What the report is based on: a median value selected from the traffic to 42 of the top US websites (it ignores those which don't have both and app and a website) from data provided by Comscore.
The term "traffic" is never defined. Web requests? Quantity of data transferred? Both of those would have significantly shapes depending on whether the client was a browser or an app.
It's also interesting to note that the top 10 "Browser Reach Advantage" websites are basically text based websites - news sites, wikipedia and blogs.
Things I learned from the report: the top website (with 207.0x Browser Reach Advantage) Blogger has an app.
A website has write access to my screen, my speakers and my internet connection. Plus limited read access. To keyboard, mouse, device orientation etc. Other rights can be granted (location, cookies etc). It can do all this only as long as it's "open".
An app usually has read/write access to large parts of my device. The whole sd-card, raw internet access, the camera, the microphone, ask me for money etc. And it can keep doing so even when it's "closed".
We will see the two converge. Websites will have more rights they can ask for. Apps will be better to tame.
But trust will keep being the differentiator and they will keep co-existing. There is friendship and there is marriage. Both have their use case.
I think you are right. And I'll add that on smartphones that are in the low- and mid-range segment, mobile version of websites are often way better than mobile apps.
For example on my smartphone (a Wiko Cink Slim), I uninstalled the Twitter and Facebook apps because mobile.twitter.com and m.facebook.com are much quicker and responsive than the Android applications. They also do not request as many permissions, and are more featureful (e.g., Facebook chat works on m.facebook.com while it isn't included in Facebook's app, and rather requires an additional very intrusive shitty app).
> An app usually has read/write access to large parts of my device. The whole sd-card, raw internet access, the camera, the microphone, ask me for money etc.
I wouldn't so firmly say "not anymore" when only 15% of Android users[1] are on the latest version of Android almost a year after its release. That number probably won't go any higher over time either because Android's update cycle is a mess.
By "that number" I meant the percentage of users who're using the latest version of Android. Lollipop's numbers will eventually get higher, but by that time there'll be 2 or 3 newer versions of Android.
No, but new phones come out all the time that are STILL using older Android versions. I think it will take around 2-3 years for all new Android phones to come with the latest Android.
It's more likely that sites became better at supporting mobile browsers and mobile browsers just became better in general.
This all leads to a better mobile browsing experience, javascript runs better on mobiles now, and site developers have learned what works and what doesn't, sites aren't Flash sponges anymore (which was the case even for sites with Mobile support 5 years ago), HTML5 and other W3C standards have taken off, the mobile screens even on low-end devices are usually similar to desktop/laptop screens in resolution.
People rarely care about their privacy, they auto accept pretty much anything that an app requests, Apple's lockdown of the permissions might help that case a bit, but again not because of privacy but because of the hassle of having to go into the settings and enable individual access for apps, this doesn't make people care for their privacy just annoyed and can't be bothered.
Content distribution and companion apps made sense when you needed to go mobile 5 years ago, when redoing your entire site to support mobile was too expensive and even then it wouldn't be a good experience due to how poor mobile browsers performed back then.
Now it makes absolutely no sense, it's another thing you need to support and worry about which costs money, it's much harder to make people download an app than press a link, and in the current "link-content-sharing" (e.g. outbrain) eco systems most sites operate in it makes very little sense since allot of the traffic is generated by those click-baity "25 reasons why Node is now considered harmful" links you have at the bottom of each TechCrunch article.
And on top of that even a low-end phone will usually have a 1080p(or close enough) screen, 2GB of RAM (unless it's an Iphone ;)) and a quad/hex core CPU which is more than capable of running pretty much any browser served content today mobile or not.
This is true at the layer of "why mobile browsers are capable of competing and beating app usage today"... the ante to get in the game...
But, the reason people are using browsers more than apps is simply because search is becoming the main use case of a smartphone.
I never look for an app on my phone beyond my home screen anymore, I just search or voice search my phone and expect it to either direct me to an app on my phone, or a search results page in Google.
I dont care where the solution to my need lives, I just want the quickest, easiest access to that solution.
Could also be the case, tho it's a bit hard to tell since you don't know how they've accessed the content, however if mobile browser would blow like they did only a handful of years ago that wouldn't be a viable approach.
Content access apps like TapTalk are pretty much dead, the only big community i know that still cares about their app is XDA-Developers which well is a very niche community to begin with.
That is such a nerdy perspective, totally stuck in the filter bubble.
I don’t think that explains the difference at all. I think it’s about convenience. People don’t want to use apps for the diverse things they do on the web all the time. Apps may be convenient if you use them frequently (and then only marginally, especially with improving speeds and more people paying attention to making websites usable on phones), but if you use something not so often?
I don't think it's a totally unique perspective, but perhaps the average smart phone user would use different terms. If you look at apps as always having new updates, and always using up precious phone storage space you begin to feel like these apps are cumbersome. It might not be what they call trust, but it is a negative experience regardless.
I have observed that non-tech people find the App Icons sitting on their home screens to be very convenient to access a service they use, kinda like bookmarks.
Sure, I don't think there is one variable that explains everything. There are a multitude of reasons that lead people to decide to use or not to use apps.
I agree with your point. Do you think it could have been phrased differently though, since some people might take offense at their perspective (and by implication, them) being called nerdy? (Obviously not the most dire of insults, and not an insult at all to many people - yourself included, probably - but still probably unpleasant to some.)
poeple especially don't need apps of 150mb+ each, and premium phones come with limited storage forcing costly upgrades on top of an hefty price.
the only competitive advantage for apps is push notifications, and the sooner browsers and phones can build that into a standard the better (including opening the right page when clicking on a specific service notification)
> the only competitive advantage for apps is push notifications
Don't know about that. What about UX? It's pretty well accepted now that the native controls in apps provide a much less frustrating UX than all but the most meticulously designed web apps, which is a very low percentage right at the top. That's just talking about controls, that's not even to mention the clunkiness of the browser itself (address bar popping up over content, popups offering to remember passwords, all that kind of stuff that detracts from your tailored experience) and to an extent the navigation model (typing in a url rather than tapping an icon for instantaneous checking of a service). Simple stuff like this really matters to the majority of people.
For technically-minded people, we're not as distracted or put off by this 'minor' stuff because we understand the trade offs, reasons for it, depth of it and even how to work around it. Most people don't, they just (probably rightly based on evidence) trust apps to deliver a better and more tailored UX and that's it, that's what they like and want more of.
Don't get me wrong, I love the web, for probably much the same reasons you do. One advantage it definitely has is the ability to avoid downloading apps in the first place. But its competitive disadvantages vs native apps go way beyond push notifications. There's are a lot of things so deeply baked in to the web and browsers that apps do so much better that it will take (software) generations to 'fix', if at all.
Unfortunately for the web, as seen in the last five years or so, these are disadvantages that really, really matter to the majority of people who actually use services. Much more so than the higher level, arguably 'more important' things like security, privacy, storage space, etc that us techies tend to prioritise.
The other thing is privacy. I don't want to give every service a "full access" into my phone. At least I have a feeling, that my privacy is in better hands, if I access services using up-to-date mobile browser (or Tinfoil for Facebook) instead of a native app running on my phone.
Looking at the ridiculous number of permissions different apps would require to run on my (Android) phone, using mobile site of the most of the services seems safer choice. Android 6.0 may change this with switchable permissions, tho.
you are right about privacy, but if you follow the thread the parent already makes a good point of user already having been desensitized about app security permission by them being confusing and only really understood by few
I agree that the current browsers represent an exceptionally good model to run remote applications securely, far better than the app store / play store 'sandbox' and 'vetting processes' model, since has been used since forever and tested again loads of different attack vectors.
but I don't know if this is the reason for non nerds not to use native apps
It went away in 4.4 but I have it again on 5.0. I had to use Xposed to get it to work on 4.4 but now it's on again by default. I see people complaining it's not present in 5.0, though. Strange...
uhm while WebSQL and IndexDB aren't mature enough, wrapping libraries that grant some cross platform offline storage do exist and are only getting better
iOS allows apps to be pretty user-hostile, too. A bunch of apps require location tracking 'always' or 'never' but without an option for 'while using.' Why would I want the Yelp app to know my location when I am not using it?
This comment makes no sense to me. The "While Using" is a user setting / choice in the Privacy options.
It corresponds to kCLAuthorizationStatusAuthorizedWhenInUse authorization status in the SDK.
If an app tries to use any of the background location API without the user having previously authorized kCLAuthorizationStatusAuthorizedAlways (and I believe the user is prompted again in a few days after the initial authorization by iOS8+) the app API calls are denied access with an error code representing unauthorized.
It has been my experience that adding the "Location updates" background mode capability indicating to Apple that your app requires background location sets off closer review during the app store acceptance process.
I have had apps rejected that a reviewer felt did not have a good enough text "reason" for requiring location tracking displayed to the user in the message box.
There are many reasons for a Yelp-type app legitimately to know about your location while not in use. It could make better recommendations, limit searches to "local" sites, etc.
None of this of course stops iOS itself from "spying" on you. The CLVisit seems to be always recording but experts with more knowledge on that topic than me might know how to turn it off.
I just checked this, and on that page some apps do in fact only offer totally disabled or Always. Seems like an oversight that could be fixed with a software update though.
> An app usually has read/write access to large parts of my device. The whole sd-card, raw internet access, the camera, the microphone, ask me for money etc. And it can keep doing so even when it's "closed".
The Play Store and installing APK's shows you a full list of permissions that the app will have access to before installing. If you're not happy with it, just don't install the app.
> The Play Store and installing APK's shows you a full list of permissions that the app will have access to before installing. If you're not happy with it, just don't install the app.
That's a shitty way to do things. You should be able to revoke access to things the app is asking and still use the (then crippled) app. There are tons of apps who ask way more access rights than they need (obviously not related to their core functions).
Only for apps that use the new API level, unfortunately. Unless they do what cyanogen does and fake the data exposed, which I guarantee Google won't do.
It's going to be extraordinarily confusing for users who end up exposing their personal data, thinking that an old-API app will be constrained when running on their new 6.0 OS.
Where? I don't see that anywhere in my settings. I'd consider installing the Facebook app again if I could yank all of its privileges. I'm 99% certain that this is NOT an option anymore, post 4.4.
I'm on 5.0 and it's under Settings > Security > App Permisions. I can select each permission in each app and switch individual permissions from the default of Allow to Deny or Ask. I'm on a stock ROM.
I have a phone with a clear 5.0 ROM, beyond a couple of Vodafone-branded shortcuts on the homescreen. Apparently this feature isn't present in Nexus ROMs, which is strange. Good on Vodafone for including this feature though.
For years it has been obvious that the web would beat apps in aggregate, but apps would have better engagement. If you wrap your website in an app using ionic and treat it like a standard site no one will use it. While soneon has noted trust (which is true) it is also fir convenience. The browser is a great app and can run many apps in parralel while quickly shifting between them, it is CONVENIENCE as mych as trust.
If your app does mot extensively make use of the hardware and can be run as a website, it has no business being an app. This is obvious, being proved now, and the app craze will pivot from shit-tier websites in iOS clothing to things that add real value and interface directly with the phones hardware because they need to provide security, access to a protocol, or leverage hatdware (accelerometer, camera, Bluetooth, etc)
I believe that mobile browser will beat apps, like it did on desktop. Mobile browsers were unusable, sites were not mobile-friendly, devices were slow with tiny screens. Now those factors change and browsing websites on modern smartphone is much more usable. And it's likely that we will see shift from mobile apps to mobile websites in the near future.
I try to find an app for things I regularly (daily) use. Some things have excellent apps which far surpass the web experience for me (like Reddit, Linkedin, Inbox, Skype) while others are crap (all HN apps I tried, including the paid ones, Hipchat (what a piece of...), almost all airline apps).
But when the experience of the app surpasses the web, it does so by such a stretch that when I start the web version on my desktop I close it again and take out my iPad. For me a company definitely only takes it's customers serious if they have smooth native apps as well as a smooth responsive web experience.
As a developer and consumer, I think that browsers are better for most use cases, where user interaction is simple enough and web technologies are more than up to the task.
But this technical consideration could be trumped by business considerations: making people pay for "mere websites" is more difficult than making them pay for apps, and if ad blocking really becomes mainstream, this could be enough reason for producers of even simple content to prefer apps - where they can have their free version with ads without having to worry about blockers.
At least for most usable apps with ads there is the in-app option to switch them off. I don't mind that for an app that is good ; on the other hand, I would've paid for that app as well in the first place.
That... or maybe it could remove all professional content from the open web, and make it some kind of ghetto for those who can't pay for apps. I honestly don't know.
You just described most of what I see on 4chan and Facebook. But in a sense what you're saying has happened to a degree with paywalls. The really good content at say the New York Times, Financial Times and the like is behind paywalls at this point while low quality stuff like Huffington Post is free.
> The really good content at say the New York Times, Financial Times and the like is behind paywalls
For now, it's "soft" paywalls - they're still indexable by Google, and visible if you come from there.
I'm afraid that, in the near future, they could really become accessible only to those who can pay. And the quality of the free internet would instantly drop many notches.
That's my fear as well, although I've seen a decline in the market over the past few years. After the banner ad crash of '08 it became harder for indie bloggers to make a living. And another problem I see these days is that not only are banner ads paying badly, but money has flooded into the field to back winner-takes-all ventures like Vox or Buzzfeed. So it's hard for the little guy to even get pizza money at this point.
Desktop apps for an average user were always a wild west with potential viruses: there was never a curated app store with reviews, which I believe is an advantage for an average user.
Additionally a large class of top phone apps is messaging, which has always been apps even on desktop: Instagram, WhatsApp, Facebook, Snapchat, Skype etc.
Also, why would I use maps.google.com on my phone instead of one click Google Maps app? Seems like torture to me.
> I believe that mobile browser will beat apps, like it did on desktop.
So now I have to "upgrade" my browser (which is basically almost a VM) that uses 1GB of RAM to run that To-do app over a series of tubes which are metered and analysed.
It's pretty great that this happened. I'm so sick of using these pesky desktop applications. Just remind me, which web apps can I use for the following?
Now of course I expect we'll here the "but it misses feature X". I'm sure that's true. Next, someone will point out that generally, 80% of people don't need feature X, and the fact that it is online means it is many times more accessible. We've seen this over and over again.
Remember when everyone thought web based email couldn't possibly work for them because of feature Z? I'm sure that's still true, for a few people.
So can we skip that whole discussion?
But I don't know if the message we should get out of this is that the web wins, or than new ways of distributing functionality to more people cheaper wins.
Let's be clear - I was never suggesting the the web can't beat native. The op was stating that the days of native are behind us. It is strictly untrue.
> Remember when everyone thought web based email couldn't possibly work for them because of feature Z?
No, I've always used web-based email clients. Did a software developer really say this?
Back in the early days of webmail, when IE6 was new and we had 28.8k modems, someone almost certainly said that. 20 years ago we used Eudora, PINE, Netscape's mail client, or (if you were really unfortunate) Outlook Express.
POP was horrible, IMAP made things much better, but the web clients had already won.
A modern desktop client is much better than a modern webmail client when it's available. It's interesting that desktop clients have been continualy improving, while web clients have been continualy getting worse.
As far as the number of products used and time spent are concerned it is. Most of the growth in the software industry over the last decade or so had been on the web.
They are built with flash, which is NOT what I would call using web techs. The fact that you couldn't tell the difference is interesting. Despite all the flash hate, there are valid use cases for it.
Those are all very specialised types of apps for content creation. If those are the only examples that came to your mind, you pretty much proved the point.
So, you really believe that nobody, other than the use cases above, use native software? I only listed the software categories which I personally use...
I think he is exaggerating, but he does have a point. How many uses native applications for reading email today? More and more are also moving to the web for office applications.
Some areas are completely swallowed. In the 90s it wouldn't have been strange to have a native application for the weather, or to browse a digital encyclopedia. Those areas are now only on the web.
No. He proved that only a specialized minority need specialized applications. For a majority of users the computer starts and ends with the web browser. The same thing will happen on mobile
How many people aside from specialists use the browser as their main "app"?
How many people use webmail, search and read info, use facebook/twitter, order stuff, use online map, look at internet junk?
The browser is the number one app on desktop without a doubt. The flexibility it offers is nothing short of incredible. You can cite specialist apps all you like, it doesn't change the fact that those apps are a niche right now on the desktop.
For anything that's processor intensive it's clear that apps will win, but flipping that think about the classic office bundle and how google moved that to the browser:
Maybe not at enterprise, but in my experience google apps have displaced office for smaller orgs and personal use. I haven't used Word or Excel in like 6 years now.
Those are all apps that actually need more access to the machine's resources. That's not really a comparable to what -most- apps in all of the major app stores do. Most of the apps in the app store are just for media consumption and generally the web version is better than the native app. See Facebook for example.
Media consumption v media creation I think is the distinct line that need to be crossed before a native app makes a whole lot of sense.
For audio production, iMaschine is a fantastic app and the main reason I stay with iOS. In fact, the $5 app was so good that I ended up buying the $200 Maschine hardware controller + desktop software for the seamless integration: create a track in iMaschine and it will export a file that you can open up on the desktop to apply advanced effects, additional voices, etc.
The thing is all those things will eventually migrate to the web. Right now the web has so much potential and those pieces of software have a small market and that market is pretty saturated.
Once we do everything we can do with the web then we will move into domains that have incumbents
I would say Facebook is much more useful to society then Audio Production.
The web is a cesspool. Its monetization is so twisted and unnatural, with individual users worth next to nothing. Companies engage in all sorts of dark patterns, tracking and other scummy bullshit to try to squeeze out an extra tenth of a penny per user. Hell, click on the link to this article and Amazon, LinkedIn, Google, and a dozen others all get to know what you just read!
I don't share the excitement about bringing this bullshit to apps. Maybe the web will improve: as websites become more sophisticated, customers will increase their expectations and start demanding better treatment. But this would require reversing the current trend.
Native apps can be even worse - they control all behavior to a more powerful degree. The bloat and battery & data drainage is why I install as few mobile apps as possible.
For Writing/debugging software definitely check out Codeanywhere.com. I am the founder so would love your feedback if you haven't tried it out yet. Thanks :)
This year, browsers are implementing the Service Worker spec [0] which is a major improvement to the previous App Cache spec [1]. Both could be used to load UI (and all other static assets) when offline.
There are also numerous headers for describing TTL for downloaded assets, and APIs for offline storage (localStorage, indexedDB).
I think the main differences are that while the web has these capabilities, you don't get them for free. You do have to go out of your way to use them. Usually, they're developed assuming an Internet connection, then made to work offline. Usually, it's the other way around for native apps.
> The problem is terminology and the exact focus of each study. Morgan Stanley’s study [browser > app] is focused on unique visitors — calling it, somewhat misleadingly, “traffic” — while comScore’s report [app > browser] is focused on actual user time spent.
Apps use less data than mobile web sites - API's over full HTML/CSS/JS stacks. When a web page on a modern news site can be ~20MB, this news is no surprise.
Most popular used apps/platforms nowadays are not related to hardware (Facebook, Twitter, Medium). Chances are, your product isn’t too.
The only thing annoying me in mobile browser experience is the UI. For example, Safari on iOS. Let’s hide URL bar and get rid of back/forwards buttons in favour of using gestures 100%. Then, most of websites will be perceived as apps for majority of users anyway.
By moving to native we officially obey corporations to control our products and it’s surely not the way to go.
It's a ridiculous to call the metric measured by this study "Traffic". All it's counting is "unique user's as counted" on all the websites.
So while I'm scrolling through twitter app on my phone, I might click on about 20-30 random links (most of which open in the browser), take a quick look and swipe away the tab. According to this study, I just presented myself as "Unique User" on 20-30 sites. All the while, I'm simply browsing twitter.
So according the methodology of this study, I just generated 30x the "web traffic" as compared to my app traffic. It's a patently ridiculous way of measuring mobile usage.
Uhm. I would say it's more notable that 1/3 of mobile traffic is seemingly under the control of a few companies. If you were ever concerned about the openess of the Internet, this is it.
One of my pet peeves are the sites like LinkedIn and scribd that push you hard to install a c-rap. If at all possible I prefer to browse desktop sites on my tablet father than mobile sites that assume I have sold my firstborn son to Verizon in exchange for the right to breath through a straw in places where Verizon feels like providing service.
And also Traffic says nothing about usage.
I mean a Browser needs to download more things than any App will do, thats why Apps are used so heavily.
the only statistics you can trust are those you falsified yourself
I've only read the headline, but it claims mobile traffic (whatever that means) is both larger and faster growing, which renders the xkcd strip irrelevant since the two religions are competing on only one of those.
I find is to so much easier to just type out the URL of what app I need - facebook, reddit, youtube, etc -
Its just faster to let google auto-complete then having to browse my own phone for the app that I need.
typing the letter "f" is faster than exiting the browser and using my OWN EYES to browse left to right row by row the apps on my phone. How cave-manish.
The iPhone keyboard famously are not very good. So are older phones.
The newer android phone have amazing responsive. I sometimes use my phone to actually write presentation speech. Something I though I would never do with my older android phones.
The term "traffic" is never defined. Web requests? Quantity of data transferred? Both of those would have significantly shapes depending on whether the client was a browser or an app.
It's also interesting to note that the top 10 "Browser Reach Advantage" websites are basically text based websites - news sites, wikipedia and blogs.
Things I learned from the report: the top website (with 207.0x Browser Reach Advantage) Blogger has an app.