Webvan

> But as Google demonstrated you can enter after the market is mature [] and dominate.

I don't think Google (founded 20 years ago) is a good example of entering a mature market.

I think a better example is Facebook (founded 15 years ago) which won against MySpace.

I thought Salesforce (founded 20 years ago) should be a good example, but their market cap is still well below SAP's.

Where do you count as "the north" ? Plenty of delivery vans around in Manchester.

I think it’s more of an income thing than a location thing.

Most supermarkets will have a minimum spend and then there is a delivery charge on top. Both are pretty cheap but if money is tight then that is one easy convenience to go without.

I currently use Instacart (as an Express member).

I'd say 10% of orders (I order ~2 times a week) contain an item that wasn't of sufficient quality. Their customer support is great when this happens, always refunding the bad item (though sometimes that's besides the point -- I needed that item for my meal!)

The $100 yearly membership fee amortizes to ~$1/order, and the markup on products seems to be ~5-10%. So I'm paying a little extra to avoid going to multiple grocery stores on a weekend/weeknight

Is that actually true, though? I've seen anecdotal reports from certain Uber drivers or whatnot, but I have not seen data. (It seems counterintuitive that such a huge amount of contractors would continue to choose that work over traditional employment in service jobs if it paid less.)

Definitely. A lot of current startup darlings are shifting costs traditionally borne by the business to the workers. Uber is exhibit A here. Taxi companies, for all their flaws, at least paid for, maintained, and took on most of the risk of the cars. With Uber, all of that is on people desperate enough to drive for Uber. This strikes me as woefully inefficient systemically; large organizations are much better placed to manage cost and amortize risk.

Or you could look at the recent DoorDash tip theft controversy, one of many such blow-ups. Anand Giridhardas has a lively discussion of that here: https://twitter.com/AnandWrites/status/1153312792964935682

I feel this is a little too cynical.

For example, maybe I want to make extra money but don't want to be tied to a shift. This type of flexibility is amazing for lots of people. It's very appealing if you need extra money, and have a lot of time. Unfortunately, this aligned in an economic cycle where people needed work, and undervalued their value as workers(and for sometime bonuses, and gamification made this type of work very profitable i.e. 4k a week driving uber in SF).

The gig economy is more about having workers without giving them benefits and with making the workers shoulder the burden of shifting demand throughout the day. With food delivery you have the lunch rush and the dinner rush and outside of those windows your demand is a lot smaller. The gig economy means the company isn't on the hook for predicting or satisfying that demand; if there's too many workers for the demand, it's the workers who take the pay hit rather than the company. And if demand spikes unpredictably the company has an easy lever to push to get more workers out there (just notify workers of the spike, and surge pricing as needed to make the unplanned demand more attractive to the workers).

So the gig economy is really just auto scaling applied to human resources instead of computing resources.

This all sounds eerily close to data, which is what I'm looking for :) What is the source for your figures?

The market for delivery has evolved comparable complexity to mobile phone contracts, sadly:

https://www.lovefood.com/guides/3444/cheapest-supermarket-on...

So, you can use Waitrose, where delivery is free, but there's a £60 minimum order, and their prices are generally higher than the other supermarkets (Waitrose is the UK's high-end supermarket for moderately posh people [1]). Or you can use solidly middle-class Sainsbury's, where there's a £25 minimum order, delivery costs £1 - £7 depending on the time and day, but it's free for orders over £100. Or something else. Or sign up to a subscription plan which will work out cheaper if you get a lot of deliveries, etc.

[1] https://www.buzzfeed.com/floperry/of-the-most-middle-class-t...

As I said in my post, I normally pay £2. But the price is variable depending on time of day etc. I can’t recall what the minimum spend is but it’s way below our typical weekly spend anyway.

> So much value in learning first hand about the experience

I couldn't agree more! I'm working for a startup doing alcohol/cannabis delivery. Before there was funding, our founders rode their bikes to do these deliveries (and have still done some deliveries in the past 6 months).

There was so many lessons learned from doing our deliveries. One such example: we had a bug where our push notifications told the customer we had arrived, but were actually still pretty far away (GPS math sure is fun :P)

Yeah, this is a common entrepreneur failure mode. You need to be ambitious and visionary to start a new thing. And investors love to hear the grand plans! Tempering that is one whole other skill.

That guy sounds like he was trouble from the beginning. But if you get in this situation again with somebody less crazy, one thing I've done is to draw a graph with a dot just above the zero point and then a circle at the top right. I get them to talk about their grand 20-year vision, and I explain that's the circle in the upper right.

Then I draw a bunch of tiny line segments that walk erratically toward the circle. I explain that each week we need to take one step in the direction of the vision. The I ask them to focus real small, and think of the tiniest thing we can do toward the goal, something that fits in a week. Once the have a few of those, I'll ask, "How can we prove that this step really went in the right direction?" (It can also be useful to pull in the Lean Startup framework of validating hypotheses here.)

Sometimes that's enough to get visionaries to start breaking it down. And sometimes it exposes that the person is only good at dreaming big, but is unable to actually execute. Either way, it can yield valuable info.

Sure, grocery delivery is commonly available here too. But what percentage of groceries are sold through delivery? That was the real thing material to WebVan's business plan. They were betting the rate of change was much higher than it turned out to be.

And I'll note that WebVan was betting on replacing supermarkets, so they needed much more dramatic numbers. A supermarket already has 90% of business up, running, and paid for. WebVan spent hugely just to get to approximate infrastructure parity -- something that turned out to be totally unnecessary.

I see, thanks for the clarification!

56k modems were pretty routine by 1999 and ISDN (It still does nothing) was a thing. But by 2000, DSL rollouts were starting to happen.

it's quite sustainable i'd bet. banks traditionally make money by lending most of its deposits to other people and earning interest on that lending.

as long as the deposits are large enough, that interest should cover the fixed costs (employees, infrastructure) and the variable costs (paying out interest, referral bonuses) of providing the accounts. internet-only banks also lower their fixed costs by foregoing expensive retail square footage.

Hire too many employees, build out too much capacity without a market that can provide the necessary demand, operating costs get too high, and you're done. Of course if the terms of investments and the discipline of the founders are combined able to regulate the growth rate, these things become less of a problem.

If you expand too quickly then you might not have the sales to maintain the large growth. Then your profits from good markets go to propping up bad markets and before you know it you aren't making any money in any markets.

You can overspend on stuff. When the money runs out you have a business that can’t sustain itself. Uber is running that risk too.

In the same way one kills a software development project by having too many developers: you feel like you need to put everyone to action immediately, so you end up with a mass of disjointed, barely functional code.

Grow too fast and not judiciously. Expand rapidly, making assumptions about expenses and operations and revenue. Then a few months later, there were unexpected expenses, operations aren't going as smoothly as anticipated, and revenue is nonexistent. The buckets of money turns into buckets of debt (possibly actual debt, or maybe debt from the perspective of the investors). Then...not much you can do.

Companies grow because they hope more growth == more revenue == more profits, but it can easily go south. Fast forward a few months or years and they're stuck spending tons of money just to keep everything going, owing tons of money, but not making any money.

Because they scaled the business before they knew how to operate the business.

Edit: In order to scale a business, the founders typically need to operate the business themselves at a much smaller scale. Once the founders prove out the business model, the next challenge is hiring people and teaching them how to do the job. The same thing happens again as the initial hires become managers or take up senior roles and transfer knowledge.

Killing yourself with a bucket of money is basically hiring people to do jobs that you (a company) hasn't figured out how to do yet.

Lose money on every sale and hope you'll make it up on volume (or more charitably, that at a certain scale profitability will emerge)

If your business model is selling $50 bills for $40, it's not sustainable.