Hacker News Reader

Comment by aklemm | Hacker News Reader

While I like and use Instacart, the model is wrong. I am paying too much for a slightly unreliable service and the employees aren't getting paid enough. The stores should provide the shoppers, delivery should be a separate service, and the cost should be about a $15-30 premium for a large grocery order (15+ minutes of order picking + delivery similar to a cab ride). That leaves the ordering platform, which might be a nice niche for Instacart to provide across multiple stores.

Anyway, this has been a tough business all the way back to Webvan!

View on HN

I will say this - at least in my local bay area Whole Foods, which I know isn't instacart anymore; there are the online shopper order fillers, they work IN the store but clearly not FOR the store.

I'm going to sound like a huge snob, but it makes the in-store shopping experience much less pleasant. The actual whole food workers are always there for you and go out of their way to help you, I don't expect the online shopper order fillers to help me, but they at best DGAF about the in-store shoppers and at worst are hostile towards them.

They speed past people and around corners with carts, cut in line at the fruits and veggies, ask WF employees to help them find items while you've waited to talk to them and in return, flat out ignore actual in-store shoppers when they ask for help. If you're out of the tech loop and don't realize that they are incentivized to fill orders, you might think the online shopper fillers are part of the WF staff there to help you. Knowing this, it's not too jarring to me to see a customer go up to one of them and have their many 'excuse me's just ignored, not even acknowledged.

They need shirts that say don't ask me for help; I know the way they are incentivized to fill as many orders is what creates this behavior - that needs to be looked at as well in my opinion.

During my last few trips to Safeway and Whole Foods I estimated about 50-75% of shoppers were delivery app drivers. I know the SF experience isn't typical, but it was still mind boggling.

Resistance to it doesn't have to be snobbish either. The entire grocery business model today is designed for people shopping for themselves. Stores occupy the largest and most expensive retail spaces in the center of the neighborhood. There are wide aisles, fancy decor and complex layouts to showcase products and keep people inside. I love going to the grocery store without a list or any agenda and coming out with a new interesting dinner option, buying loaves of bread on impulse because they just came out of the oven and smell great, talking to the butcher about fresh cuts of meat, getting a recommendation for a new wine or cheese to try out. Like myself I imagine grocery shopping is as much a hobby as bare necessity for others as well.

If delivery becomes a norm then none of this will exist anymore (at least at an affordable price). Grocery stores will just be large warehouses outside the city optimized for serving pre-decided orders as quickly and efficiently as possible.

Now I finally what this scene [0] from the HBO show was about. As a European, I never understood!

[0] https://www.youtube.com/watch?v=CrDjRZ9AfQ8

The Brooklyn Wegmans is the same for what it’s worth, 60-70% instacart shoppers some times of day.

I was just thinking about this. You would think that some of the grocery suppliers would get wise to the fact that these customer bases need drastically different experiences.

A consumer grocery shopper wants like with like (including in multiple locations. e.g. I want chili powder in the spice aisle, and with the mexican food), lots of variety, things organized by price. etc...

A food order picker wants something that looks like an ikea warehouse. Pick items x,y,z from bins a,b,c in aisle 3. They could probably care less if like is with like, as long as there is an easy way to find substitutes. They want to work with a bunch of people that are 'on a mission' like they are. And not have to deal with people like me who read the marketing copy on each type of Newman's pasta sauce.

Hey - It's me, your friendly supermarket supply chain guy! (I now consult on this stuff)

Picking in store has a huge number of cost base efficiencies compared to using a 'dark store'. These are as follows:

* Reduced property costs and property overhead.

* Having 5 vans in 300 locations is better than having 300 vans in 5 locations from a transport perspective (you get much closer to the customer, or have to tranship through hubs).

* Depending on locale, store workers may have more relaxed terms and conditions and unionisation than industrial warehouse workers, which 'warehouse' workers may get dependent on setup.

* Every darkstore you have needs to be fully stocked. This isn't so much of an issue with long-life products, but is an issue with fresh products that need to be sold before their expiry. A darkstore has a 'minimum viable size' with these fresh products dependent on the range you want to offer. Bigger darkstore with more customers means a bigger range - but you are also further away from the customer so either you have a bigger stem distance or need to do trunking to hubs and tranship from there.

So it's not about getting wise, trust me we think about these things and will model what is the most optimum way to deliver, at least in the UK :) An incredibly rough rule-of-thumb of logistics in the UK is that 1/3 of your money is spent on transport, 1/3 is spent on warehouse labour and 1/3 is spent on fixed costs (building, rent, utilities, equipment) - so your saving on labour has to be significant to pull back the fixed cost benefit of running out of an existing retail estate.

This is in the UK - most supermarkets have their own in-store pickers and will deliver with vans in the back. We have one company (Occado) that has a few large DC's and will tranship to outbases with vans. Some supermarkets have these dedicated 'darkstore' facilities in super-dense areas (e.g. Tesco, Asda), but they don't work in most areas of the country and in most cases are actually more expensive to run than in-store picking operations (they are done for capacity reasons rather than cost reasons).

The advantages of the centralised approach tend actually to be about service. If I can centralise to a big enough warehouse, I can offer a better range and offer things like visibility of expiry dates.

Something tells me that online grocery shopping may evolve into this format- for the reasons pointed out, it doesn’t make a ton of sense long term for these two formats to be bundled once there’s enough demand for both (which at this point, it seems like there is). These seems to be what Amazon Fresh is doing.

Exactly. Like segregation of B2B and B2C traffic and route them through servers that have different SLA and capabilities.

I guess the current model where couriers shop at the retail store is a result of hacking up an MVP. It’s very inefficient at scale. Once the grocery stores and retailers deeply integrate with the delivery companies a more efficient workflow will emerge. Something closer to what you just described.

I personally think this deteriorated the in-store shopping experience. There were too many folks running around - it's a different vibe. Never felt that even when stores were packed with regular customers. In addition, some stores blocked out quite a big area within the store for fulling these orders as staging area.

I wonder if they could have that staging area in the back (warehouse or what ever it is called), which wouldn't directly impact the in-store shoppers.

My local WF used to be a nice place to shop before the acquisition. But once Amazon took over, the place turned into an assembly line! It's like you were stepping into a turn-of-the century factory (well, turn of the _last_ century). I just stopped going there.

For me, other than the online order filler folks, I've been surprised by how other little change I've noticed. Aside from promoting kindles in locked glass cases and for some reason ice chests, I hadn't noticed a change.

Yah, Amazon changed the vibe of the place. It's gotten a lot more visually distracting with the visible blue prime stuff everywhere and all the loud reminders for prime rewards and other marketing peppered around the store. Different physical presence too with the Prime Now bagging/staging areas, and the Amazon lockers too.

+1 on the stores providing the shoppers-- I've used Instacart a few times in Washington DC and I'd say it's more than a little unreliable here. I've had such amusing results as:

- Four bunches of ~7 bananas instead of 4 individual bananas

- A single cherry tomato instead of one pound of cherry tomatoes

- Candied walnuts substituted for fajita seasoning

I'm not overly upset about it, but the main problem is definitely the quantity issue-- several different drivers have purchased absurd quantities that seem like something one would obviously question and double-check. Or that something would be flagged when the price is so far outside the expected price on the app. I would think this would be totally solved by store-provided shoppers though.

I loled at "Candied walnuts"

Agree with this. The premise is nice but the execution and shadiness of raising prices so you don't really know how much you are paying for the instacart premium feels bad.

You'd think this business would be going through the roof during the pandemic. If it can't work now, how is it ever going to work?

Well, just to cite one factor, it might work better under conditions where we aren't literally putting people's lives at risk by asking them to shop instead of us shopping for ourselves.

In other words, conditions in which the pool of labor is less heinously and obviously exploited and better workers are easier to recruit for the job in question.

While I agree with you about that, I don't think most people have that logic when they consider ordering on insta-cart. If demand for insta-cart is dropping I would wager that doesn't bode well for its use in the future regardless of the level of talent/exploitation of the employees.

In other words, if all of a sudden insta-cart was much more generous with their employees and managed to hire better talent, that wouldn't likely translate into more demand for their product. Yes places like Trader Joes treat their workers well (and build brand around that) - the lions share of demand revolves around on merits/price. I wish it was more along your factor but sense that it isn't a factor with an order of magnitude change associated with it.

This is how it works in Texas at HEB and Central Market. It's about $10 to do curbside grocery pickup, more around $15-$30 I believe for delivery, which is done through Favor, a delivery subsidiary of HEB.

HEB is $4.95 for curbside, if you want same day or next day times. It is free if you book two or more days out.

Delivery is $5, with the same $4.95 same or next day surcharge as pickup. It is a great deal.

Instacart has a heavy presence in HEB, at least in Austin.

True, but IME it is almost always a much better deal to use the HEB/Favor service for HEB delivery or pickup since the per-item markup is low and fixed for the former, plus it provides and allows for coupon use.

This is how grocery delivery works in the Netherlands. The groceries generally aren't delivered from individual stores but directly from the regional distribution centers. Delivery costs range from free to €10.

Wouldn't Kroger save money by doing direct delivery? Groceries could be shipped direct from the distribution center.

Not really. It still needs to be broken down from large lots to smaller lots. My local store, Metro, uses their physical locations as a delivery staging area - they do large-lot breakdown and distribution as normal, but for an online order, they put it on the "online delivery" cart instead of stocking. This means all deliveries need a 24H advance notice, but it's done for a flat $12 fee for to-your-door delivery, with items at in-store prices.

Once they have all the logistics in place sure, but even for a company like Kroger, which is on the larger side, it's a huge undertaking. Once you tack on liability and other issues, it's a huge project just to get going.

People going to Kroger solve the last mile problem, picking, and substituting for free. These are all expensive problems to pay for solutions on a low margin business.

It sounds insane to me that Instacart even exists.

I guess Instacart was a trailblazer? I don't see why supermarkets wouldn't offer this service and pocket the difference.

In the UK major food supermarket do deliveries or local collection. They're prepared by the supermarket staff in their warehouse, without having to browse the supermarket. They're quite reliable (you order online and you have to pick the products manually) but they will apply substitutions sometimes.

Collection is free if you order over a certain amount.

Covid messed up things (queues got super long / expensive) but overall it's a well working model.

Having used instacart (through Costco) and walmart grocery delivery almost exclusively for the past 10 months.. I'm more amazed at how cheap it is, and I'm a bit worried that employees are not getting paid enough.

Costco is probably the more 'reasonable' one. They mark up the products probably 5-10%. So this one I feel that the delivery people are probably being paid ok.. although my nearest costco is 16 miles away (about 25min).. so it kind of feels like a steal to be able to get delivery from them for what feels like not a whole lot more.

Walmart is closer, but there are no markups.. and I pay whatever $10/mo for free delivery from them. Order multiple times a month.

I'm honestly not sure I'm going to go back to the stores as often even when things are normal.

> Having used instacart (through Costco) and walmart grocery delivery almost exclusively for the past 10 months.. I'm more amazed at how cheap it is, and I'm a bit worried that employees are not getting paid enough.

Are you really? What have you tried to modify this as a result?

Many gig workers have their tips taken away or obscured in some cases, as in the lawsuit settled by Doodash recently, which is most of their total income from these things: the algo sets the base salary as close as the lower minima for it to get picked up.

Have you ever considered just paying them more outright: as in leave an envelope at your door with the sum of money you think they're worth? Or simply asking them to contact you directly to be their personal shoppers if things were done to your satisfaction? I saw the former more often, with token sums like $5/10 by some well intentioned people but really the latter would be the more impactful solution and had been done prior to this for the affluent/exec class long ago.

Just so it's clear, the markup on those apps isn't going to the contractor's pay, its their to generate revenue for the core business to pump its valuation with fluff and prove its MVP has the potential to reach profitability... someday in the undetermined future. The gig worker gets as little as the market will bare most times, the app is optimized that way, and often has to be 'stacked' to make it even worth their time or will remain bouncing around in the limbo if no one picks it up.

I spent a lot of time looking at Doordash's business model during the pandemic while I went back to school or supply chain management after having done supply chain and logistics in the Auto Industry. I even did it myself to understand it's implications as a 3PL supply chain solution ahead of its IPO last year and it was super eye opening.

These things are horrible and are specifically targeted to the ever growing under underclass of Society. I'd say its borderline predatory given how much they gameify it in their favour.

> I'm honestly not sure I'm going to go back to the stores as often even when things are normal.

And quite frankly these apps/platforms are relying on you doing exactly that to justify its business model and will operate at a loss to do so in order for a massive IPO valuation and more VC money.

It's really quite sick, but it seems almost inevitable at this point given how many businesses have been shut down during this pandemic.

If I had the time and resources I would really want to do a talk at the Chaos Computer Club on it, as I think it really needs to be exposed for what it is amongst the Technorati who help built this monstrosity. I saw a Software guy who wanted to get the analytics on rejected offers from various regions with an app that tracked the acceptance ratios and other metrics, so the data exists out there somewhere.

Reply on news.ycombinator.com