I got yelled at for telling people that :-) Or more seriously when I worked at Google a third party came to me excited about a new Google service that I knew was like the half baked 20% project of a new employee and getting little traction. I suggested they hold off on committing until it was more fully baked. Long story short it got back to the guy and I got scolded for 'sabotaging' a project (which vanished without a trace 9 months later when the person moved on to a project that didn't encourage 20% time).

I completely understood the idea of 'trying' something to see if it would work, but worried about the difference between the project "inside" the plex vs the view of it outside the walls. Personally I think that bringing some more discipline to that was one of the better things Larry has done as CEO so far.

That said, payments systems at Google have always been really really weird. Weird in the sense that they tend to 'force' a user support infrastructure which has historically been anathema to Google product deployment. People will put up with "Go to this forum and ask your question, maybe another user can help you out." for a lot of things, but when it involves actual money, not so much. Not to mention the rules that come along with handling money. It's really too bad since Google could be a force for great good here, instead PayPal continues to own the space. (I'm rooting for Dwolla and Square though, they are showing great strength here).

> but worried about the difference between the project "inside" the plex vs the view of it outside the walls. Personally I think that bringing some more discipline to that was one of the better things Larry has done as CEO so far.

Wasn't that the point of Labs? To allow experimental projects to get used by people outside Google?

Yes, as far as I could tell that was part of the Labs mission (this was between 2006 and 2010). The confounding factor at the time was that "labs" was one of about four different ways something went from idea into "production" (aka visible outside Google), a tremendous amount depended on the project manager you worked with. There were people who had been at Google long enough that they could get an idea from first successful build to production in a month, knocking down all the various stubs of process that were in the way (like Marissa's UI review). These folks were well regarded by the fact that they got a lot of stuff done. Pretty much anyone, with the right set of people, could ship nearly anything. The downside was that sometimes things that came out were not as baked as they might appear to someone on the outside.

As someone previously burned by this business model, let me be the first to tell you how right you are. I built a small business around their SOAP search API some seven years ago, and received a commercial license and increased API usage to bring my product to market.

After easily lining up enough customers to exceed my daily limit, I lobbied for higher usage and a method to share the revenue (they had no structure to charge for API calls back then). For years I tried to contact anyone I could within the organization, but there was no response. After an eternity, they simply shut it down with no notice.

You just never know how long they're going to enjoy experimenting with their APIs, and unless you have a backup plan, it's a deal breaker.

Yeah, but how many business was using Google Checkout as only method of payment? This decision will hurt only those businesses. Rest of businesses using Google Checkout will remove one of payments method. Not a big deal, and we may guess that this method was probably not so popular already.

I had a client that used Google Checkout as the only method of payment. It's not rare. It is a mom & pop (truly) web site selling products online.

Indeed. I did a similar setup years and years ago for a mom and pop shop (well before Braintree and Stripe) and I had to choose between Paypal and Google Checkout. Glad I set them up with Paypal.

I feel the problem lies in the fact that Google shuts down whatever service they deem not "successful". Because of this, I'm starting to weigh my options differently when Google services are in the picture, and I don't think I'm alone in this.

That being said, any other services could be shut down via an aqui-hire, or other reasons. Maybe this is the prevailing problem of small time developers like myself, and there's no other way around it apart from getting your idea bootstrapped with these services, and hopefully by the time any of those services gets shut down (if ever), you'll have a migration / backup plan of some sort.

My startup for a time being was only using Google Checkout. Here lies the problem. Can't trust Google with building APIs when they reorganize.

This is surprising to me, but I wonder if Google Maps could close down....

> I wonder if Google Maps could close down....

No. Google has bought what must be dozens of companies at this point just to support and improve Maps (Waze being the latest), it's reportedly profitable & popular, with the continued rise of mobile I'd expect it to become more important, and my little survival analysis (http://www.gwern.net/Google%20shutdowns) gives it 87% odds of surviving to May 2018 (which IMO is too low).

Also there are a bunch of government people who work on it I think.

I saw a bunch of public linkedIn profiles with NSA analysts and people in the private sector with google maps listed as experience for whatever reason

Google maps and Google earth are used extensively for imagery analysis and intelligence work. They aren't working with the maps team so much as using it as a tool.

That page is always really interesting when it comes up, but have you considered adding data for existing product improvements by acquisition? You include a variable for whether it was initially acquired or not, but perhaps big acquisitions clearly for the purpose of improving a product are a strong signal of longevity?

That's not a bad idea, but at this point it would be hard to go back and be able to classify each product by how many other products were merged into it. That information is not always clearly laid out publicly (to say the least).

The more relevant question is not whether Google Maps could close down but whether it will change in a manner that adversely affects you. For example, with their new vector-based system, will Google phase out tiles at some point? There are some apps that might be incompatible with vector-based data, so this could matter. Or more relevantly, will Google restrict API access? Think of Twitter -- it's never closed its Tweet stream, but it has changed the rules for API usage, to the detriment of certain Twitter clients.

As a (for context, Android) user, changes to Google Maps have already adversely affected me. I (until very recently) lived in NYC.

The best way to tell what direction you were facing when you came above ground from being on the subway was to look at the direction of traffic aligning with the arrows on Google Maps... but they decided that was no longer an important map feature and got rid of it.

I would frequently be zoomed into a specific location, and relatively soon after need to switch contexts on the OS (to check an email, answer a text message, etc.). When I returned to the app it used to be that it would retain the zoom level, but after updates it seems that Google thinks it's useful/funny/cute to zoom back out to the full zoom of the overall trip. For the life of me, I can't imagine a context in which this is useful.

They've got a monopoly on mapping data, and insist on taking a great app and making it worse. I can't understand why.

Umm... Google maps has a compass... which rotates with you as you turn. Just look at your icon on the map, at has an arrow facing the direction you are.

Well, no. Some devices have a compass in them, and Google Maps uses it if possible.

If your device has no inbuilt compass (think iPhone 3G, from memory), there's nothing Google Maps can do about it.

Oh, my bad. I actually thought it was a feature of the GPS, not a separate sensor.

To be fair, that does require your device has a compass-sensor.

They recently got rid of topographic data for the mobile Maps application. For serious stuff, relying on your phone is a bad idea - you need a real map - but for casual walks, having some idea of elevations was really handy, and now it's gone.

Mapping is such a core part of the mobile OS experience I can't see them shuttering it. Separate from that, they use it to show ads and promote their other services (like Google Plus).

I doubt they'll kill it, but would you be confident the APIs will be there in a few years? Between Apple and the mass defections to OpenStreetMap et al when they raised pricing, their third-party use is not what it was.

That's insane, maps collects huge amounts of data, google's raw material.

But you'd have to assume that Checkout also collected lots of raw data about people and their purchasing habits. It wasn't as if nobody used the service, it just wasn't the most popular.

Not enough data to be useful though. If they can't make behavioural predictive models of you to make ads & sell to the NSA it's not useful

> I can't imagine ever building something that relies on a <any company name> service at this point.

Seriously, if you don't build it or own it, you can't rely on it. SaaS is convenient, but it's a flawed business model if you depend on any startup/product lasting forever. If you want to last forever, you have to innovate for yourself or buy the people you will depend on.