I tried being a Delivery Rider, and thought about how to maximise earnings. Here's what I found.

What is the most optimal trip a delivery rider can take?

Delivery riders have risen to prominence during the pandemic. As our lifestyles changed, we relied on these?everyday heroes?- for food, groceries, documents, and more. They sometimes?brave inclement weather?to fulfil orders. One delivery rider even made the news for?receiving cheers after unintentionally joining the mobile column?during NDP 2020. But most times, their sacrifices and hard work goes unnoticed.

Having a bit of spare time, I thought I'd try being a delivery rider. One Friday afternoon, I?signed up on Lalamove?and, within the day, was approved to take orders. That was the beginning of another way to?Rediscover Singapore?and recoup some motorcycle costs.

Rediscovering Singapore

Border restrictions allowed me (and most of us) more time to explore our own backyard. Singapore's extensive Park Connector Networks are great for traversing different neighbourhoods. Delivering items created a need to go even further and added a new dimension - interacting with shop owners and individuals. One trip took me from a light industrial area in Alexandra to a condominium in West Coast and all the way to an apartment in Admiralty - close enough to see the Johor Bahru skyline. There's still much to see in Singapore.

What I loved most - was witnessing everyday Singaporean lives. I saw so much - from delivering food, birthday cakes, cookies from a home baker, and even medicine from a telemedicine dispensary. In a city where life is so fast-paced, affording the time to reflect and immerse is immensely valuable. There are so many stories and energies connected to this "delivery". Stories that I could only begin to imagine.

I watched an?uncle?resting his arms on the ledge along the common corridors, looking out at the skyline. I paused for a few moments to share his view. It was calming. This was a reminder to take more moments to be still and be present.

The Approach - Optimising Trips for Returns

After taking a few trips, I began to wonder if there were more 'optimal' trips to take. I began to look into the pricing structure and time taken to perform deliveries. The question I tried to answer was: "which deliveries would give me the maximum return on time?"

Some base data and assumptions used:

1. Lalamove's Courier Charges?(for Motorcycles): The base charge is $10 for the first kilometre (km) and $0.35 for every subsequent km. Every extra stop is $3. The commission rate is 15%, in addition to a per-trip driver's tax of $0.20.
2. Time Taken: By taking an average of my turnaround times, each stop (picking up or delivering items) takes 7 minutes. The average travelling speed I clocked was 40km/h.
3. Petrol Costs: Taking historical fuel consumption data from my Yahama YBR125, I averaged a cost of $0.05 per km. Also noteworthy is that Lalamove riders get 25% of petrol at Sinopec outlets - the highest discount rate in Singapore. I used this to determine the net returns from the gross amount received from Lalamove.

I define net returns as the amount received from Lalamove (after deducting commissions and driver's tax), less the petrol costs ($0.05/km). Motorcycle depreciation and other ancillary costs (insurance, road tax, etc.) were not considered for this analysis because they are not incremental costs.

Combining the above data and assumptions, I computed the net returns by kilometres and estimated the amount of time required to complete a trip. A sample calculation:

Findings

Based on data gathered and computed, I synthesise some points to note. In addition, there are some observations I've made from riding. These are listed below.

1. Shorter trips are more efficient

Generally, shorter trips (by distance) take less time. A 1km trip, however unlikely, gives a net return of $0.53 per minute. Getting quick jobs aren't frequent; it's like striking the jackpot if you do. I noticed that shorter jobs spawn more frequently at areas closer to the CBD (e.g. Robertson Quay, River Valley, etc.) during lunch and dinner time.

The crossover point for which trips with more stops become more time-effective is at the 10km mark. For jobs shorter than 10km, you'll earn more per time spent on jobs without extra stops.

This is an excellent observation to keep in mind if you see two jobs originating from the same location. These could be food orders arising from the same café going to different places.?Given two jobs stemming from the same establishment, the shorter one gives better returns.

Are shorter trips always better? I initially thought so. However, remembering that the initial goal was to maximise the return on time, I replotted the data points.

2. For any given time spent - the more stops, the better

For any given time, the more stops there are, the better. This is because each extra stop brings a post-commission bonus of $2.55. Given an assumption that each extra stop is 7 minutes, that works out to $0.36 per minute. Recall that in the chart above, the "crossover" where jobs with more stops were more time-efficient was at the 10km mark, where the return was $0.36 per minute.

Therefore, my observation is that if you're working long shifts,?jobs with more stops maximise your time returns. As the job gets longer, the effects of having the "extra stop" charge do not amount to a significant difference anymore.

3. Time return per minute converges to ~$0.22/minute

That works out to roughly $13.20 per hour. After an 8 hour shift, the hourly rate net of overheads (depreciation, insurance, road tax, season parking) works out to approximately $11 per hour. This assumes that you can get jobs one after another.

From my experience, jobs don't come so immediately. Even if they do, they involve a bit of travelling time in between your last drop-off point and the next pick-up point. Effectively, that means inevitable inefficiencies, and the hourly rate will most likely be less than $11 per hour.

If you're riding for a long shift, the goal is to ensure that you have a?quick turnaround between trips. Shorter trips are a bonus. But ultimately, it is still better to be on a job in the long run to maximise long-term earnings.

4. Some locations are better than others

I noted that jobs ending in primarily residential areas, like Woodlands, Punggol, and Pasir Ris, are less favourable. The following job's start location is not likely to be anywhere near there. Picking a?job that ends in a more central location?gives you a better chance of a faster turnaround. Lalamove sometimes states which sites have more orders. Of course, that being said, these jobs are good if you're explicitly heading to a particular area.

5. Riding to malls and condominiums significantly slow you down

The additional measures that are taken by malls, condominiums significantly add to the delivery time. Trips to these establishments sometimes easily double the time taken.

For instance, this condominium disallowed delivery riders from entering the car park. Furthermore, there was a lengthy check-in process at the guard post. Waiting in line, giving all the details, and walking further into the residential blocks quickly added 7 minutes to the delivery time.

While attempting the delivery, another GrabFood rider quipped how some condominium or office building lifts required individual authorisation. i.e. For every level you want to go to, you need an access card to tap you in, or residents need to allow you in via the intercom. During peak periods when many are lined up, this can be frustrating.

How may these additional time incurred be addressed?

• Perhaps the fares could be adjusted to the ease of accessing the location. This is similar to "zone charging", where orders originating from the CBD incur a higher charge. An additional charge may be levied if the order originates and/or ends in a mall, condominium, or office building. Similarly, those at more accessible locations can be charged less.
• It may also be possible to set up a common delivery point. E.g. at the mall entrance, the security post for condominiums, and the sender leaves/recipient collects the items from that point. Customers who opt for these centralised points would not be levied an additional fee. A faster turnaround results in a greater capacity for more jobs to be taken. More potential earnings for riders and more capacity for delivery platforms.
• In practice, this is difficult to determine because each postal code in Singapore will need to be categorised. But a step towards more fair pricing based on location accessibility could start from analysing the amount of time a rider's GPS location remains at a particular spot.

In the meantime, my observations are that?landed properties/shophouses are the most accessible stops that offer the fastest turnaround time, followed by HDB apartments, and lastly, condominiums and office buildings. I would also qualify that you don't have the luxury of time scrutinising the type of property most of the time. Riders only have 8 seconds to indicate their interest in a job before it gets allocated to someone.

The Pecking Order

To sum the discussion up, I propose a few points on deciding which jobs to take:

1. Opt for?jobs that end in more central locations?for longer shifts, thereby ensuring a faster turnaround time. In the long run, having more time on a job earns more.
2. Wherever possible, opt for?jobs with deliveries to/from more accessible locations?like shophouses or landed properties. In reality, this is rarely possible. It is ultimately still better to be on a job than not.
3. For two or more given jobs arising from the?same location, the?shorter job?generates a better return on time. If longer jobs are being considered, opt for jobs with?no stops for those less than 10km?and?more stops for those more than 10km.
4. If?time?is a limiting factor, consider that?jobs with more stops?give better returns. For example, if you only have 60 minutes - a job with 2 extra stops gives $0.31 per minute, 1 extra stop gives $0.28 per minute, and no additional stops gives $0.26 per minute.

While the model used to determine time spent is relatively simplistic, it serves as a sufficient basis for getting initial insights on what are the most optimal jobs to take. Given more time, I would consider taking actual data points (instead of a simple average) to revisit this analysis.

Postlude: Be kind!

If you (consumers) can help riders improve their turnaround time, that would make a rider's day. It could be as simple as being more specific with your delivery instructions. Or even offering to meet at the ground floor/car park to receive the item. These acts help to maximise their turnaround time and allow them to earn more. Any act of kindness goes a long way.