns. As more accept/decline/complete signals flow, the matching model’s ranking gets sharper. The standing model’s tier-boundary calibration tightens. The negotiation engine’s posteriors converge to the offer patterns that actually close deals. You don’t see this directly. What you see is the platform becoming better over time at surfacing work that fits you. The first month feels useful; the third month feels surprisingly good.What to read next
How it works
The model and its signals.
The four tiers
What each tier unlocks.
The matching model
The architecture behind the ranking.