Mobile browser traffic is 2X bigger than app traffic, and growing faster
venturebeat.com
venturebeat.com
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.
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).
On Android.
[1] https://developer.android.com/about/dashboards/index.html
What? People don't upgrade to an older phone. That number has nowhere to go but highter.
The upgrade cycle being a mess just make this take more time.
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.
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.
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.
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?
it's 2015 man, we took the word back. nerdy is cool.
I don't even know anymore what apps I have on my phone.
I simply search and let Google search my phone and the internet at the same time and I click on the most convenient result.
it is definitely a difference, but to act like this is the reason for these numbers is so misguided.
"I use mobile web because apps have too much control" - no normal person ever
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)
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.
we're not that far from a good user experience. even the bar problem is only ios specific, android browsers have a better support for full screen apps
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.
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
And on iOS, even speakers and high bandwidth content are loaded only on a per user request basis.
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.
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).
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.
By stock ROM, you mean you have a nexus device and the original or OTA upgraded firmware on it?
PS: This is the device http://www.gsmarena.com/vodafone_smart_ultra_6-7313.php
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)
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.
_This comment sent using Yarn for HN_
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.
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.
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.
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.
We've really done some progress here yeah.
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?
-Audio production
-3D animation
-Video editing
-CAD design
-Writing/debugging software
3D animation: https://clara.io/
Video Editing: https://www.wevideo.com/
CAD design: https://www.tinkercad.com/
Writing/debugging software: https://codenvy.com/
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.
> 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?
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.
Anyway, both have their use.
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.
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.
Most people with a phone?
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.
- Word Processing
- Spreadsheets
- Slideshows
And Google Apps hasn't displaced Office by and measure. Companies are still depending on and using standard SOE stacks which includes Office.
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.
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.
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.
That's a bizarre value judgment to make, but social apps have different technical needs from multimedia apps:
Bandwidth to the cloud will never beat bandwidth to the local disk. And RAM.
This is completely different on an iPhone 6, Galaxy S6, Nexus 6...
On my devices, including desktop, web is for reading and watching movies.
If there is an application for what I am trying to do, it will be used instead.
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.
[0] http://www.w3.org/TR/service-workers/ [1] http://www.w3.org/html/wg/drafts/html/master/browsers.html#a...
Explains the surprising result.
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.
Hm, do you really think this is a mainstream concern that is driving this?
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.
Hope someone finds it useful.
2013: 1.8x
2014: 1.7x
2015: 2.1x
Get two up years in a row, and maybe there's a trend. You could get that much increase from increased bloat on a few key sites' home pages.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
Aren't you overreacting a bit? This premise is easily proven wrong by anyone driving growth in most startups.
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.
As for cavemanish, some might argue "clicking a large icon" is less cavemanish than having to resort to something as unnatural as a keyboard.
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.