Orders come from wherever the kitchens and the stores are, which is rarely where the money is. That is the single most useful thing to understand about delivery demand: there are two maps, not one. The first shows where jobs originate — restaurant clusters, supermarket car parks, dark stores, warehouses — and it is dense, predictable and crowded with other drivers. The second shows where the drop-offs pay well, and it is a different shape entirely. Chasing one is not chasing the other, and a driver who knows which map they are reading makes better decisions than one who is simply driving toward the busy-looking part of the app.
Where do the orders actually come from?
From supply density: the more kitchens, shops and fulfilment points there are in a square mile, the more jobs originate there. This sounds obvious and is routinely ignored. A high street with two dozen takeaways, a retail park with a large supermarket, a cluster of dark stores serving rapid grocery — these are order factories, and they produce jobs at a rate that has nothing to do with how nice the area is or how many people live in it. On the rideshare side the equivalent is trip origin density: stations, airports, hospital shift changes, nightlife strips.
Which is why the first practical question is not "where is it busy?" but "what is near me that makes orders?" A driver sitting between three restaurant clusters is in a structurally better position than one sitting in a residential area with a single chain outlet, regardless of what any heat map was showing when they parked. Density is a property of the built environment; it is the same next Tuesday.
Why is the pickup map different from the tip map?
Because orders are created where the food is and generosity lives where the customer is, and those are different places. Restaurant clusters are commercial; the households ordering from them are spread across a wide ring of residential streets with very different incomes, building types and distances. The result is that the place you wait is chosen by supply and the place you are paid is chosen by demand, and no amount of sitting in the right pickup spot guarantees the right drop-off.
Drop-off characteristics matter in ways that pickup density never shows you. Distance out is unpaid distance back unless you get an order in the other direction. A tower block with no parking and a slow lift costs you minutes that never appear on any earnings screen. A long suburban run may pay well and then strand you somewhere with no orders. None of this is visible from the map of where jobs start, and it is most of what determines whether an hour was worth working.
What actually moves demand, hour to hour?
Time of day first, then weather, then events — in roughly that order of reliability. Meal times are the backbone and they are the most predictable thing in the job; everything else modulates them. The pattern is worth knowing as a shape rather than as a rule, because it differs by city and by what kind of delivery you do.
Events are the most profitable and the most dangerous of the four. A stadium emptying, a festival, a bank holiday, a big televised match: demand spikes hard, and so does congestion, road closure and the number of drivers who read the same news you did. The events worth working are usually the ones that create demand without creating gridlock — bad weather on a weekday evening is the classic example, because it suppresses the supply of drivers at the same time as it raises the number of people who would rather not go out.
- Meal times: the lunch window and the evening window, with the evening usually longer and denser
- Weather: rain and cold raise orders and thin out the bike and scooter fleet at the same time
- The working week: weekday evenings, Friday and Saturday behave nothing like a Monday lunchtime
- Events: matches, concerts, festivals and holidays, each with its own traffic penalty
- Local rhythm: university terms, shift changes at large employers, seasonal tourism
Should I chase density or chase tips?
Chase density for volume and the tip map for value, and know which one your situation needs. A driver whose problem is dead time — long gaps between offers — needs pickup density, because the fix is more opportunities per hour. A driver whose problem is that they are always busy and never ahead needs the other map, because the fix is fewer, better jobs and less unpaid distance between them.
The trap is treating the two as one strategy. Repositioning across town toward a neighbourhood that tips well is a real cost: fuel, wear, and the jobs you did not take while you were driving there. Repositioning toward a dense pickup cluster is cheaper but commits you to whatever the cluster offers. The honest answer is usually to sit where density is adequate and be selective about direction — take the jobs that move you toward the areas you want to be in, and treat the ones that strand you as more expensive than they look.
How do I read this from my own history?
By putting pay and distance on the same trips and then grouping them, which is something no platform dashboard does for you. The app knows what it paid you. It does not know the unpaid miles you drove to reach the pickup, the miles home from the last drop-off, or the jobs you did for another platform in between. Your own records are the only place where the complete picture of an hour exists.
The groupings worth building are simple and you only need a few weeks of data before they start saying something. Group by hour of the day and you find your real windows rather than the ones everybody assumes. Group by pickup area and you find which clusters actually feed you. Group by drop-off area and you find where the jobs that paid well ended up — which, over enough trips, is the tip map drawn from your own evidence instead of from forum folklore. The effective-hourly-rate guide on this blog covers the arithmetic of turning pay and unpaid distance into a number you can compare.
What do I do with the answer?
Change one thing at a time and let the records tell you whether it worked. Start an hour earlier for a fortnight and compare. Park at a different cluster for a week. Refuse the longest outbound runs on the nights you are trying to stay dense. Each change is testable because you have per-trip pay and per-trip distance, and the comparison is over your own driving rather than someone else’s advice on the internet.
BrewGig builds exactly these two maps — where orders originate and where the well-paying drop-offs are — from the anonymised trips of paying drivers, alongside the trip and earnings records you keep for tax. BrewGig is an independent product and is not affiliated with, endorsed by or partnered with any delivery or rideshare platform; all platform names mentioned here are the trademarks of their respective owners.