Hacker News Reader

Comment by keiferski | Hacker News Reader

The main issue with review sites, in my experience, is that I end up being hesitant to try new restaurants which don’t have a large number of reviews or even one really bad review. They may be delicious but there is always a hesitation factor; “Do I want to spend $20 and risk the food being awful, or do I want to go with the safe option, the place I’ve been to a hundred times?” This problem didn’t really exist prior to review sites, but somehow I don’t remember being all that bothered by it.

It makes me think: new restaurants would benefit from a sort of insurance program. I would be willing to try far more new restaurants if I had a guaranteed way to receive my money back in the event of disappointment. Just like insurance, the less times I use my refund, the lower rate I pay per month.

/just a thought.

View on HN

> “Do I want to spend $20 and risk the food being awful, or do I want to go with the safe option, the place I’ve been to a hundred times?” This problem didn’t really exist prior to review sites

Actually, it very much did.

The answer is that you suck it up and go to the new place--live a little.

I eat at a lot of different places. Normally the food isn't awful--most are just mediocre and I won't be back without some other reason (There is an Asian chain restaurant in London that was so BAD I can't even remember its name--that's one of the exceptions). Some are good, but too expensive for what you got--they go on my "If someone really wants <Food X>, I can take them there."

And then there are the few that are "good and reasonably priced (or maybe even cheap)". And the occasional "Expensive, but worth it".

Those are the ones you will recommend when your friends ask.

> Just like insurance, the less times I use my refund, the lower rate I pay per month.

Huh, what insurance works like that? If you pay a rate based on how much you used it, you might as well pay for the thing you're having the insurance pay for. I get that such an insurance will catch outliers still (saving you from a big boo-boo) but that's the only kind of thing insurances are for: big boo-boos (chance events that cost you disproportionately much compared to how much reducing the risk of them to zero would cost, like surgery or your house burning down). If you use it for frivolous stuff like, say, covering for a $10 meal, imagine the overhead, the management costs of all those claims, the arbitration, and the little you get out of it unless you order at a shitty restaurant and get unlucky nearly every time because you travel a lot for work and don't know the good places around -- oh wait, in your model that means you pay more anyway and it's not worth it, so every subscriber is really only lining the manager's pockets.

Just eat mediocre food for once or order elsewhere if the food is actually inedible. The lack of repeat custom is surely worth more to a restaurant than refunding those ten dollars you're sour about.

> Huh, what insurance works like that?

Yes. For example no-claims bonuses on car insurance means you pay less if you claim less.

> If you pay a rate based on how much you used it, you might as well pay for the thing you're having the insurance pay for.

No. For example say my car insurance is £500 or something. It goes down £50 a year every year I don't use it, due to no-claims bonuses. But it's still less than paying for a major accident out-of-pocket that could cost £5 million.

> Just like insurance, the less times I use my refund, the lower rate I pay per month.

Then the problem shifts to: do I want to risk my no-claim discount, or do I want to go with the safe option?

Sure but presumably the restaurants would be vetted in some way to minimize negative experiences, and the likelihood of you wanting to use your discount fairly low.

I’m not familiar with insurance companies but I imagine this is precisely their business model: minimize the likelihood of a claim but make the benefit of doing so worth it for the consumer to pay a monthly fee.

As far as I can tell, insurance only makes sense if the cost of the insured event would ruin you. Otherwise, you're better off just setting aside some cash into a savings account and calling it your insurance fund.

Insurers will always set premiums so that you're on the losing side of probability. Their real product is protecting you from the damage of going negative.

"ruin you" may be a bit strong. Take travel insurance.

Do I take travel insurance for routine trips that might involve me losing $200 for having to cancel a plane flight and maybe some hotels. Nope.

But if I have a high 4 figures trip that I'll have to cancel if I sprain my ankle or because a family member who is having health problems needs me to stay home? Or because I know there could be expensive evacuation if something goes wrong. That seems a pretty good use of a few hundred dollars.

Of course, in my case, this is pretty much the definition of adverse selection but maybe the relatively reasonable rates I can get are being subsidized by all the people who buy the insurance routinely.

So essentially, travel insurance doesn't protect "being ruined" it protects from making a shitty situation even shittier.

I think that's true of a lot of insurance for people who aren't on the edge. Travel insurance basically covers two scenarios:

- You've prepaid an expensive trip and can't go on it for some covered reason (injury, family situation)

- Or you're on a trip and have to be expensively evacuated/repatriated (pulmonary edema in Nepal/break both your legs in Paris)

For most of us who normally travel a lot, it's not about having to spend $1K out of pocket because luggage got lost/stolen or you had to do some complicated rebooking because of a delayed flight.

> Do I take travel insurance for routine trips that might involve me losing $200 for having to cancel a plane flight and maybe some hotels. Nope.

The cost of the flight and hotels is irrelevant though.

If you have a medical emergency while in a foreign country you could be looking at millions of dollars of costs to treat and repatriate you. That would bankrupt most people. Even if you're just popping to another city for a day - if it's in another country I think you really must be insured.

I do have health insurance (which one might need to get money back from after the fact) which does cover when traveling and good luck getting just about any travel insurance policy to pay out "millions of dollars."

Most of the time I'd be covered by my company anyway and I do buy travel insurance when doing activities with greater than normal risk. But I'd be surprised if anyone I know routinely bought travel insurance for foreign travel. I literally have never heard it come up in a conversation.

This seems like a fairly balanced overview of the subject: https://www.ricksteves.com/travel-tips/trip-planning/travel-...

> Insurers will always set premiums so that you're on the losing side of probability. Their real product is protecting you from the damage of going negative.

This is also a legal requirement as underselling the cost of the underlying risks can put your company at risk of insolvency.

States hate having insurance companies go bust in case of a real emergency so there are liquidity, zip underwriting and other preventive measures to keep companies kicking because the alternative causes the state to eat the costs.

There are also cases where insurance works by aligning incentives, to help make early intervention easier.

Take health insurance where you get free stop-smoking plans. It makes the barrier for signing up to such a plan much lower, which is good for you. It also makes the chance of people suffering from lung cancer lower, which is good for your insurer.

I think this is an experience common to enough people that it's frequently used as a prime example of the explore/exploit concept in machine learning.

I try new restaurants whenever I can. I wouldn't hesitate too much to order from a bad rating one if I don't see too many recent bad reviews and explanation. That said, I have been bitten by this far more than I would like. I apply multiple different filters now and then order. Ruined food costs you a good amount of mood.

What would make me try new restaurants?

Something new and novel. Can you make me something that the other restaurants can't? Do you provide ability to customize my dishes? Maybe call your customers with opportunity for big discount or food, tell them about your restaurant more and listen to their feedback. I recommend a single restaurant everytime someone asks me about food place here because I know their service is consistent and great. They don't care about reviews as much as my word.

As a thought experiment, do you have the same reaction when considering visiting a new restaurant in person?

I think so, yeah. I’ve often walked by a new restaurant and thought, “That looks interesting” but then end up looking at the reviews before I decide to eat there.

Reply on news.ycombinator.com