Crucial insight, there. From the perspective of an observer who has no special affection for Apple, it's surprising that a company of that competence would succumb to such a basic process error.

0 - http://www.theatlantic.com/technology/archive/2012/09/how-go...

This just amazes me.

For the longest time there were really only two suppliers of data: Tele Atlas and NAVTEQ. Everything else wasn't worth touching b/c the quality sucked. Then Google collected their own in their little cars and stopped paying suppliers. I'm not sure why Apple didn't have the foresight to understand this was an enormous engineering effort from Google - not only collecting their own data but the whole platform itself.

I worked as a consultant for one of the two major data suppliers for 3 years rebuilding their backend. 400 years sounds like a reasonable swag.

He lays out very well that a human element is required to bring these streams of data together. And Apple is not a company with the DNA of big data.

Just algorithmic manipulation of the data is not possible or sufficient - they will need an army to integrate these streams and bring them up to par. And Apple is even less happy to deal with an Army than Google was.

It's just that it is hard and that it will take Apple a lot of time and investment to bring the mapping functionality to a better level. I'm pretty sure Apple knows that.

It's just the first iteration. Apple works that way. Bring a product or service in a first iteration and then improve from there. This is how the iPod evolved, for example. Aperture, another example.

The current maps application in iOS already has a feedback interface. This helps to improve the data.

But there are a few things which need more consideration:

* the 3d view looks ugly when looked at close to the objects. The 3d reconstruction algorithm which creates a 3d view from images is problematic.

* the angle of the data from Tomtom is for car users. Other users have less benefit: there is a lack of detail and the usage perspective renders the map in a certain way. For example here in Europe there are a lot of local public transit users. They have a hard time identifying useful informations on the current maps.

* combination with all kinds of POIs. You need to get that data and have it constantly updated. Where is a shop, when is it open, where is a museum, where are interesting views, where is a difficult road condition, ...

Probably the mapping domain is the toughest Apple has touched in years. You need a really good idea how to deal with the challenges. Personally, I think it is worth it, but it will be a lot of work (and not of machines, but also of humans) and very expensive for Apple. I'd wish they would use more of Openstreetmap and that there would be a benefit for the Openstreetmap community.

All things considered, given how complex maps must be to implement, it seems like Apple did a pretty good job for day one.

Furthermore, this furore reminds me of the storm over Siri. Tech pundits work themselves into a frenzy proclaiming that Apple is losing its edge. Average consumers however pay no heed and the company rolls on to the next product launch largely unharmed. The critics miss the bigger picture: a company with so much momentum that it can easily afford to crowdsource the refinement of challenging big-data projects such as Siri and Maps. While the critics stand around prophesising doom, Apple iterates, improves, and by the time the next big hit comes out the last "disaster" is ancient history. Ignore their strength at your peril I say.

Nobody, including Google, consistently delivers functional map data to mobile devices every time, at least not in the Atlanta metro area. I travel there with some frequency for various soccer tournaments as a referee or parent. It is not uncommon for one or another device to provide faulty routing to the people involved in a match. Sometimes it's Garmin. Sometimes it's Android. sometimes Bing.

People don't primarily buy smartphones for the maps any more than people primarily buy smartphones based on call quality. Apple knows this. Apple sells phones because of iTunes and brand positioning.

The issues with maps didn't even garner a comment among the Apple fans in yesterday's Facebook feed. The edge cases among iPhone users that will be lost over poor quality maps is more than offset by the Genius's sales pitch about how easy Apple's map application is to use.

Reminds me of the days when Microsoft used to release software that customers would call "beta-quality" and wonder if they were being used as unpaid QA.

I wonder how long it will take for the outside world's expectations of Apple to drop to what they should be in the permanent absence of the gaze of Barad-dûr?

This could be a very interesting case study on sales of an otherwise great product being hurt by a move that cripples a critically important element of said product.

I've already met several people with iPhone 4's (or older) who said they are not upgrading to either iOS6 or iPhone 5 because of the mapping issue. I am part of that group as well. I'll have to buy an iPhone 5 for development purposes but I don't think I'll have it as my primary phone until the whole maps business is sorted out.

Mapping seems to be one of those things that you can't design your way around. In other words, nobody cares about beautiful inaccurate maps. This could be one of the first challenges on Apples's desk that can't be solved with cute commercials and pretty design. It has to be good and at least equal to, if not better than, Google's offering.

Regrettably sometimes the only way to get good at something is to start doing it. At first you'll probably suck at it but, with time and effort, you'll get better and better. This is Apple getting on that path to excelling at mapping. It'll take time. There's no doubt that they have the financial resources to make it happen. Now it is about execution.

Apple lacks the ability to mine vast amounts of local search data, as Google was able to do when it started its mapping project.

Does anybody know what this means? Don't the queries I make go through their servers, and isn't my location a relevant parameter for those queries? I don't see why this can't be mined.

EDIT: Here is an archived version found elsewhere: http://www.allhatter.com/showthread.php/13017-Google-Maps-am...

This is ridiculous in the year 2012. Nothing about this should be a headache (apart from people keeping their data secret for various reasons) yet it's still bafflingly hard.

And they can improve, as Google did. But customers are very forgiving of mistakes and iteration when your product is the first and best (so far); plus, Apple's key branding is quality.

> Apple lacks the ability to mine vast amounts of local search data, as Google was able to do when it started its mapping project

This is an interesting competitive advantage Google has that is really really hard to beat. Similar to what Facebook has. But... perhaps Apple also has data? e.g. from Siri queries? They can certainly gather it now, now that their mapping app is being used.

I just wonder why fix something that ain't broke and make it worse!

Google must be pissing themselves laughing. Just as Apple get their act together google will probably wade in with a killer app? OR will they? Perhaps they'll not bother and use it as a marketing ploy to push Android?

Fascinating article.

When you try to compete with an entrenched product that is feature rich and highly polished (such as google maps) you either need to differentiate your product or come to the table with an equal level of polish. Apple has failed to do this. They have made a classic "strategy tax" blunder.

This move is straight out of the playbook of the old, bad Microsoft and the old, bad IBM. It's the sort of fundamentally bad idea that a big, dumb, lumbering company makes. If there's anyone in the world that is happy about this it has to be the people on the surface team at Microsoft. Because it shows how massively vulnerable Apple is to competition now. Whatever magic Apple used to have, it seems to be gone.

Sure, all software works this way. We all know the "release early, release often" mantra. The difference is that Apple replaced functionality that many users depended upon with something completely inadequate. It'd be like Sun or IBM going in and replacing all the super specialized custom microprocessors in your server farm with experimental souped-up 386es. Sure, they're the same type of thing, and hypothetically they should serve the same function, but you don't just go out and replace completely functional stuff without at least the basics firmly in place.

It is clear that Apple released iOS Maps very prematurely, mid-development. This was stupid, and there's really no use splitting hairs about it. Some bugs or problems are to be expected, but a critical function that most of your users depend upon, that has real, serious consequences if failure occurs (lost in a dark alley, etc. etc.) ought to be better than what Apple has put out. Add to this Apple's draconian policies that forbid Google from offering "Google Maps" as a third-party app on the App Store (though I guess today it looks like they may be waffling a bit on this), and you have a very serious problem.

I guess we can say that it may be good for Android adoption. :)

From the article

1. Completeness – Features are absent and some features that are included seem to have erroneous attributes and relationships.

2. Logical Consistency - Expecting the data across different sources to be completely consistent ie 'An example of this could be having a store’s name, street number and street name correct, but mapping it in the wrong place'

3. Positional Accuracy – is considered the closeness of a coordinate value to values accepted as being true

4. Temporal Accuracy – particularly in respect to temporal validity – are the features that they map still in existence today?

5. Thematic Accuracy – particularly in respect to non-quantitative attribute correctness and classification correctness.

The first iPod didn't do a lot, but what it did, it did correctly. If the first iPod had had the scale of quality issues we're seeing out of Apple Maps, there wouldn't have been a second.

And why exactly should Apple care about this? Isn't the vast majority of their profit coming from higher then normal margins on hardware? Their bread and butter is suppose to be making compelling tech products that have high build quality and are easy to use for any type of user. Why would Apple care where ad dollars are going as long as they kept doing what they are suppose to be doing to justify their high margin devices? Apple isn't in the ad business.

Of course, with their map move they've just compromised the quality of one of the core services of a mobile device, that should be a bigger deal then any of this nonsense about location based recommendations being the future of ad revenue.

I don't know the particulars of the agreement they had with Google and other providers, but couldn't Apple have just rolled out their own driving directions application with restaurant and location search services but kept Google on for the time being? This would have allowed them to launch, then roll out a fully fledged Maps application of their own.