Hacker News Reader

Comment by bombcar | Hacker News Reader

You don't have to build a search engine that is better than Google's - you may only need to build one that is roughly in the running and keep it around until such time as you have to divest from Google.

And working for years on a project that may never see the light of day can be hard.

View on HN

> You don't have to build a search engine that is better than Google's…

You're right, of course — most companies would be happy with a "perfectly fine" DDG-like offering. I'm presupposing that "best" is Apple's bar for an M1-like slam dunk launch, since an underwhelming launch (think Apple Maps 1.0) can be hard to recover from.

> And working for years on a project that may never see the light of day can be hard.

For sure! Moonshots require a level of faith that I'm not sure could maintain for a decade either.

I think, much like maps, a search engine requires an iterative development process. I don't think you can reasonably launch either product (search nor maps) and surpass the quality of Google on day 1 because there is so much user feedback on both products which you can not capture without actual users. Apple, with Maps, has shown an ability to stick with things and iteratively grow them to be really great (Apple maps IMO now far surpasses google maps and except for some specific data layers it's far and above OSM).

It's a great point. Apple Maps was pretty rough when it launched but they've stuck with it and iteratively improved it. It's not like this is a self-driving car where you have to get it right and any mistake from the get-go can be fatal, so there's no reason not to apply the same iterative approach to search.

A colleague who trusted Apple Maps 1.0 a bit too much once found himself driving towards the edge of a cliff because his actual destination was at the bottom of that cliff down a mountain road. So yeah, maybe not at the same level as self driving cars but Apple Maps 1.0 could have been fatal for people less situationally aware than my colleague.

Tesla certainly didn't feel the need to get their Fake Self-Driving right from the get go.

Apple Maps are plain useless. Google Maps are bad but bearable. OSM is miles ahead Apple Maps which is surprising because they could just pull data from OSM and not being that bad.

Is this because of your area? Apple Maps has been tangibly improving over the last few years.

I assume that's because of my area. Doesn't make it any better for me. It doesn't even have names of villages around (which OSM does have).

That's very interesting, where are you located? I'm in north NJ in the US.

I'm in Kazakhstan, Astana. City itself is not that bad, but surroundings are bad.

Nobody's tried Brave Search?

I pretty much only use brave. It seems to give me more relevant results.

> I'm presupposing that "best" is Apple's bar for an M1-like slam dunk launch

Well Apple has been iterating Ax Series Chips, M1 is no different to previous A(n)X Chip. i.e M1 wasn't really a "launch" for the M Chip. And Search could easily have been something else in the mean time before a "Google" launch.

Apple already does a lot of "Searching" with Siri Suggestions. And arguably it is the 2nd largest search engine on the planet.

> Well Apple has been iterating Ax Series Chips, M1 is no different to previous A(n)X Chip. i.e M1 wasn't really a "launch" for the M Chip.

Through an engineering lens, yes — one can look at the M1 SoC in a vacuum and characterize it as a logical evolutionary step in Apple's 20 years of silicon design. What made its introduction shocking (and its launch a slam dunk) is that Apple leveraged its experience in silicon design, OS design, hardware design, emulation, etc. to create new products, based on a completely new architecture for the desktop/laptop market, that definitively trounced their Intel-based predecessors.

You may have seen the clues before the M1 launched (which are crystal clear in retrospect), but for most people the unveiling of the M1 was as much a surprise as learning the identity of Keyser Söze.

> Apple already does a lot of "Searching" with Siri Suggestions. And arguably it is the 2nd largest search engine on the planet.

Exactly! What will be "Apple Search" exists in a nascent state today, making its eventual launch as a branded product very analogous to the "develop secretly in plain sight" strategy Apple used for the M1.

The money doesn't come from search quality, it comes from selling ads and Apple is a long way away from building up ad inventory and an ad salesforce like Google's. They absolutely are making money on ads today, but they're a long way from swapping out Google.

Yeah, I suspect part of it is keeping an emergency option if Google were to not only stop paying them but actively blocking them (it's unlikely to happen, but if you have infinite money you're going to be paying to reduce unlikely risks).

Apple will start using their own search tools in smaller places that they can monetize better, I suspect, things like "Siri find a good restaurant".

There’s plenty of other search engines eagerly waiting to take googles space on ios, should google decide to block ios users

First impressions matter. Has Apple Map's reputation caught up to the reality after the initial launch was underwhelming? Apple might be trigger shy.

General consensus in this thread seems to be it's better than Google now, and for me (at least) I never bother opening Google Maps anymore; I did often early on.

That may be true in the US, but for example in Berlin, cycling directions are still unavailable and POI search is still behind google maps. I would prefer to use Apple Maps, but it’s lacking.

There’s also no way to share a route with someone else.

Although apple maps has had the better client for quite a while, it can still be whack sometimes in actual directions.

I have not found this to be the case in my couple of years since I moved to Apple and would never have guessed Google was better at some point.

Google maps had been a pain point coming from a pretty rural area, Apple does not have this problem.

Rural is a huge issue and Apple surprisingly seems to actually respond and fix issues - I reported one once and it was fixed the next time I bothered to check, so good on them.

Google still routes you to the middle of a cornfield there, and I don't know how to get that fixed.

Very much dependent on the country. Where they have rolled out their “new” maps based on their own mapping (and not licensing a mix of others like TomTom), they are much much better than people think.

https://www.apple.com/au/newsroom/2021/12/apple-rolls-out-al...

In Australia, I rarely go for Google Maps anymore. Apple Maps search and driving UI is cleaner, the LookAround streetview equivalent is great, and spoken directions are more natural. Google Maps still has the edge in business location data though - like finding “a barber nearby”. As more people use it, it’s clear Apple Maps is catching up there too.

But when I was in southern Europe, which I guess haven’t been upgraded yet, I found they’re a long way behind Google Maps in terms of routing, detail map data, and updates to business locations.

Apple's turn by turn instructions are 10x better than Googles.

With them working so much on building up their ad business, having a search engine and switching the iOS default search to it might make up for the loss in Google money they get right now.

Just like with Google Maps / Apple Maps it doesn't seem very Apple-like to depend on a partner this heavily for such a core feature.

Reply on news.ycombinator.com