Why I don't like ad blockers

As the comments on the articles show, there is an audience on Habré who is quite fond of ad blockers. In this article I will express my opinion, for which you can strongly dislike them.

What are the claims?

It’s impossible to block ads without ever touching the rest. In many cases this is possible, but not done due to the poor quality of the blocker. Also, often the functionality is generally implemented crookedly, which only exacerbates the situation. But this is not the claim. It’s clear that the developers are not gods, and can’t magically fix everything. But you need to warn the user about this feature?

Let me explain: some users really understand that a blocker can break a site or part of its functionality, and when something is wrong, they automatically reach for the button at the top to pause the blocker. It's good. But some users may not even think that this is the fault of the blocker. They will either leave such a site or be left without the service they need. An honest blocker should, during installation, display a large bright warning about such an important feature that there are no users who do not know about it. The absence of a warning is the concealment of information and misleading, which indicates that the blocker is not honest. Of course, an dishonest blocker owes nothing to anyone, but this article just describes why I do not like blockers - for their dishonesty.

As the creator of the site, I want the site to work correctly even for users who decide to block ads. But in reality, something can break down on a regular basis, which is why I 1) incur losses from the outflow of users, 2) incur reputational losses. To fix this, I have to do the work that dishonest blockers shift to me, or put up with losses. Moreover, the implementation of this work does not bring any benefit to society: blockers as a whole are not good and impede the development of sites, that is, I am forced to do work that does not benefit society. This reason is the second reason for the condemnation of blockers.

It is worse when correction is impossible in principle. Yes, it happens (it will be described below). And this is the third reason - even if you are willing to spend on efforts to support, it is often impossible to correct incorrect work in principle (yes yes, it happens!).

And the fourth reason is not enough to test the site once. It can break at any time without warning. And sometimes none of the users will even notify you of this (for example, because the site does not work for him at all). Yes, you can put 10 blockers and regularly check the site, but this will not save from the 11th blocker, it will not save from the lack of a warning, and it will not save from the fact that you can simply not notice something.


Of course, not only advertising sites break down, but even sites on which there is no advertising at all. The point of view "well, everything has broken, probably because the whole site was littered with advertising" is fundamentally wrong. Among my projects, problems with advertising sites arose more often, but not by much, than with projects without advertising in general.

It should also be noted that some blockers contain a disclaimer in which it is written that they may break functionality or sites, but usually this disclaimer is hidden very far away and is not easy to find - this is very different from the idea to "print in large text immediately after installation."

When does the blocker work incorrectly?

  1. When, instead of using specific rules for a particular site, the general rules for all sites are used. It is unlikely that the container with ads will be called “this-is-container-for-ads”, which means that the rules can even affect sites without ads at all. Developers may not always add specific names to the rules, for example, there are cases when applications become completely inoperative simply because some critical request contained a random value starting with “ad98b ...”. This is the “ad” in the beginning and the blocker reacted.
  2. When a rule is added for a specific site, but this is done in haste. Often, blocker developers do not even conduct minimal testing after adding a rule.
  3. When the blocker itself does not work correctly. For example, there are blockers with the ability to block ads with the mouse, but instead of advertising, they block the entire iframe so that it does not even begin to load. It is already impossible to bypass such a lock, in principle, at least if the iframe is not yours (and most often it is not yours). To the user, it looks like a white screen. 99.99% of users, seeing a white screen, think that the site is not loading, and only 0.01% checks it from another browser, and sees that everything is normal from another browser, which means that there is something wrong with the browser.
  4. When there is a mouse lock function. If the rules are created by the administrator of the blocker, he can test the rule. But users have very little understanding in technology (or rather do not understand anything), so even if the site after that completely stopped working, even in this case, not everyone will guess that this is the fault of the virtual rule they created incorrectly. What can we say about if the site has lost only part of its functionality and in the beginning it is invisible. The site administrator can no longer conduct testing, because each user can cut elements in different ways, and you won’t be able to put 40 blockers to check the work. You can try to ensure work when cutting any element (with a progressive loss of functionality), but this is not always possible, or requires significant costs. Moreover, the solution to the problem is actually very simple: when the user selects an element with the mouse, he should see a warning that this may break the site, and after completing the action in the add-on menu he should be able to quickly cancel the rules added in this way. But do current blockers display warnings? If not, then why?

What can be done?

If you want to make an honest blocker, you just need to display a big warning immediately after installing that the blocker can cause sites to work with errors, break part of the functionality or sites completely. Further, when the user knows about this feature, he can make a choice: either he turns off the blocker on this page, because he understands that the blocker does not work correctly, or he punishes the site for leaving that its developers did not spend their time for testing, and if you spent it, then did not check the work again (unless of course it was possible at all). Website developers decide which blockers need to be supported, and whether it is necessary at all. Everyone has a choice, and he clearly understands what he is doing.

As a site developer, you can display a warning in the spirit "Functionality is not guaranteed when ad blockers are enabled and in browsers prior to IE 11, Chrome 4.0, FF 4.0." Of course, this is not always possible (due to the fact that the site is no longer working for the user, or simply there is nowhere to display such a warning). Also, part of the user may not understand this, because users are sure that it’s the authors of the sites who are trying their best to defend their advertising, and not that the blockers themselves may not work correctly - it turns out that reputation losses again, even if problems have not yet begun.

You can maintain an official blog or forum - if something is wrong with the site, users will know where to go. Otherwise, you may never even find out about problems.

Other reasons for denouncing blockers

There are also other reasons for condemning blockers:

  1. After blocking, the blocker does not notify the owners of the site so that they themselves check the operability, even if the contacts are explicitly indicated at the bottom of the page. Your site may break down at any time, and there is no time to fix it, not to mention preparing in advance and not letting it happen.
  2. It is difficult to turn to the support of the blocker: if it is an honest blocker, then the link for site developers should be in the most visible place, and if dishonest, this is another reason for the condemnation. I’m not saying that you can’t apply at all, but you need to understand that there are not one, but a lot of blockers, and it turns out that everyone needs to study. An honest blocker should simplify the task as much as possible.
  3. Some users even think that the developers of the site are trying to deal with blockers, and therefore it does not work, and in vain they leave the site. This is the most offensive. If the blocker had warned that it was because of him that the site might not work correctly, this would not have happened.

Can blockers fix the situation? Of course they can, after all, only a few warnings are enough for the user to understand that this site does not make tricks or is just buggy, and the blocker is imperfect, and to fix the situation, you need to disable the blocker on this page (if the user is satisfied with this). Do blockers do it now? Not.

Why is user notification after installation so important?

One of many real cases: the user comes to us, complains about the white screen. We see from the screen that he has a blocker, and we say that it does not work correctly. The user understands that if you turn off the blocker, then everything will work, but I’m sure that we specifically made it so that if he blocks ads, then he won’t work. Or we fought with advertising, fought, and now the blocker has blocked us altogether, in short we are to blame, and instead of fixing or removing all the ads, we say disable the blocker.

In fact, due to the carelessness of the blocker, we were slandered. This is very unpleasant. At the same time, it is impossible to fix the problem - the white screen cannot be fixed technically.

“I've never seen broken sites, it's a rarity”

Crashing sites is no rarity. The fact is that applying the statement “I almost never meet broken sites” is absolutely incorrect. Let's say you find 1 broken site out of 500. But what does it really mean?

  1. Firstly, it is likely that you generally visited the site for the first time or used it for a short time. For example, you use the site for a month, but the site could exist, for example, for 5 or even 10 years. During this time, the site could break even not once, but once 10. But it will seem to you that the site is not broken.

    To determine the coefficient, let's take the breakdown time 1 time in 3 years, the breakdown time is 1 month, the site’s operation time is 5 years, and the time you use the site is also 1 month. Then the probability that in the first month of operation the site will never be broken - 0.972 (and broken - 0.028). For 3 years, not broken - 0.363, but broken - 0.637. Please note: yes, in 3 years on average the site will break 1 time, but at the same time the probability of failure is not 100%, but only 64% - this is because the site can break down 0 times or more than 1 time at once. Over the 5 years of the site’s operation, the probability that it is not broken is 0.184, and 0.816 is broken.

    But you use the site for a month, and for you the probability of failure is 0.028. This is 29.358974 times less. It turns out that when you see 1 broken site, the total number of broken sites you visited is approximately 29 times more, because it is unlikely that you visited them every day from the very beginning of the site.
  2. Further, you can hardly use absolutely all the functionality of the site, and if you do not see a breakdown, this does not mean that it does not exist. It depends on the size of the site - there are large sites in which you can not use 10% of the functionality, but there are small ones in which if there is a breakdown, it will be most likely to be noticed. I will not greatly increase the coefficient at this point, but take about 2-3 times (take 2 times). The likelihood of noticing a breakdown is reduced by 2 times - it turns out that if you see 1 broken site out of 500, you need to consider this as at least 58.717948 sites out of 500.
  3. The next point is not the fact that you will think of a blocker at all, because not only it can cause incorrect operation, but also server problems, errors, and more. The probability is especially high at the first opening - you can just close the site and forget. Many breakdowns may not look like breakdowns from a blocker. I will take coefficient 2 at this point. It turns out that if you see 1 broken site out of 500, you need to consider this as at least 117.435896 sites out of 500.
  4. Well, the last point - there is not one blocker, but many. A site can be broken with only one blocker out of 20, and most likely it is with you that it will work fine anyway. I’ll take a coefficient of 5 at this point. It turns out that if you see 1 broken site out of 500, you need to consider it as 587.17948 out of 500 sites. The

    figure is more than 100% - this means that we either made the wrong assumption at the very beginning (for example, we may be able to see a maximum of not 1 site out of 500, but only 1 out of 2000), or some of the points above does not apply to you - because when calculating we did not evaluate in any way what is the likelihood that in any particular paragraph for some reason we cannot apply it.

As a result: even if you come across 1 site for many years, the actual number of broken sites you visited can be even more than half. It is impossible to estimate the number of broken sites according to the judgment "since I have not noticed this for 20 years, then this does not happen."

Another short explanation

Most users expect the ad blocker to block ads. But in fact, this is not so. This is what the user should be warned about. Such a deception is unpleasant to many.

Also popular now: