AMP For E-commerce ?

[Solipsism]

Hi everyone, i create this topic to have your feeling about this question.
Is AMP worth for an e-commerce website ?
(I have a lot of doubt about the crawl budget and the thing that you have an other website to admistrate. And for the speed's argue, i think that a PWA is more appropriate ! )

[not2easy]

We recently had a discussion (OK over a month ago) asking the same question here: [webmasterworld.com...]

It does not seem to be the best format for e-Commerce sites.

[tangor]

AMP by definition, strips a lot away to achieve a purpose for g's ability to deliver fast ... likely does not play well with ecommerce sites which need a bit more content to make the sale.

YMMV

[NickMNS]

AMP by definition, strips a lot away to achieve a purpose for g's ability to deliver fast

This is an over simplification, that ultimately leads to false statements.

AMP allows only those things which are necessary and sufficient to delivering the first page view. In addition to this it allows resources required for future page views or user interactions to be pre-loaded such that subsequent page / interactions are as fast as the initial one. That said, creating static AMP page for Ecommerce website is less than ideal and most likely a waste of time and resources. But...

...And for the speed's argue, i think that a PWA is more appropriate !

Yes and no, a PWA will have a very slow first page load because the user will not have access all the resources required and thus these will need to be loaded. After that point, yes things will be faster, as much of the computation will be done client side, and resources will be cached/stored.

[...back to the but]
AMP and PWA are not mutually exclusive, they are in fact complimentary, by design. Creating AMP pages that preload your PWA resources (service worker) gives you the fastest possible execution that would be hard to replicate any other way. The down side, is the level of complexity of the coding that is required to make all the moving parts work seamlessly. If you intend on creating a PWA regardless, including AMP is likely a good investment. If all you intend on is having is few static product pages, then AMP may not be the best choice.

Here is a video from Google that explains things much better than I have:
[amp.dev...]

[Solipsism]

Thank you @NickMNS !
The fact is, that i'm working as an Inhouse for an e-commerce. And the previous SEO has implemented AMP.
1) I have a lot of problem with AMP for the tracking in Analytics .
2) I have a log analysis tool, and when i look at our AMP, they are really really crawled, like 90% of our budget crawl are on our AMP pages. But, only 5% of our AMP pages have a visit from Google.

--> I don't know what to do. They invested a lot in AMP but sometimes i don't think the ROI is worth. I know that the good thing from AMP is that it is more crawled by Google and the speed score is good. But our conversion are bad and our product pages are just empty static page.
Should i remove AMP from our product pages ?

Moreover, AMP really work for our blog and it's logic because it's a media content.

[NickMNS]

@Solipsism

I have a lot of problem with AMP for the tracking in Analytics

Analytics is another big down side in terms of AMP. Stats from AMP pages are heavily biased due mostly to where and how the technology is applied. It is important to understands these biases that are inherent with AMP before making any decisions.

In most cases the AMP technology only applies to landing pages. This is specially true for static amp pages.

But our conversion are bad and our product pages are just empty static page.

When a user from search lands on AMP page a page view is recorded but if that page is static, so there is no action for the user to take other than to navigate to another of page of the website. In most cases that page is not likely to be AMP. So the question then becomes are conversion being correctly attributed to the AMP page, or is the subsequent page getting the attribution. In the later case it would significantly bias AMP pages to lower conversion rates.

Then AMP mostly applies to mobile users, typically mobile users have a lower conversion rate than desktop, so you need to compare the conversion rate from AMP pages to non-pages with mobile users.

...like 90% of our budget crawl are on our AMP pages. But, only 5% of our AMP pages have a visit from Google.

First, crawling and ranking are two different things. Obviously if a page hasn't be crawled it can't rank, but a page that has been crawled does not necessarily have to rank. The second part of the statement, on its own provides little information. Statistically the statement of all AMP pages only 5% get traffic is fine but you need put this in context to glean some value from it. You need to ask of all non-AMP pages what % get traffic. And what % of those pages are the same. Then, as above, you need to segregate out mobile users vs desktop, and then compare AMP vs non-amp mobile users. With these figures you will then be in a position to make a more informed decision.