How a Singapore Ride-Comparison Startup Streams Real-Time Fares from Grab, Gojek, Tada, Ryde & CDG Zig

5 providers

compared on one query

On-demand

arbitrary origin–destination pairs

500-row trial

to validate before scaling

Client: Singapore ride-hailing comparison startup (name withheld)

Industry: Mobility / Consumer App

Use Case: Real-time fare, ETA & surge comparison • Providers: Grab, Gojek, Tada, Ryde, CDG Zig

Delivery: On-demand API; trial-first, then managed feed

About the Client

Our client is a Singapore startup building a ride-hailing comparison platform — the "which app is cheapest right now?" layer for a market with five serious ride providers. The product's premise is that fares, ETAs and surge differ meaningfully between Grab, Gojek, Tada, Ryde and CDG Zig at any given moment, and that a shopper who compares before booking saves money. Delivering on that premise requires fare data that is genuinely real-time and genuinely comparable — for whatever route the user types in.

The Challenge

• Real-time, or worthless. A ride fare is valid for minutes. Unlike product prices, there's no useful "cached" fare — the data must reflect the quote a rider would get right now, for their exact route.
• Arbitrary origin–destination queries. The platform can't pre-scrape a fixed set of routes; users enter any pickup and drop-off. The data layer had to support on-demand OD-pair fare retrieval, not a static dataset.
• Five providers, five behaviours. Each app computes fares, surge and ETAs differently and exposes them differently. Normalizing them into one comparable structure was the hard part.
• Feasibility before commitment. As an early-stage founder, the client needed to validate that this was even reliably possible — and at what latency and cost — before building a product on top. A free trial was the natural first step.

The Actowiz Solution

1. On-Demand OD-Pair Fare Retrieval

Rather than a static dataset, we built an on-demand retrieval capability: the platform passes an origin, destination and time, and receives current fare, ETA and surge/availability signals for each provider serving that route — the live quote a rider would see.

2. Cross-Provider Normalization

Grab, Gojek, Tada, Ryde and CDG Zig fares are normalized into one schema — base fare, surge multiplier/indicator, estimated total, ETA, ride tier — so the platform can present a true apples-to-apples comparison instead of five incompatible quotes.

3. Latency-Optimized Architecture

Because the fare must feel real-time to the end user, the retrieval path is engineered for low latency and concurrent multi-provider queries, so a single comparison returns fast enough for a consumer app.

4. Trial-First, Then Managed Feed

We started with a free trial (a capped sample of OD-pair queries) so the founder could validate feasibility, latency and comparison quality against real routes before any commitment — then structured a managed feed with a clear path toward the client eventually owning more of the retrieval stack if desired.

The Results

• 5-in-1 — a single comparison query returns normalized real-time fares across all five providers — the core feature the product is built on
• Any route — on-demand OD-pair retrieval means users aren't limited to pre-defined routes — they type any pickup and drop-off and get live quotes
• Validated first — the 500-row-style free trial let the founder confirm feasibility, latency and quality before committing budget — de-risking the whole build
• Trial → managed — a clean path from proof-of-concept to a production managed feed, with the option to bring more of the stack in-house as the company scales

Client Feedback

"Every provider quotes differently, and fares are only good for minutes. We needed proof this could work at all before we built a company on it. The free trial answered that — then the managed feed made it real."

— Founder, Singapore ride-hailing comparison platform

Why It Worked

• On-demand, not static. Ride-hailing can't be pre-scraped — building for arbitrary OD-pair retrieval matched how the product actually queries.
• Normalization is the comparison. Five providers only become "cheapest vs rest" once their fares sit in one schema — that normalization is the product's value.
• Prove before you build. A trial-first approach let a first-time founder de-risk feasibility before spend — the right way to start a data-dependent product.

Frequently Asked Questions

Which ride-hailing providers and markets can you cover?

Grab, Gojek, Tada, Ryde, CDG Zig in Singapore; Uber, Bolt, Careem and regional players across SEA, MENA, Europe and beyond. We confirm feasibility per provider/market during the trial.

Can fares really be retrieved for any route on demand?

Yes — the retrieval capability takes an origin, destination and time and returns live quotes per provider. It's not a static dataset; it's built for arbitrary OD-pair queries.

How is "real-time" actually achieved?

The path is latency-optimized with concurrent multi-provider queries, so a comparison returns fast enough for a consumer app. Because ride fares expire in minutes, no useful caching is applied to the fare itself.

Do you offer a trial before we commit?

Yes — a capped free trial of OD-pair queries is the standard first step, so you validate feasibility, latency and comparison quality on real routes before any contract.

Is ride-hailing fare data collection compliant?

We collect publicly available fare-estimate and ETA information as presented to a prospective rider, with no rider accounts or personal data involved, under Actowiz's responsible-scraping framework. We scope each provider's specifics with you during the trial.