Best Practices for Indexable Progressive Web Apps (PWAs)

[engine]

Google has published a guide to best practices for indexable Progressive Web Applications.

It's a relatively comprehensive guide detailing both best practice and a series of do's and don'ts.

It talks about crawling for indexing, clean urls, canonical URLs, device types, and progressive development. It goes on to mention schema, Open Graph and Twitter Cards, testing with multiple browsers, and testing site speed and performance.

It's worth bookmarking the guide.
[webmasters.googleblog.com...]

[ergophobe]

Some excellent resources there if you follow the links down the rabbit hole... too busy for the next couple of months, but after that? It looks really easy to make some simple PWAs.

[thedonald123]

A big draw with PWA's would be the ability to send push notifications like a native app. Is anyone using notifications for a Content style website? Seems pretty intrusive. I imagine the churn rate must be high.

[iamlost]

Progressive Web Apps (PWAs) will likely become quite the 'thing' over the next few years. However it is critical that webdevs not misconstrue why Google has got behind the idea over the past year: it is the other shoe dropping.
* the Accelerated Mobile Pages (AMP) Project is their response to Apple's News App and Facebook's Instant Articles - under guise of improved user experience build a defensive walled garden of content for ad delivery.

* the heightened push that to truly have an impact, it's important that they’re [PWAs] indexable and linkable. Incoming protocol HTTP/2 provides several new things including 'push'; W3C's Web Applications Working Group has been developing Service Workers, which became a Working Draft in June 2015; etc.

The reason that Google is so solidly behind PWAs is that typically web apps are search black boxes and that is a threat to their very foundation.
Note: from OP linked article:

...they’re one of the most exciting new ideas on the web. But to truly have an impact, it's important that they’re indexable and linkable. Every recommendation presented in this article is an existing best practice for indexability...

Google desperately needs apps to be indexable. And so they are pushing PWAs, which are coming regardless, in order to maintain access; access that is largely missing with current mobile apps despite enormous efforts to be let in.

Devil's Advocate:
If one has a website and it is indexed in Google (and other SEs) then it will bring in a small percentage of return visitors (nav/brand queries) but mostly new visitors. If such a visitor loves the site enough they might decide to download an app and come directly; in essence apps are the web variation of a loyalty card. Given that these people have already found and liked your content...why let Google (or anyone else) be MITM and interfere with that one to one relationship?

That is not to say that they must not or even should not be partially or wholly indexable in SEs. It is to say that being indexable is NOT a must or even a should requirement - and most definitely not a best practice. Whether and to what degree one allows a SE access for indexing purposes must or at least should be a business decision of the app publisher just as it is with their web sites; and the two should not simply be accepted as one and the same.

It is quite possible to have a PWA that is opaque to SEs, it may even be desirable.
Until Google or anyone else can show why their being a MITM of an established relationship will 'truly have an impact' that significantly outweighs the hazards/negatives...sorry, but a good many of those so called best practices in the linked article quite simply are not, imo.

[ergophobe]

The reason that Google is so solidly behind PWAs is that typically web apps are search black boxes and that is a threat to their very foundation.

Absolutely. Apps are a threat to Google and many others. Dries Buytaert, founder of Drupal, recently made an appeal for "the open web."

Ultimately, if you can get users to install them, the "search black boxes" solve the problem of your competitor being one click away. Once I have the Expedia app running, it's much less likely that I'll open up the Kayak app (I don't even know if there is one, but just by way of hypothetical).

To argue with the Devil's advocate...

So if you run a website and you are not a huge player, the closed web is as bad for you as it is for Google and in the same way that PWAs give Google a tool to fight the app black box, be it Amazon or Expedia, it gives that same chance to webmasters. Face it, our small businesses are almost never going to motivate users to download our app for the one-off purchase, but a PWA gives us a ghost of a chance against Expedia or Amazon and their apps. That is, if the web experience is as good as an app, there is much less incentive for users to lock themselves into an app.

The app download "loyalty card" you mention works against the small merchant. It works in favor of businesses that you will buy from repeatedly (Expedia, Uber), not one-offs (Sunset Motel, Joe's Taxi).

Despite all the moaning about how Google loves big brands, I would say the opposite is true. Before 1995, your ability as a small merchant to attract a national audience was essentialy zero. Big brands owned the market. Then we had a decade or so where access was *really* easy for the small guy. And now we're in a world that's halfway in between. As a website owner, that's a problem. As a consumer, though, we're in a pretty good place (I remember when I would want to get to the brand website for a product and it would be on page 3, which was great for affiliates but horrid for customers).

Relative to 2005, Google has made it hard for the small guy. Relative to 1995, Google has made it easy.

The move to apps is, at least in my industries, a much bigger threat than the increased competition within Google (not to say that isn't a problem - it's a big scary problem - but I can see things I can do there, whereas short of PWAs I don't see how I can compete against an app).

[ergophobe]

>>MITM

Planet Money just had an episode on the degree to which we're becoming a MITM economy. Uber doesn't produce anything but a connection. AirBnB doesn't produce anything but a connection. EBay, Expedia, Lyft, Kayak, VRBO, Etsy, Amazon Marketplace and so forth.

So just to be clear - I am very sympathetic to that part of your post too. I just look at the main businesses that I'm involved in and *most* sales come through some MITM who skims off a percentage in exchange for connecting me with the customer. I have been working hard just to keep things rolling without having time for broader effort, but after January 1, my situation will change somewhat and one of my goals is to work harder on direct relationships.

Again, I see PWAs as a tool to help with that. AMP as rolled out is a disaster - if I copy and paste the URL, it's a Google URL. It basically reduces your site to a Twitter card, but on Google. You can't share the source URL (at least I can't figure out how with AMP new articles).

PWAs should be wholly ownable on your own server rather than served always out of the Google cache.

[NickMNS]

@ergophobe I you referring to this planet money episode?
[npr.org...]

[iamlost]

ergophobe:
Note: my bold emphasis.

So if you run a website and you are not a huge player, the closed web is as bad for you as it is for Google and in the same way that PWAs give Google a tool to fight the app black box, be it Amazon or Expedia, it gives that same chance to webmasters. Face it, our small businesses are almost never going to motivate users to download our app for the one-off purchase, but a PWA gives us a ghost of a chance against Expedia or Amazon and their apps. That is, if the web experience is as good as an app, there is much less incentive for users to lock themselves into an app.

I'm sorry but I really do not see how giving Google (or any SE or other third party for that matter) access to apps, including PWAs, offers small players a chance to fight back against huge players that is not already cooked into their site. And if the site is truly that awesome that they come and then return enough to consider getting your app... sorry, it just doesn't compute.
Note: on a related note I have absolutely no idea why sites, with the visitor locked in clicking to download their app, are almost always referred offsite to Apple or Google app stores. Idiocy to the n-th.

Of course no one is going to download an app after a one off purchase (eCom) or read (info). And I wouldn't want them to. Of course if my business model is solely Google (or other SE) new traffic and singular purchases... sorry but I refuse to owe my soul to the company store.

I'm fortunate in that I'm a decade ahead of you in the 'going direct' revenue business. Part of my direct ad space sale agreements is (for example) often some variation on coupons/specials for a companion affiliate relationship. And that is a very tasty hook for my info sites to tempt return visits and proffer an app to make getting and using the latest and greatest limited time offers simple, easy, and painless. 3% to 5% of my visitors are coming via app - and those very frequent very loyal customers/visitors are all direct all mine.

The trick to competing in niches dominated by behemoths is pretty much as it is in the B&M world when WalMart comes to town: do NOT go head to head; go for quality, value, and nibble the edges and holes.

Yes, if the web experience is as good as the app (and I like to think my sites and apps info and UX are comparable) a good many people who return regularly can't be bothered; only about a third to half of my sites' frequent return visitors come by app (although they do mostly come much more often) but some app methods are different, some find simpler. Just as some folks prefer one browser over another, so too some folks prefer an app over a browser or vice versa.

[iamlost]

For those interested and not averse to reading:

* What, Exactly, Makes Something A Progressive Web App? [infrequently.org] by Alex Russell, Infrequently Noted, September-2016.
Note: he (a Google dev) and his wife Frances Berriman (freelance designer and front-end dev) coined the term 'Progressive Web App'.

* Web App Manifest: Living Document [w3.org], W3C Working Draft 10 November 2016.

* Progressive Web Apps: The definitive collection of resources [dev.opera.com], Dev.Opera, 11-November-2016.

Note: Important iOS Caution:
* Don’t use iOS meta tags irresponsibly in your Progressive Web Apps [medium.com] by Maximiliano Firtman, Medium, 07-June-2016.
Note: Apple/iOS/Safari are still only considering several aspects that go into PWAs. For the next while at least full features will be primarily Android without mucking up iOS. Progressive, right?

[graeme_p]

Note: on a related note I have absolutely no idea why sites, with the visitor locked in clicking to download their app, are almost always referred offsite to Apple or Google app stores. Idiocy to the n-th.

I do not understand that. getting them to install the app is what locks the visitor in, and the only way most people can (with IOS) or will (with Android) install an app is from the Apple and Google app stores respectively.

I really do not see how giving Google (or any SE or other third party for that matter) access to apps, including PWAs, offers small players a chance to fight back against huge players

Because it will encourage people to use Google instead of going straight to the app, and if they start from Google it levels the playing field a bit.

[ergophobe]

@NickMNS - that's the one

@iamlost - see @graeme_p's answers with respect to app stores (exactly) and levelling the playing field

Because it will encourage people to use Google instead of going straight to the app, and if they start from Google it levels the playing field a bit.

Exactly. If people go straight to the Amazon app or Expedia app, I have NO chance to get in front of those people or, in one case, I do have a chance to get in front of them, but Expedia is constantly putting reasons in front of the user to have them book somewhere else. On Google, I at least stand a *chance* that they will come to my site - be it PWA or responsive or whatever - and I have a chance to "speak" to them without interruption.

If users get used to the web being an experience as pleasant and effective as an app, it reduces their desire to go straight to the app.

Now, I do understand that much of this is based on a reticence to spread their credit card numbers around everywhere and PWAs in and of themselves don't solve this.

[ergophobe]

PS @iamlost - that's a great reading list. Thanks! It would take me a month to follow up on the links on the Opera page, but those are excellent articles.

[iamlost]

graeme_p, ergophobe:
I understand your point about the various lock ins that make the two app stores the easy solution. I do say, however, that however difficult Apple, browsers, and some ISPs have made app downloads directly from sites there are perfectly legitimate methods around one, two, and in certain edge cases, all of those hurdles.

Every click step in a required sequence loses roughly 10% of an audience. Just that simple click from one's site to an app store has consequences. At some volumes those consequences can be significant.

And that is without even considering that apps aren't only for mobile. Granted, few desktop users, either by percentage or absolute number, download apps but some do, given the opportunity and sufficient perceived value. Progressive Web Apps will not (if one is sensible and forward looking) be a strictly mobile benefit and it is worth remembering that some apps (not PWAs) are already by design and the sue of existing prior attributes workable both on and offline.

Further, it is important to remember that while the PWA hoopla has been and is rising, Apple and Safari are not yet on board. And while Android is popular iOS is generally by far the better revenue source.

As to the benefits of allowing SEs access to one's apps, that is a business decision. In my circumstances the decision is absolutely not; I can, however, see the logic in your reasoning and trust that it works as well as you hope.

[ergophobe]

>>trust that it works as well as you hope

Thanks! And certainly your larger point is the important one. Everyone has a certain set of circumstances, challenges, visitors and so forth. It's up to all of us to figure out how to best address all of those and there is no magic bullet. What's essential for one person is a waste of time, possibly detrimental, for another.

But the nice thing about a forum and some healthy debate is reading those other points of view that allow us to pool experience and see beyond our own limited circumstances.

[NickMNS]

Here some more great info presented by John Muller about seo and web-apps.
[youtube.com...]