Amplitude for Gig Economy Marketplaces

Why Lifecycle Optimization in Gig Marketplaces Is Different

You're operating a two-sided marketplace. Every lifecycle decision you make has to account for supply and demand simultaneously — which means a standard SaaS retention playbook will fail you.

When a rider churns from your rideshare platform, you lose demand. When a driver churns, you lose supply. If both happen in the same city at the same time, your marketplace liquidity collapses. Amplitude gives you the instrumentation to see these dynamics before they become emergencies — but only if you set it up to reflect how gig marketplaces actually work.

This guide covers the specific Amplitude setup, event taxonomy, segmentation logic, and automation triggers that gig economy growth teams should build.

---

Setting Up Amplitude for a Two-Sided Marketplace

Separate Your User Types from Day One

The most common mistake is treating all users as a single population. In Amplitude, create distinct user properties that identify user type at the account level:

• `user_type`: `"provider"` or `"consumer"`
• `market` or `city_id`: geographic unit of supply-demand balance
• `provider_tier`: `"top_performer"`, `"standard"`, `"at_risk"`
• `consumer_segment`: `"power_user"`, `"occasional"`, `"lapsed"`

Build every chart, funnel, and cohort with user type as a baseline filter. Mixing provider and consumer data in the same funnel analysis produces numbers that mean nothing.

Establish Your Marketplace Health Events

Before you track engagement, define what "healthy" means at the transaction level. Your core event schema should include:

Supply-side events:

• `provider_onboarding_started`
• `provider_verified` (with `verification_step` property)
• `provider_first_job_completed`
• `provider_session_started` (with `city_id`, `time_of_day`)
• `provider_job_accepted`
• `provider_job_declined` (with `decline_reason`)
• `provider_app_opened_no_job_available`

Demand-side events:

• `consumer_signup`
• `consumer_first_booking`
• `consumer_booking_completed`
• `consumer_booking_abandoned` (with `abandonment_stage`)
• `consumer_repeat_booking`
• `consumer_churned` (defined as 30 or 60 days without a booking, depending on your category)

Marketplace match events:

• `match_attempted`
• `match_successful`
• `match_failed` (with `failure_reason`: `"no_supply"`, `"provider_declined"`, `"timeout"`)
• `transaction_completed`
• `transaction_disputed`

These match events are your leading indicators for liquidity problems. A rising `match_failed` rate with `failure_reason: "no_supply"` in a specific city tells you exactly where to run a supply acquisition push.

---

Segments That Actually Drive Decisions

The Provider Lifecycle Segments

Build these as behavioral cohorts in Amplitude, not just property-based filters:

1. New Provider (Days 0–14): Signed up but not yet completed first job
2. Activated Provider: Completed 3+ jobs in first 30 days
3. Established Provider: Active in 3 of the last 4 weeks
4. At-Risk Provider: Was established, now inactive for 14–21 days
5. Churned Provider: No session in 30+ days

The activation threshold (3 jobs in 30 days) is a placeholder — run a correlation analysis in Amplitude between early job completion count and 90-day retention to find your actual number. Platforms typically find this threshold sits between 2 and 5 completed jobs.

The Consumer Lifecycle Segments

• Acquired: Signed up, no booking
• Activated: Completed first booking
• Retained: 2+ bookings in 60 days
• Habitual: Books on a predictable weekly or bi-weekly cadence
• Lapsed: No booking in 30–45 days after previously being retained
• Churned: No booking in 60–90 days

Use Amplitude's Compass feature to identify which early behaviors predict long-term retention. For most gig platforms, booking within 48 hours of signup is the single highest-leverage activation moment.

---

Funnels and Charts Worth Building

The Dual-Sided Activation Funnel

Build two parallel funnels — one for providers, one for consumers — and monitor them at the city level weekly:

Provider funnel: `provider_onboarding_started` → `provider_verified` → `provider_first_job_completed`

Consumer funnel: `consumer_signup` → `consumer_first_booking` → `consumer_repeat_booking`

Track conversion rates at each step by city. A market where provider verification completion drops below 40% needs a different intervention than one where first-job completion is the problem.

Supply-Demand Ratio by Cohort

Use Amplitude's user-level segmentation to calculate the ratio of active providers to active consumers per city per week. This is not a native Amplitude chart — you will need to export this data or build it via Amplitude's Data Tables feature and combine it with your market-level properties.

When this ratio drops below your viable threshold (typically 1 active provider per 8–12 consumers for on-demand services), match failure rates spike within 72 hours. This is your early warning system.

Retention by Job Category

If your platform spans multiple service categories — deliveries, home services, rides — build separate retention curves per category. Consumer booking frequency and natural churn windows differ dramatically between a weekly house cleaning and a same-day delivery. Treating them with the same re-engagement timing destroys your communication relevance.

---

Automations to Build Around Amplitude Data

Amplitude's Audiences and Sync features let you push cohorts directly into your CRM, push notification tool, or marketing automation platform.

Set up these syncs:

• At-risk provider sync: Any provider who was established and has had no session in 14 days → trigger a personalized outreach within 24 hours. Do not wait for 30-day churn.
• Unactivated consumer sync: Consumer signed up 24 hours ago with no booking → push notification or email with a first-booking incentive.
• Lapsed consumer win-back: Consumer with 2+ past bookings, no activity in 35 days → re-engagement sequence. The 35-day mark consistently outperforms 60-day re-engagement on conversion.
• High-value consumer flagging: Consumer who has completed 10+ bookings → suppress from promotional discounts and route to a loyalty or VIP flow.

---

Industry-Specific Challenges in Amplitude

Geographic data granularity: Amplitude's default location tracking is not precise enough for marketplace liquidity analysis. Pass `city_id` or `zone_id` as explicit event properties — do not rely on Amplitude's geo-IP resolution for operational decisions.

Provider dual-app behavior: Providers on delivery or ride platforms frequently run multiple apps simultaneously. Your session data will have gaps. Account for this in retention definitions — a provider with no sessions in your app for 7 days may still be active in the market.

Seasonal supply fluctuations: Build annotation markers in Amplitude for known supply events (holidays, weather events, local promotions). Without these, your cohort analysis will produce false signals around retention drops.

Delayed job completion data: If job completion is confirmed asynchronously (hours after the transaction), your funnel completion events will appear delayed. Use server-side event ingestion for completion events to ensure timestamp accuracy.

---

Frequently Asked Questions

How should I define "activation" differently for providers versus consumers?

Activation for a consumer typically means completing a first transaction. For a provider, completing one job is not enough — they have not yet formed a habit. Use Amplitude's retention analysis to find the minimum early job count that predicts 90-day retention. For most platforms, this is 3–5 completed jobs within the first 21 days.

Can Amplitude handle marketplace supply-demand analytics natively?

Amplitude is strong at per-user behavioral analysis but was not built for marketplace liquidity metrics. You can approximate supply-demand ratios using Data Tables and user counts, but for true real-time liquidity dashboards you will likely need a secondary tool like Looker or Metabase pulling from your data warehouse, with Amplitude focused on behavioral segmentation and lifecycle work.

What is the right re-engagement window for lapsed gig consumers?

It depends on your booking category. For high-frequency services like food delivery, lapsed starts at 10–14 days. For low-frequency services like home cleaning, lapsed starts at 45–60 days. Build separate lifecycle definitions per category in Amplitude rather than applying one universal window.

How do I use Amplitude to reduce provider churn without over-incentivizing?

Segment providers by lifetime job count and earnings before triggering any incentive-based re-engagement. A provider who completed 50 jobs and went inactive has different economics than one who completed 3. Use Amplitude's cohort analysis to identify which at-risk providers have historically reactivated without an incentive — and only deploy incentives for the segments where behavioral nudges alone do not move the needle.