Indeed, it passed the annoying stage some time ago. Is it even a real usage? I've only seen it in press articles.
We took "unicorn" out of this HN title and will probably just keep taking it out of all HN titles until it finally goes away. One could almost write code to replace it, but I think not quite.
I'm currently consulting in the food delivery space, and I can tell you one thing -- don't believe it. It's certainly possible to create solid local businesses in the space, but not businesses at a scale that a billion dollar in valuation demands.
And when you add in the lack of barriers to entry for many of those companies that are in existing sectors overrun with competitors, I think you can cull the list by 50% right off the bat.
But I'm sure they are excited to see themselves on this list, as I would be, even if I knew our fundamentals that got us there were only based upon clearly unrealistic hopes and prayers for future users and revenue.
And here I was honestly hoping for a real 'unicorn' list. I should have known better.
What are the two biggest issues in food delivery? I imagine it's the production of food and delivery of it, right? (Funny written out, but I'm sure you get my point - labour costs.)
Staff to load automated delivery vehicles and the patrons unloading at the other end will eventually solve one half.
Half decent automated food production could solve the other. A lot of Indian or Chinese dishes (curries, stir-fries, etc) could surely be automated? Adding and stirring, portioning out into containers, etc? Could do many pasta dishes in the same way.
With the right technology, and a limited selection of meals, could you realistically staff a restaurant delivering out to hundreds each night with 1-2 people? I could see unpack-and-go setups making this industry easy to scale.
If delivered meals were decent and cheap, I'd get it a few times a week. At the moment the price puts me off - labour costs are high in Australia.
There seem to be a few different models, maybe worth iterating them (UK examples).
1) An ordering mechanism for existing delivery services (e.g. JustEat, HungryHouse). The problem here is the delivery experience is opaque, non-standard, and sometimes unreliable. If your food is late, you have to call the restaurant.
2) An ordering and delivery mechanism for existing restaurants (e.g Deliveroo). I think this is the sweet spot because it controls the entire UX, is able to offer a reasonable range of food, and is able to list higher-end restaurants so potentially expands the market for food delivery.
3) All-in-one service (e.g. Deliverance). I think the main problem here is that they're competing with restaurants, which in my opinion will mean they're only ever a niche thing.
My guess is that the scalable 21st century businesses will be in Type 2, and that food delivery (which could eventually be drone-based) and resturants (some of which may eventually use automation) will remain separate. I think the delivery & restaurant businesses seem to have different incentives; restuarants need to optimise for quality & value, and delivery businesses need to optimise for cost and breadth of choice.
That's why I think #3 may win out over restaurants which are sluggish to react. They'll own the whole process and will be able to optimise exactly what needs to be optimised. Restaurants have baggage.
If I just want half-decent food, I'm not going to be especially fussy. Speed and price will be huge factors. A restaurant might not feel comfortable running a curry-making robot out back whereas a food delivery service will be all about the result and whatever gets them there.
Firstly, what happens before all these automations happen. In that world there's no efficiency advantage over restaurants so a delivery service set up today has no reason to make its own food (other than to control the whole process).
Secondly, once we have those robots, restaurant churn is high. A Cornell study suggested 50% of restaurants fail within 3 years. To me that suggests that the switch between hand-made and machine-made food could be as rapid as culture allows. The factors will be things like the capital required to buy those robots, the cost of paying engineers to maintain them vs the cost of paying chefs, etc.
TBH, I'm somewhat wary of robots smart enough to cook food equipped with meat cleavers and rolling pins. I'm not sure the cost structure is the thing I'd worry about in that business.
I'll describe the robot I'm imagining. It requires power and a flat service (floorspace or bench). It makes an Indian curry in a large quantity and is managed by one operator who can probably handle 3-4 of these without breaking a sweat.
It has a large pot, a stirring method, and a few canisters above loaded with relevant ingredients. These arrive preprepared. Sliced onions, peeled and quartered potatoes, spice mixes, water, stock, yoghurt, pureed tomato, whatever. Each canister assesses weight. An app predicts demand (based on day of week) and instructs the operator as to the quantities required. He tips in the prepared onions, potatoes, etc - the required weight of each. The robot doesn't have to peel or chop anything. The ingredients arrive in bulk.
The machine is given the recipe (basically temperature graph and time to introduce each canister. Oil, stir, stir, onion, stir for x minutes, potatoes, protein and so on. Imagine a Thermomix that you can preload with ingredients.
The operator has loaded up one machine, then does another, and another. While waiting, he starts to prepare packaging. As machines finish and switch to warming mode, he serves portions to packages and loads the delivery method which has backed up to an open window/door nearby. The vehicle is like a post office mailbox wall with a warming method - it drives up to a house indicating the order is in a particular box. The recipient grabs their order and a server is notified.
The operator back in the "restaurant" wipes down the machines as required. The app calculating servings auto-places orders for replacement ingredients based on what was used.
If you have a sole driver, most of that could be built and solved now. We could make a set of warming drawers that sideload into a stock van. We could build an apparatus that sits above a giant Thermomix or KitchenAid-type appliance.
I think the keys to it all are not trying to solve the most difficult bits of each equation. IMO, those are: robots that can prepare ingredients, delivery from street to the front door, and a menu that includes things that are difficult to produce and serve en masse. This approach could still handle countless stir-fries, curries, soups, and a few salads (a Thermomix makes a half-decent coleslaw in 3 seconds - we do it all the time).
I live in China, and I am a user of ele.me - the food delivery service, and have worked with 17zuoye. 17zuoye is a solid business. They have worked for years to penetrate the K-12 education market, which, being public funded move at a rather slow pace, but once you get your foot in the door, you have a natural barrier to fend off new entrants. Ele.me is a rather curious story. They get to where they are by heavily subsidizing restaurants and customers. For all I can tell, they lose money on every delivery they make. and the moment they cut back on subsidies, people stop using their service. As a customer I welcome the subsidies (it works like a wealth transfer program, where the VC's are footing lunch bills), but as a sustainable business I am not sure where it could be headed...
100% agree on the food delivery space. I've told a few investors not to touch it. Basically it will parallel the restaurant industry at best. Lots of players without much margins.
Does that make twitter, amazon, and tesla still startups? I swear I'm not cherrypicking data. Those, plus google and facebook, are the first big companies I thought to look up P/E on. Public startup seems like a contradiction.
I agree company size is a bad metric, but there's something weird about calling some of these huge companies with thousands of employees startups (the unicorns, I mean, not amazon and twitter).
Was tongue-in-cheek, but that's why I used the word "first". All three have turned a quarterly profit at some point (Twitter only if you discount losses from stock-based compensation).
Twitter and Tesla have not yet turned on the money spigot, so you could argue they're still startups. Amazon is a grown-ass business that can turn the spigot on anytime it wants to impress investors.
I'm with you. It's ridiculous. I call hitting a billion dollars a good finish unless I'm wanting to gamble. A start-up are people with a plan, some money, determination, and maybe some customers/users. They're just getting started with no clue if they're going to make it or they've existed a while with uncertain business model.
The $10-100+ mil club are typically established companies still growing, still adapting, and with huge valuations.
A lot of the tech community is "magic" to the uninformed and uneducated. I think the term "unicorn" is a play on this sentiment.
There is nothing mythical about well funded startups having high valuations.
You don't stumble into one while galavanting through an enchanted forest.
You can't drink the blood of a successful SV CEO to prolong your life.
If you say "unicorn" in reference to tech startups I immediately assume you are a trendy idiot who's vocabulary doesn't extend past memes and buzzwords.
I've worked at several "unicorns", we all think you sound dumb.
edit: "you" is directed at the financial news community.
It's a misnomer, but is someone's attempt to define any tech company which is so far apart from the rest of the pack in terms of business practices and reach that they don't warrant being called 'yet another tech company'.
This usually involves under handedness, tax loopholes, and secret handshakes that 'normal' startups are too scared to execute on. 'Normal' startups have a market safety net to fall back on and can pivit anytime they want without much fanfare caused. Unicorns are not afforded this luxury because they are, well, unicorns.
