4 Ways to Improve Dispatch for On-Demand Rides and Deliveries

For an on-demand business, implementing effective dispatching can be a challenge. Matching your customers to the right agents happens during the dispatch process. And due to low decision windows and ETA thresholds, it’s crucial to make accurate decisions at the get-go. This is especially true at higher volumes, where your mapping platform needs to keep up and still stay cost-effective for your business.

Sounds challenging? It’s worth paying keen attention to.

Effective dispatch leads to better customer-agent matching right from the start, which has a direct impact on your product efficiency, customer experience, and in turn, your bottom line. Done right, good dispatching translates to:

• Faster turnaround time
• Lower operating costs
• Fewer cancellations
• Higher asset utilization

Getting better dispatch with better maps

Here are 4 ways to improve dispatch efficiency of on-demand rides and deliveries through an effective mapping platform.

1. Measure accurate driving distance during dispatch — the speed-accuracy trade-off

When you’re mapping distances, especially during the order placement, you need to ask yourself “Do I want this done fast or accurate?”.

When it comes to calculating ETAs, there’s a trade-off between how fast you can calculate it and how accurate it will be. Two key distance mapping methods are:

• Aerial distance: Also known as straight-line distance, this draws a line between your driver and the destination without considering roadways. Due to this, you have extremely fast agent-customer mapping. However, the estimated arrival time can have high variation.
• Routing distance: This calculation measures the actual on-road distance between the driver/agent and customer. This takes more time and processing than aerial distance measurement, but is far more accurate. While it may take more time than a radial search, it is far more accurate.

While aerial distance can be a lucrative mapping method since it’s fast, the accuracy trade-off is generally too high to make it useful for on-demand service requirements. Pairing routing distance calculation with a tailored mapping solution like NextBillion.ai is often the best way to go.

2. Set up accurate and relevant POIs

Accurate and context-specific POIs can greatly ease up first and last-mile challenges.

Most locations with multiple entry-exit points and parking lots are difficult to track on your maps —especially if you are using plug-and-play map providers. For a shopping mall or large apartment complex, cabs or food delivery vehicles can have completely different points of entry and different parking spaces. When these POIs don’t show up on your maps, it adversely affects your navigation, ETAs, and at the end, customer experience.

For example, in a large apartment complex, your agent can end up wasting precious minutes trying to locate the right entry, parking space, building or house, severely impacting the last-mile of your dispatch.

Through our map data management platform, you can easily add a layer of contextual POIs to your maps —such as precise pickup and drop-off points specific to your use case. Besides ensuring smooth delivery and pick-up experience, accurate POIs also lead to fast arrival times and low cancellation rates.

3. Know where your drivers actually are with Map Matching APIs

We’ve all been there — you book a ride, find a vehicle that seems close, keep looking around trying to find it until you realize you’re on the opposite sides of the road.

Even for city-wide operations, you can rarely expect perfect GPS accuracy. Add fluctuating network quality, on-road speed, and varying agent/driver devices to the mix, we have a big challenge to overcome.

Thankfully, the solution is far easier than ‘fixing GPS’ for your entire fleet. The right location data solutions can offset noisy GPS and make your routing and dispatch far more accurate.

Our ‘Snap to Road‘ API can solve this challenge for you. These customizable APIs are designed to offer accurate map matching in scenarios where you need to:

• Determine the side of street that the vehicle is on
• Predict the turn taken at an intersection
• Determine if the vehicle is on a highway vs on a service road

4. Pick scalable mapping solution that keeps up with your company growth

The silver bullet to offset large API calls is, of course, batching them together. The iconic ‘finding a driver’ message has now become an industry standard for on-demand rides and deliveries.

Since dispatch needs to finish in a few seconds, searching for nearby drivers can frequently demand large volumes of API calls. But without the time necessary for large batches, typical mapping platforms cannot handle them. For reference:

• Internal mapping systems often peak at 20-40K queries per second
• During peak hours, an on-demand delivery or ride-hailing product can see almost 100K+ queries per second in a metropolitan area
• Plug-and-play mapping platforms like Google only support up to 1000 queries per second

Our location data platform can help you tailor your dispatch to your use case and business needs to allow for high platform efficiency and customer experience without needing to pay unreasonably high API costs.

Compared to plug-and-play generic mapping APIs, here’s what our APIs have to offer:

• 3-5x lower latency
• 10-20x higher throughput
• Flexible pricing and deployment modes that are built for scale

For on-demand rides and deliveries, optimizing dispatch comes hand in hand with moving to a better map data management platform, where you can focus on:

• Optimizing ETAs
• Implementing better routing and navigation
• Rapidly adding POIs and updating your maps without being dependent on map providers

NextBillion.ai has helped companies like Gojek, Freight Tiger and L&T with their mapping technology. Reach out and see what we can do for you!

Build better maps with NextBillion.ai