Grubhub is slowly shrinking delivery pay on higher tip orders
I posted my first impressions on the new Grubhub Pay Model here.
At the time, it appeared that they were paying slightly more per trip than they would have under the original model. I mentioned at the time that I was curious if that would slowly change.
At first it was hard to really catch any real trends. In the three weeks since, I've noticed a few trends. The change seems to be happening week to week.
I tracked as many orders as I could. Google maps is great because you can measure the straight-line distance between points. That's how Grubhub calculated the old model: In my market they paid $3.50 plus 50 cents per mile based on the straight line distance from restaurant to customer.
It's interesting now to see the trends take shape.
Week 1: Fares were looking better.
I was able to capture screenshots of the offer of 33 deliveries in the first week. That allowed me to measure the point to point distances and calculate pay under the original model.
In the first week, the average fare I received (not including tip) was $4.51. Under the old model, the average fare would have been $4.27. So it was basically a 6% increase in the Grubhub portion of our pay.
Once you add in tips, that only actually amounted to an overall 3% increase. Still, an increase is a good thing, right?
Week 2: Fares are better still… but not by as much.
The second week of the new model was the one where I could calculate delivery fees. I had 47 trips with data that I could measure. I averaged $4.53 in fares, where those same trips would have paid out only $4.33 on average. That's still an increase, but now only a 4.5% increase
Week 3: Things, they are a changing.
Now to be fair, I did have a smaller data set, with only 17 measurable deliveries.
Week three shows a noticeable decrease in pay compared to the old model. The average fare per delivery was $4.27. Under the old pay model it would have been $4.41.
For the first time, I was making less than under the old model. It was a three percent DECREASE.
I wondered if this was an anomaly because I had a smaller data set. I only had 17 trips that I could measure. However, when I looked at each week on a trip by trip basis, the individual trips largely line up with the overall trend.
The pattern in my market shows that Grubhub seems to be decreasing pay from one week to the next, but in very small increments.
What about any other patterns?
All three weeks (and even so far this week) I have noticed that there were trips where the old model would have paid better, and others where the new model would have paid better. So far I haven't seen enough of a pattern to figure out if there's a new formula in play.
After the first week, I didn't notice that deliveries with a longer anticipated wait time were paying more than others. Grubhub said time for delivery would now be a factor. In my opinion, that's a farce.
Then one thing caught my eye. In the past couple of weeks I had some Taco Bell orders where the tip was low but the payout was significantly higher (about $2.25 higher than I normally would have received). So I started wondering about whether tips would make a difference.
I sorted the trips by tip amount. Here's where it gets interesting, especially week by week.
When the tip was $5 or higher in the first week, the average delivery pay was $4.55, This compared to the $4.45 I calculated under the old model.
If the tip was less than $5, the average delivery pay was $4.45. The old model would have paid $4.07.
When the tip was higher, the new model was paying 2% less. When the tip was lower, the new model paid 8.5% more than the old model.
In week 2, when the tip was $5 or higher, the average fare was $4.58 compared to $4.52 under the old model .
If the tip was less than $5, the average fare was $4.47, compared to $4.14 under the old model.
So now I'm at a 1% lower pay if the tip was good, and about 7.5% more when the tip was lower.
So a slight scootch of a difference for the second week. It's not enough of a variation from week one to week two to draw any conclusions with a small set of data. However, it's interesting that the pattern was consistent between higher tip and lower tip deliveries.
In week 3, when the tip was $5 or higher, the average fare was $4.04, it would have been $4.38 under the old model.
When the tip was less than $5, average delivery pay was $4.45. It would have been $4.42 under the old pay model.
When tips were higher, the new model now was paying about 8.5% LESS than the old model.
In other words, on higher paying trips, I was now seeing an 8% DROP in my earnings on higher paying tips. When tips were lower, pay was nearly the same. It averaged less than a 1% increase over the old model.
So far in this fourth week, 100% of my higher tip orders have been paying out less than they would have under the old model. The decrease on those has been a whopping 18%. On lower tip orders, I've seenabout a 4% increase.
Grubhub is following Doordash's lead by using large tips as an excuse to pay less
Under the old model, $3.50 was the bare minimum. We received $3.50 PLUS mileage.
So far in this fourth week of the pay model, my average delivery pay on orders where the customer tipped $5 or more was $3.37. In other words, my total delivery pay on those orders was 4% lower than the BASE pay under the old model.
It seems pretty clear Grubhub is using higher tips as an excuse to pay less out of their own pockets.
Now it's not as bad as Doordash, at least not yet. You can read more about their pay model here, but Grubhub is getting as low as $3 per delivery when the tips are good. At least Doordash has been more up front about this practice.
I'll keep tracking, and looking for patterns.
If you are one of the markets that has had the new model (where the mileage breakout is taken away) – have you seen any patterns yet in the new pay model?