Huge drop on revenue after a redesign

[tasoskavala]

Hello everyone,

If I am right, it is against the rules to share statistic numbers from my adsense, so I will try to describe my problem without them.

I had a website with adsense and last month I decided to make a redesign since it wasn't responsive and the view on tablets and mobiles was awfull.

So, instead of fixed sizes on adsense, I used also a few responsive units and hiding and showing other adsenses in base of the viewport of the device.

I published it 4-5 days before, but since then my stats has had a huge decrease. I hoped that I would have an increase instead.

The problem is that even the number of page views is totaly different from Google Analytics. Almost the half of it. Is there anything else I could do to monitor it better?

[netmeg]

Well for one thing, you have to wait a while for things to settle out. Did you keep all the same URLs? And if your number of pageviews is that far off, then I'd look for a technical problem first to make sure everything is rendering properly.

[not2easy]

Hello tasoskavala, welcome to the forums. I believe it must take a while for Google to process the changes because I have read the same complaint from several different people here in the recent past. Even down to the GA mismatch in numbers. I would not be too worried since it is so common. Like netmeg said - make sure there are no technical issues, but apparently it is not a 24 or even 72 hour process. Read down through some recent threads here and you will see it is not rare.

[tasoskavala]

Thank you all for your answers. I have kept the same URLs. I did the following changes from technical view:

1) On first version, the multilanguage feature was handled from Javascript. Now, it is from backend and load the page in the language the user choose.

2) All the website (+ the adsense banners) follow a responsive view way.

3) Now, I noticed something that may interrupt my stats. In the first version of the site, I have several forms (about 20) which is the main purpose of the website and after the submission, I load another page after a few functions from the backend. Now, I moved those functions on Javascript and I don't refresh the page to show the results. This could lead to lower page views I think. But not to the difference from GA and adsense stats.

I will wait a few days more to let it stabilized and then I will see if something else happened.

As for the technical issues. What else do you think that could lead to this problem?

[AlexB77]

So, instead of fixed sizes on adsense, I used also a few responsive units and hiding and showing other adsenses in base of the viewport of the device.

When you say you are hiding non-responsive add units do you actually mean you are not loading them at all on all responsive screens or you are hiding them from user view via css?

If you are hiding them via css, then this is a huge problem.

[tasoskavala]

This is how I did it:

<div class="adslot1">include adsense code</div>
<div class="adslot2">include adsense code</div>
<div class="adslot3">include adsense code</div>

@media only screen and (min-width : 993px) {

.adslot2 {
display: none;
}
}

@media only screen and (max-width: 600px) {
.adslot1 {
display: none;
}

.adslot3 {
display: none;
}
}


@media only screen and (min-width : 601px) and (max-width : 992px) {
.adslot1 {
display: none;
}

.adslot3 {
display: none;
}
}

I asked on StackoverFlow how I could do it without violating the rules of Google and they said that this is the right way

[AlexB77]

It is actually wrong and could be a potential cause of your problem.

First of all, even when you are hiding it via css the code for an add block is loaded anyway and for google adsense it will record a page view when in reality it was not displayed at all. Ypu should only load add block that you want to be displayed and not one that you don't.

[tasoskavala]

And how can I do it instead?

Whatever, I have 3 banners and one link ad. In base of the viewport, I show max 2 banners and one link ad. I don't pass the limit of 3 banners in the same page, even if I calculate the hided banners.

[AlexB77]

I am not judging you, only trying to help. It is not about exceeding limits it is more about hiding adds instead of showing them. Also, what if you need to have 3 add blocks on desktops an 2 on mobiles?

You could use PHP library to detect the device and push the right add code.
I can PM you the link to that library it is very light and easy to implement.

[tasoskavala]

Of course I didn't think that you judge me. I am not a professional developer and I take every post as a valuable feedback.

Unfortunately, I use Python instead of PHP for my backend.

[AlexB77]

There should be some library for Python also, similar to PHP. Let me do a quick research on this and i will get back to you.

[AlexB77]

Not sure if this would help but after a quick search i found this pymobiledetect 1.2.2

This should help

[RedBar]

Did you keep all the same URLs?

If you did not do this then G could take months to update and re-evaluate your site and you may have to wait until the next Panda update for it to take effect.

[bhukkel]

Here is a official google example of hiding an responsive ad:

[support.google.com...]

[AlexB77]

@bhukkel this is correct but only if you want all of your adds to be responsive on all of the devices including desktops, but if you want your adds to be responsive only on the mobile devices, then this is useless as you will need to make some adjustments to your back end, first to recognise device via UA and then respond with correct set of add code. For example if you have your responsive ad in one include file and regular code in another, you can load desired on page request via PHP or Python.

[bhukkel]

@alexb77 i understand your situation but it is not really a responsive website. Responsive ads can also benefit on desktops. There are so many screen resolutions today.

[Mike343]

fyi if you are using async tags then all you have to do is the following.

inside some media query:
.hide { display: none; }

then the async tag
<ins class="adsbygoogle hide"

[tasoskavala]

I change the css code and the class on the async tag too to avoid violating any rule.

To sum up, I haven't changed any URL. They are the same as before. But I think I did I huge mistake. The main idea of the website is to have 20 web forms with specific formulas to calculate different things each one. On the first site, the formulas were calculated on backend. Now, I moved them all on Javascript. As a result, the pageviews number has been halfed when my users number stays the same.

Could this be the reason of the huge loss on income?

[AlexB77]

It is very obvious, because by calculating something on the back end you were forcing user to see all results on the new page, which means all new set of adds were displayed and all new page views were recorded, the only question that you probably should ask is when your users were actually clicking your ads, I guess after they had their results and not before then because they were obviously interested in getting the result first and were staying on your site until that moment. But again this is just my theory may be I am completely wrong as I do not know your site as well as you do.

[tasoskavala]

This is a nice answer for brainstorming and find the problem. However, I am not 100% sure that this could be the case here. The reason is that the form and the result page is the same. In the previous version, the screen was splitted in the half. The left part was the form and the right part was a blank place to populate the result after the submission. The difference now is that I use JS to populate the blank space.

Also, the results are only numbers which means that it could not influence the kind of banners.

Another thing that I saw today was that on adsense the Website Health had a drop. I checked and the reason is that I didn't cached the new JS files and I had a negative score from Google.