A study about content blockers for iOS 9

[Updated: added ‘Distilled’ as a recommendation. Keep reading to find out why.]

INTRO

Much has been said about content blockers and the morality (or lack, thereof) of using them, and it’s not my place to go over it all over again. When I set myself to make this study, I set two simple goals:

– Inform iOS users, who are still on the fence on which content blocker to use, or that are not totally satisfied with their current option;

– Contribute with my feedback to any developers who are willing to look at this study as an opportunity to improve.

All content blockers were tested following a strict set of rules, in order to be as precise and fair as possible. Please note that performance may vary, according to your setup (device, network speed, router loads).

For practical reasons, I wasn’t able to test all of the content blockers at the App Store. At the moment, 16 blockers were tested, of (currently) 29 at Dave Mark’s curated list at The Loop. More blockers will be added to this study, in time.

If you’re a developer, and your content blocker is not part of my tests, or if you spot any mistakes, DM me at Twitter (@carlos78oli).

Tested content blockers (in alphabetical order):

  • Adblock Fast, from 10x Studio (Free)
  • Adblock Multi, from Roman Shevtsov ($0.99)
  • Blockr, from Tim Poller & Arno Appenzeller ($0.99)
  • Chop, from TalkAbout Design (Free)
  • Clear, from Tightrope Interactive (Free)
  • Clearly, from Gist LLC (Free)
  • Crystal, from Murphy Apps ($0.99)
  • Distilled, from Sean Murphy ($0.99)
  • F-Secure Adblocker, from F-Secure (Free)
  • Freedom, from Zach Simone (Free)
  • Just Content, from Blackwater Park Studios (Free)
  • Lionz, from Appsolut Secure GmbH (Free)
  • Privacy, from Obsessive Software ($0.99)
  • Refine, from Luke Li (Free)
  • Silentium, from Francesco Zerbinati ($2.99)
  • Vivio, from Jan Cislinsky (Free, with IAP)

Note 1: Prices valid at the time of posting.

Note 2: As I’m publishing this article, Lionz’ v2.0 update was released on the App Store, adding new features and improvements, and Distilled v1.1 is awaiting approval from Apple. When the opportunity comes, I will test these two updates, and reflect any changes in my study and conclusions.

SETUP

iPhone 5S running iOS 9.1 (Public Beta) connected by wifi to a 100Mbps optical-fiber network

SITES TESTED

I’ve tested content blockers in 14 popular websites, always using the same article from each website. The reason I’ve tested article pages, and not homepages, is because homepages are dynamic, and might change substantially between tests, misrepresenting the collected data for any individual content blocker.

After a long consideration, I won’t be revealing exactly which sites or articles I’m using to test the content blockers, the reason being that I’m still open to test and add new content blockers, and by revealing which sites I’m using, developers could take the chance to optimize their apps to those specific sites.

When I decide to close the tests to new content blockers, I’ll publicly announce which sites I’m using, and will even make available all tables and charts to general public and developers.

METHODOLOGY

  1. Before running tests on content blockers, I’ve closed all of the apps in my device’s app switcher, cleared all history and website data, restarted the iPhone, and ran a test with no content blockers turned on and no ‘Do Not Track’, to set a comparison baseline.
  2. When testing a content blocker, I would previously clear all history and website data.
  3. Proceeded to load each individual links/articles from bookmarks I previously created.
  4. In each article, I would scroll to the bottom of the page (or end of article, if an “infinite” page is present), in order to allow any eventual hidden content to load.
  5. I would then take note of the time the page took to load, the number of files loaded, and how much data was downloaded (data collected through Safari’s web inspector, on the Mac).

BY THE NUMBERS / THE RESULTS

Having collected all the stats, regarding data consumption, loading times and number of files/scripts loaded, I’ve put all that information into charts, which give us a fairly good first impression on content blockers performance, the one we can measure in numbers (more on that later on).

Data Usage, in Mb

Captura de ecrã 2015-10-14, às 02.11.52

Lionz and Silentium are the winners regarding the amount of data they both managed to save (51Mb), but followed really close by Adblock Multi and Privacy (48Mb). However, even the less efective content blocker, managed to save 29Mb from being loaded. If you have a limited and small data package, savings in the order of 29-51Mb (and remember, this is just from 14 web views) could mean a huge financial saving at the end of the month.

Loading Times, in seconds
Captura de ecrã 2015-10-14, às 02.13.03

Regarding loading times, differences are much more subtle among the better performers, but might mean the world for anyone in slow network environments. Silentium, once again, manages to be the best, but only by a slight margin to its closest competitors, beating Just Content and Lionz by a very slight margin (4 seconds), with Privacy just 2 seconds behind the duo. These three are the only ones that managed to cut loading times by over 50%, with others not far behind (Privacy, Chop, Crystal, AdBlock Multi, Blockr and Distilled).

Files/Scripts loaded

Captura de ecrã 2015-10-14, às 02.12.18

Running no blockers, the number of files/scripts/trackers astonishingly passes the 3500 mark. Even Clear, the less capable content blocker regarding loaded files, managed to cut that number by more than a third. But what about the better performers? Well, Silentium and Lionz went even further, and cut the total number by more than two thirds (68% to be more precise).

Now, these numbers don’t mean everything by themselves. What good is a content blocker that’s fast, removes a lot of files, and saves a big amount of data, if pages get “broken”? And what does each blocker offers, beside the blocking of ads? That’s why, besides a quantifiable evaluation, a quality inspection is also needed.

FIRST CASUALTIES

It’s my personal opinion, that a content blocker must always provide the means to whitelist and support our favorite sites and authors, and anyone who respects their readers’ privacy (and wallet!), while trying to provide them with the best possible user experience. If you share that same line of thought (and you should), keep reading along. If not, just make an educated choice based on the above charts, and your personal tastes, and you’re good to go.

By removing all content blockers that don’t offer any whitelisting option, we’re left with Adblock Multi, Blockr, Clear, Distilled, Just Content, Lionz, Privacy, Refine and Silentium. But while all these offer the option to whitelist our favorite sites, only five of them go a step further into facilitating our lives, by allowing us to whitelist sites through Safari’s share sheet: Adblock Multi, Blockr, Clear, Lionz and Silentium. Although I won’t exclude any content blocker for not having this feature, it’s a nice bonus that should be taken into consideration. Not having this, forces the user to copy the URL, open the content blocker app, add the site, then return to Safari (I’m betting most users give up whitelisting sites after going through this procedure more than a couple times).

THE PROMISE

Now that we’re left with 9 choices, it’s time to dig into what each one has to offer, and what they promise to block. Some people may like having lots of options and customization available, while others may like to keep it simple, with as few options as possible. Take this table just as a reference, it’s up to you to decide which features (or lack thereof) might interest you the most.

Captura de ecrã 2015-10-13, às 04.25.54

Looking at this table, one can easily deduct that Silentium is one of the most complete and customizable options. It’s also the most expensive, so it’s up to you to decide if you’re willing to pay the price (although, its performance and feature set are, in themselves, worth the admission price). A case could also be made for Refine, and its custom and public rules, allowing for an even greater customization, but it’s not as accessible and understandable to the general audience as Silentium.

If you don’t want to have to worry about which options to turn on or off, only two of them are completely free of any setups and options: Clear and Just Content. You install them, turn them on in Settings app, and you’re good to go.

(Please note that just because any given blocker is shown as not having EU cookie banner/social media buttons/comments blocking, doesn’t mean they won’t. As a rule of thumb, if no option is present, it means that they don’t, but in some rare cases it might just happen that they DO block them, but the developers don’t advertise the feature.)

DO THEY KEEP THEIR PROMISES?

Well… Yes and no. Ad-wise, all of them keep up with their promises, and they do block the majority of ads, save for some very well concealed hidden ones (which I’m yet to find, in all honesty). As for comments sections and social media buttons, it’s a case of hit-and-miss.

Some comments sections are removed (mostly those from Disqus or Facebook), when the content blocker offers that option, but others are not. Most of the times that happens because some webpages use their own code for comments sections, and that would require individual rules to be added, per site, to the content blocker rules database… And of course, creating rules for all webpages in the world is an impossible task.

Same could be said for social media buttons. Some manage to get rid of them, others don’t, and when they do, most of the times they leave behind an empty box or frame. There are no perfect solutions.

(Bonus tip: if your choice doesn’t offer comments blocking, or if you’re unsatisfied with its rate of success removing them, check Shut Up. It’s free, and can be used together with other content blockers, without affecting performance in a significant way. Check Federico Viticci’s article in MacStories)

WHAT’S BROKEN

Broadly speaking, not one of these nine content blockers broke anything, nor did any pages stopped working. The most common bad behaviors I could find, were blank spaces being left by some content blockers where ads used to be (picture a newspaper with holes where printed ads used to be), and sponsored links not being removed. In that regard, Blockr, Lionz, Privacy and Silentium were the most efficient, respecting pages layouts and formats, while simultaneously removing all ads and sponsored links (in the sample sites I’ve used, at least).

CONCLUSIONS

I won’t dare to pick a clear and definite winner, because… there isn’t one. There are just too many variables at play, that don’t allow me to make a judgement, and be fair to all developers. That’s a choice you’ll have to make, because what I value might not be what you’re looking for. I value a content blocker who blocks comments in sites, and social media buttons, but you might not want to block those. I don’t mind paying for apps, but you may want a free app. The list goes on and on.

There are no bad content blockers, but in my humble opinion (and at the time I wrote this), I feel like Lionz, Privacy and Silentium (purely alphabetical order) are some of the safest bets. This may change in the feature, as new content blockers get released and others get updated, and I’ll try to reflect that in future updates to the study. Also, don’t rule out Distilled, for a very honorable reason: all of the proceedings from Distilled sales will revert to Sean Murphy‘s other job: rescuing and taking care of animals on a farm.

THE BOTTOM LINE

Having spent a huge amount of hours (too many to tell) testing these blockers, there’s one thing I’m really certain of: your web browsing experience will greatly benefit from having a content blocker. Any content blocker. When going back to no content blockers turned on, I was reminded how ugly the current state of the web really is, and I really hope that advertisers and publishers alike take this chance to really improve upon the way they do business. So, please look at the numbers, choose how much (and if) you want to pay, and what options suit your personal tastes. Just make sure to choose one with an option to whitelist your favorite sites and authors, and show them your support.

It’s getting tough being an European Apple user (a rant about prices)

On Tuesday, Apple made some headlines, with the introduction of a 21.5″ 4K Retina-display iMac (Macworld review by Jason Snell), and also a new set of Mac accessories: Magic Keyboard, Magic Mouse 2 and Magic Trackpad 2 with Force Touch (read also Jason Snell’s review of the Keyboard and Trackpad over at Six Colors).

However, these new products once again brought forward a terrible trend for European Union Apple users: an escalation in prices.

Let’s take a look at US prices for the new products:

  • iMac 21.5″ 4K Retina display – $1,499
  • Magic Keyboard – $99
  • Magic Mouse 2 – $79
  • Magic Trackpad 2 – $129

In Portugal, these are the prices for those same products:

  • iMac 21.5″ 4K Retina display – 1.749€ ($1,991)
  • Magic Keyboard – 119€ ($135)
  • Magic Mouse 2 – 89€ ($101)
  • Magic Trackpad 2 – 149€ ($170)

On average, these Apple products cost over 30% more to a Portuguese customer (that number slightly varies in regards to other EU countries).

Yes, a case could be made that prices in Portugal include a hefty 23% tax load, but the fact is that prices in Portugal (and generally speaking, in the European Union) are progressively getting higher and higher, for the end-client.

Still not convinced? Here are some numbers for one of Apple’s  crown jewels, the iPhone:

  • By October/2014, an unlocked 16Gb iPhone 6 had a Portuguese price-tag of 709€ ($806, by rates at the time).
  • That same date, we could also get an unlocked 16Gb iPhone 5S for 609€ ($692).
  • Today, a brand-new and unlocked iPhone 6S costs 759€ ($864), an increase of 50€($58).
  • Also today, last year’s 16Gb iPhone 6 can be obtained by 649€ ($750), having lowered 60€($56), but representing an increase of 40€ ($58) to last year price for the (then) 2nd model in the line, the iPhone 5S.

A well-informed consumer knows that there are many factors in play, that can lead to these price fluctuations, besides the obvious currency exchange rates issue (a weakened Euro in regards to the US Dollar), but how can we expect that the public, in general, looks at these  numbers, not thinking this is just Apple getting greedy, just looking for a larger income?

I consider myself a smart and well-informed Apple user/client, but even I start finding myself lacking the arguments to try to explain and justify this.

(Oh, and don’t even get me started on the lack of Portuguese-speaking Siri… I’ll save it for another time).