DoorDash is requiring engineers to deliver food | Hacker News Reader

DoorDash is requiring engineers to deliver food

sfgate.com

Previously https://news.ycombinator.com/item?id=29716186

Thanks! Macroexpanded:

Every DoorDash employee, from engineers to CEO, will make deliveries - https://news.ycombinator.com/item?id=29716186 - Dec 2021 (152 comments)

DoorDash will require all employees to deliver goods - https://news.ycombinator.com/item?id=29675076 - Dec 2021 (133 comments)

Requiring the engineer to actually use the product they are working on in real life is good. Maybe it can help make the experience better for the dashers, the restaurants, and the company.

That $400,000/yr engineer makes more money in a day than a DoorDash worker makes in 2 weeks working full time. I do not feel any sympathy for the 1 dash they will have to make a month.

Since I first heard about this, this is what I pictured in my head:

Alex: I’m headed to Bob’s for some lunch. Want anything?

Tory: Sure. Hey, let me put in a DD order for it.

Alex: Sweet, that’s this month down…

what does the engineers salary have to do with anything

"what does the engineers salary have to do with anything"

The engineers salary (actually total compensation, not salary) was quoted in the article.

"An engineer with a reported total compensation, or TC, of $400,000 a year griped about the responsibility of having to do a once-a-month delivery. “What the actual f—k?” the engineer wrote on the platform. “I didn’t sign up for this, there was nothing in the offer letter/job description about this.”"

I am suggesting that if you make $400,000 a year, one minor inconvenience a month by actually using the product you are working on really isn't something you should complain about.

What if the engineer doesn't have a car or a disability? That would be a problem, if it's not on contract

Addressed in the article you must not have read yet somehow have opinions on.

Nope, sorry, i read only the first part. Thanks for addressing that

That’s clearly not the issue at hand.

That they get paid a lot? And sometimes it’s to deliver groceries.

Getting paid a lot means you can demand anything now?

It's not exactly anything. It's actually using the product they are developing. Is that really so crazy?

While not exactly the same, we were responsible for developing an in-house app used by our retail employees. Every single engineer working on the app would every so often shadow one of these employees to see firsthand how it was being used in the real world.

Yeah, and they can quit and work somewhere else that’ll pay them just as well if they don’t like it. They’re not being asked to scrub toilets, they’re being asked to exercise the company’s product. The thing that pays their salary.

Let’s see how many actually quit over this. My guess is few to none.

> Yeah, and they can quit and work somewhere else that’ll pay them just as well if they don’t like it.

I think that's the main sentiment in those threads and the main response to that is "oh woe how dare you think like that, you prima donna".

Another thing to consider is one less dash, per employee, available for the dashers who might need the money. Perhaps there are not enough to be significant but if you dash by the office I bet you’ll notice.

One less dash hardly makes a difference, but you benefit from the engineer feeling the experience first-hand by dog fooding the product.

Multiply by the number of employees mostly saturated in a radius around the office. Perhaps still not a big impact but surely noticeable if that's your neighborhood.

400,000 / 52 = 7692.31 More like a month, if the dasher is lucky

Seems like a great way for employees to become more intimate with the product they’re developing. This, I would expect, would lead to developing a better product and user experience.

Seems like a great way to assist the engineers in just finding new jobs...

Seems like a great way to filter out prima donnas who can't be bothered to put themselves in the place of their users. I would be happy to be rid of anyone complaining about this policy.

It definitely is. Any engineer that won't eat their own dog food (if at all possible) is net a mediocre engineer, or otherwise immature (hasn't progressed far enough in their career; doesn't yet realize how much they don't know). DoorDash is better off without many of those types. It's an excellent way to clear the house of engineers that are more likely to hold back the product than advance it. Which of course doesn't mean firing everyone that complains; there are people that are rigidly unwilling and people that are uncomfortable (for various possible reasons) and need to be brought along.

In all of my experience, it has never been engineers misunderstanding the users. It has always, ALWAYS, been C-suites, Executives, and VPs pushing product features to

1. Entice investors 2. Planning for some external company event like a merger 3. Pandering to analysts 4. Pandering to the sales department, who again, sell into VPs and not to users

I think it's incredibly out of touch to suggest engineers are the ones that lack empathy for their own product, and not the myriad of individuals who haven't touched a line of code or a design file for the past decade. As an engineer I field support tickets every day, I know what problems the system has, no amount of me using the application is going to change what gets consistently submited through help desk.

You'll have to forgive me for being bitter after a decade of trying to innovate in the products I worked on only to be shut down every single step of the way by upper management; then to have to swallow this self-righteous PR stunt where the world seems to blame "out of touch" workers for not knowing what their job is and not having empathy for their users.

Beyond that, don't get me wrong, I don't think doing deliveries is beneath anyone. I would probably accept half a million dollars to do deliveries. However, I do think it's incredibly immoral to suggest that you can pay some people $200/hr to do deliveries, but others only $7/hr. Instead of focusing ire on those that are upset that their job duties are not what was described to them when hired; perhaps we can be upset that doordash has the capacity to pay their laborers much much more, and chooses not to in order to ensure they can pay half a million dollars for what everyone here considers "mediocre engineers".

Also, again, I personally would not work at doordash because of this requirement. However, I wouldn't have any issue with the company if they CLEARLY outlined this job requirement during their hiring process; which it looks like they did not do.

Article says ceo does dashes too.

Can you explain to me how someone who spends their entire day in k8s and helm charts would benefit by exploring the UX of an application through the lens of the user? Beyond a normal study, that is?

I've run every one of my companies with some form of everyone participating at a user level regularly. Either through support or using the product. Each type of role gets something out of it. A k8s engineer can achieve more understanding of their mission and connection to how their seemingly siloed activity has a real world impact while more directly product focused roles might gain actual product insight. In my experience employees who genuinely care about what the company is seeking to achieve do gain something from the experience.

If the application actually uses that k8s cluster, then you might actually want to see the real-world performance of it all instead of just aggragated response time graphs.

There's more to real-world performance than k8s cluster.

Totally agree. I wonder what other companies can do similarly.

As an engineer (product creator) - it behooves one to continually be involved in experiencing how the product is used. Not just how we believe it, or think it.

Use cases, flows, and product paths seem logical on a whiteboard. However, the real world often introduces nuances that significantly alter use in micro ways.

The objection to interact with customers from engineers I believe is more due to a social aversion, introversion, etc rather than a defiance of not doing the delivery. Engineers can often be afraid of unknown social situations - which is fair.

If I were the CEO, I'd have engineers go in packs of 3 engineers at first for 3-4 months. Then pair to group of 2 with one as a "ride along, no interaction" for moral support.

These engineers are paid incredibly well. If they can't handle spending 30 minutes once a month taking a delivery, then they can work somewhere else.

Yes, precisely, they will get paid extremely well somewhere else where this is not part of the expected job duties.

And in my opinion, the doordash product will suffer for it.

That is even better then. This will filter out the complainers and drama queens, leaving the employees who actually want to make a better product.

That's exactly what they said?

Has anyone considered that this is simply a viral stunt created by the gossip website (which I'd rather not name) in order to promote itself?

Even if the commentary isn't fake, it's just one engineer within a huge organization. For those upvoting this story, shouldn't our outrage/attention be directed elsewhere?

> which I'd rather not name

Why? It's the primary source where this was first "reported".

Many many DoorDash employees confirmed that this was really happening on Blind.

https://www.teamblind.com/post/Doordash-making-engineers-del...

I'm not disputing that the policy is real. I'm saying there's a possibility that the engineer's disdainful reaction to the policy is fake -- and done to stoke outrage and promote the gossip site.

> Why [not name the source]?

Because my intuition is that there's a high probability of this being a viral stunt -- in which case, I'd rather this site not exist.

This was absolutely a promotional piece.

They did this at chick-fil-a corporate as well. An extremely competent product manager described the yearly experience as “standing around not knowing what I should do and trying not to get in the way.” Those kitchens get hectic, super different than being in the office.

I wonder how society would change in general if we were all forced to step out of our comfort zones once a month? Make me become a garbage man for the day, or a barista or a gas station attendant.

Perhaps it would engender some empathy between different areas of society and reduce the overall sense of entitlement some folks seem to have.

Personally I’d like to see a return of mandatory draft or civil service. I think merging people across all boundaries to forge common experience could only serve to improve society.

Something I was thinking about where if I was the day-to-day employee suddenly forced to work with a brand new employee that you know will only be working there for one day (meaning they won't be super motivated to really dig in), I'd be pissed. Lots of comments in the thread about low morale for those engineers being forced to do this, but not much discussion about the reverse direction. (Makes since since this is a site focused on those engineers.)

This person claiming to standing around not knowing what to do just shows that they really are trying to do anything. You've been assigned a task, do it. If you don't know, ask. They aren't trying to be better because they know at the end of the one day their nightmare will be over.

I assume you were supposed to watch the employees interact with the technology, and ask them a few questions on their break or something. Trying to see if there's anything about the point-of-sale, employee scheduling, payroll, or other tech that seemed like a good idea back at Headquarters, but doesn't work well in real life. Or new ideas that would make things easier, etc.

That extremely competent product manager lacks imagination. I always learn things from watching users.

If anything the problem is I get too many ideas for things that could be improved and in reality you can't actually do a fraction of them.

But all this "extremely competent" guy gets out of it is "standing around not knowing what I should do"? What a joke.

When designing a dogfooding process try not to have 50 engineers do the same thing 12 or 120 times a year. For example, having engineers in kitchens fulfilling orders, having engineers perform store management tasks, and having engineers run a full day of deliveries will probably provide more insights even if done less frequently. I’m pro dogfooding, but put some thought into making sure the entire app is being dogfooded. Finally, when running a dogfooding program in your company be sure to communicate the importance of focusing on your customers and and cultivating customer empathy.

Speaking of dogfooding, let's not forget the restaurant's side of things. The DoorDash payout report is a $#!T show:

  TIMESTAMP_UTC_TIME, TIMESTAMP_UTC_DATE, TIMESTAMP_LOCAL_TIME, TIMESTAMP_LOCAL_DATE, PAYOUT_TIME, PAYOUT_DATE, STORE_ID, BUSINESS_ID, STORE_NAME, MERCHANT_STORE_ID, TRANSACTION_TYPE, TRANSACTION_ID, DOORDASH_ORDER_ID, MERCHANT_DELIVERY_ID, EXTERNAL_ID, DESCRIPTION, FINAL_ORDER_STATUS, CURRENCY, SUBTOTAL, TAX_SUBTOTAL, COMMISSION, COMMISSION_TAX_AMOUNT, DRIVE_CHARGE, MARKETING_FEES, DELIVERY_FEES, SMALL_ORDER_FEES, TAX_FEES, FLEXIBLE_FULFILLMENT_FEE, FLEXIBLE_FULFILLMENT_FEE_TAX, TIPS, CREDIT, DEBIT, DOORDASH_TRANSACTION_ID, PAYOUT_ID, TAX_REMITTED_BY_DOORDASH_TO_STATE, SUBTOTAL_FOR_TAX, DOORDASH_FUNDED_SUBTOTAL_DISCOUNT_AMOUNT, MERCHANT_FUNDED_SUBTOTAL_DISCOUNT_AMOUNT

What the heck is a 'TAX_FEES' anyway? You have taxes, or you have fees. Why use the terms 'CREDIT' and 'DEBIT'? Too clever by half.

What's worse is it doesn't add up correctly.

  SUBTOTAL + TAX_SUBTOTAL + DELIVERY_FEES + SMALL_ORDER_FEES + TIPS

  does not _always_ equal

  CREDIT + COMMISSION + DRIVE_CHARGE + MARKETING_FEES

Before DoorDash gobbled up Caviar, their report was different and terrible. Then they updated the report to what you see here, and it's better, but still terrible.

This is all-around a good idea. As a worker I would be happy for a paid opportunity to get away from my desk and dash across the city once a month.

The protesting engineer cited sounds like a bad team player who needs opportunities like this to develop character.

I remember one comment here in HN that mentioned how in the future society will be divided in those above and bellow the API. About the difference in quality of life of those that make the platform, and those that work for it, like the gig workers. It made me angry because was too cold, but in line with reality.

I really hope this type of strategies help in this regard. For the people developing those platforms, to remember that in the end those who work for those Apps are humans too, and not only a number in an A/B test or KPI spreadsheet.

Do you remember what thread that comment was in?

https://news.ycombinator.com/item?id=29609542

I once worked at a company that mandated everybody do help desk one day a month for a similar reason. Everybody thought it was a good idea including me. There was a 5 month span when I was working 10-12h days (yes it was a startup and I was young and dumb) to get something shipped. By the 3rd time of having to take a day to work help desk I was furious that on the one hand I was expected to work myself bald to ship a product but on the other they felt it was a good use of my time to take a whole day off the project every month. After this and a couple other incidences we stopped requiring it. This is to say there are certainly some upsides but there are also morale downsides.

I feel it has good intentions, but the realities just don't line up in my mind. Admittedly super negative outlook

The CEOs and executives that actually decide what the product IS will most likely either feign participation, or keep 'rearranging deck chairs on the Titanic'

The leagues of engineers and other relatively disconnected workers being encouraged to do this won't collectively Do Better.

They'll keep trying to keep their heads above water in their sprints/what-have-you, while a vague entity sets the direction

Let's do some math.

So, say you're paying me $15 an hour. Really, if you own a business by the time you pay for taxes, infrastructure, capital outlay and insurance you are really paying about $45 an hour for me to hang out and do what you tell me to do at varying levels of competency. You need to make that much more value from my labor per-hour or you are going to go out of business.

Say I have a piece of software that, because of a miscommunication between the developers who wrote it and the people who have to use the software (ideally this is what management is for but, lolz) has added 30 seconds to a transaction I do 20 times an hour. This means that, suddenly, there are 10 minutes of labor added to this task on aggregate an hour.

If you have a piece of software that wastes 10 minutes of my time, you just wasted $7.50 worth of my labor. But, of course, I'm using this software for 8 hours a day. So that is $60 of money wasted a day.

So, really, you thought you needed to make $45 an hour from my labor, but you just added $7.50 to the total. So, $52.50, and now you have a morale problem to boot: because the folks who are using the software are now accomplishing less than they used to for the same amount of work.

How are you not double counting your time?

Regardless of how you spend your hour, it still costs them $45/hr for you. If $7.50 of that is paying for idle time, that's still part of the $45, it doesn't get added again. If they could save those 10 minutes, that would just make you more productive and allow them to reach $45/hr easier.

That is a good point. You're still cutting out 1/6th of my productivity, which has to be quantified.

How would you quantify this?

It's essentially lost profits.

Let's say that your $45/hr number accounts for all the overhead. Just to keep the math simple here.

Let's then say that you generate $60/hr value for the company. That's $15/hr of profit for them right now.

If they could eliminate that delay, that's $10/hr of value you could be adding with that time. That's straight to their profits if they could eliminate it at zero cost.

Now, we both know that it's not zero cost. There's some cost associated with eliminating that delay. If the cost of eliminating that delay is more that $10/hr, it'll cost more to remove it than to realize it. And that'll eat into the $15/hr they're currently making.

We also know that while it'll cost now to fix the delay, at some point it will stop costing because the delay will be fixed. At that point, we get to realize the full profit.

Then it becomes a calculation of how long will it take. Because if it would take $11/hr for an entire year, then it might be worth it to suffer the loss in profits this year because the extra profits in the new year will make up for the shortfall this year.

And if you have multiple people using the app, the cost to fix it gets spread out across all the employees. With two employees, that's now $20/hr I could capture and as long as it costs less than that, we'll be good. Well, less than $30/hr because we aren't capturing that $20/hr until the work is done and we have to pay for the labor in the meantime and that means affecting current profits.

And of course, that scales. The more employees you have using the app, the more you can afford to throw at the problem and the more profitable it becomes to capture that extra. At just 100 employees, that's $1500/hr you can afford to fix this issue and you'll see returns of $1000/hr.

Thanks for this!

P.S. I have seen software deploys add 30 seconds to transaction before, it certainly has an effect on a companies bottom line.

Without experience, an engineers understanding of a problem space is suspect.

In other words, the difference between a good engineer and a bad engineer often comes down to experience and understanding.

My guess is the company is also taking note of the engineer's attitude and reaction to this experience requirement as well.

Ah, I remember reading about this idea in "The Phoenix Project" (or "The unicorn project") book, where all employees work in the shops for a week servicing customers, to get an idea of how the people work, who they are building the software for.

It'll be interesting to see how long it takes for someone ill suited to food delivery gets injured, robbed, assaulted, or worse.

I wouldn't do it. I wouldn't do it during normal times never mind during a pandemic. Seems like a bold strategy during the most job seeker friendly market I have seen in 20 years.

I made life choices so I wouldn't ever have to go door to door for anything. No offense to hard working delivery people but when I did that sort of work I chose to mop floors and clean toilets over anything food service.

DoorDash is a good idea in principle, but it's poorly executed. Among their competitors, Grubhub is superior in every way.

The reasons I dislike DoorDash and stopped using it:

- It has frequent outages.

- Has been the subject of controversies regarding tips.

- Does not have live customer support. They get back to you in 3 days.

Grubhub has none of the issues above, and they are always on top of orders.

Amazon Restaurants was another yet superior competitor to DoorDash but was discontinued as a service.

I have no affiliation to any of these companies, just talking from the perspective of a user.

Whichever app you are using, you are only considering how it affects you. This is natural, and not a judgment. Rarely does the user consider how it affects the restaurants themselves. I have yet to read a positive experience for the restaurants.

I notice that none of the comment think about scenario where the engineer doesn't know how to drive or doesn't own a car

Probably because the article addresses this.

> For employees unable to do deliveries, there are other programs in place to work with service employees and businesses

Interesting point!

I wonder if a good compromise would be for them to ride along with another driver?

The amount of people both here and at DD who think that the task is "beneath them" is appallingly large.

"Today you, tomorrow me" should be what everyone says to themselves when confronted with this sort of thing.

Yet another manifestation of the callousness of SF tech workers

One engineer who is getting nearly $0.5m total compensation and when asked to do one delivery a month, they are now throwing a tantrum over that.

But you know what? Good.

Maybe that is what they need since the drivers and riders get 0.01% of what these engineers earn and I'm sure that there were complaints about that. So when one is getting nearly $0.5m for the engineering work and told to do just ONE delivery a month, they are STILL complaining.

Just deliver a pizza to your friends house once a month and eat it and that would count.

Job done.

One of the issues I have in the whole software industry is the lack of empathy for regular folks and a lack understanding of how hard the average person has to bust their ass just to make it in life with just the basics. I personally believe it leads to products that are often disconnected completely from the people who actually use them and damaging to our social fabric.

I applaud DoorDash for this move and quite frankly this would make me more likely to take a job there.

> One of the issues I have in the whole software industry is the lack of empathy for regular folks and a lack understanding of how hard the average person has to bust their ass just to make it in life with just the basics.

And how is that industry issue?

My guess is they're talking about money.

Many parts of life are much easier when you're financially secure. And the challenges to navigate vary by wealth level.

Same can be said about any white collar job, why is that suddenly a problem here?

Don't want to speak for the original commentor, but I'd guess that the size of the income gap is very large.

> Just deliver one pizza

> Just work a bit more one day

> Just take a paycut

Nothing good comes out of mindset like this.

I'm pretty sure some engineers had ideas previously about how to improve the product, they just have no power to do so. It would be weird if it wasn't like that. I wonder if all this is just a way to create a massive quit

The crybaby rationalizations are not only not a suprise, they aren't even clever.

The very people trying to invoke the value of their time, are in that very process exposing how valuable it actually isn't.

> But a 1,500-comment thread on Blind

Nothing good ever starts off that way. Astonishingly sociopathic website.

That said, I do think that the media are deliberately seeking out the weirdest and most unpleasant employee opinions; I can’t imagine most employees of the company have an issue with this. And it does seem to be something they did before; they just suspended it for a bit.

Curious if you're allowed to leave early or come in late for this, or if they are asking employees to do it outside of their regular working hours.

I have seen a dozen adds from Doordash promoting becoming a “Dasher” in the last several days; perhaps they are just having issues with staffing?

Provided there's also a way to surface ideas that come from this experience, this sounds like a great move.

I have a hard time feeling sympathy for these engineers.

Reply on news.ycombinator.com