I am running into a strange issue with the goal funnel that I set up for my Shopify store.
I used the following URL for step 3, which corresponds to the customer choosing whether they'll use a guest checkout or if they'll create a new account:
/account/login/*
I used the wildcard because Shopify appends that URL with a unique ID.
Problem is, when I look at the Funnel Visualization report, I am seeing a lot of exits from step 2 (view cart), and the exit URLs are all of the type "/account/login/0f53d3781dcdf4feie6a3", which actually means they didn't actually drop from the funnel, they went to step 3, and should have been captured.
Any idea why this is happening? Did I just set things up incorrectly?
Thanks in advance!
Lukas
Google Analytics does not use wildcards, it uses regular expressions (not all regexp features are supported, though).
if you want to you regular expressions in your funnel steps the
matchtype for the destination url needs to be set to regular
expression, too.
The star character (asterisk) means "match any number of the preceding character". So "/account/login/*" would match e.g. "/account/login//" or "/account/login/////" but not "/account/login/0f53d3781dcdf4feie6a3"
You should be okay if you use the "any" character which is "." (dot), so your regexp for the funnel step would like this:
/account/login/.*
(match anything that includes the phrase "/account/login/" and has an abritrary number of abritrary characters after that). Or you might try to craft a regexp that matches only valid session ids (personally I would not bother).
You are facing this issue due to problem in Regex. Due to improper Regex, pages which should have been counted in Step 3 are appearing as exit pages in Step 2. Let me explain in detail.
The Regex you have applied is-"/account/login/*". Here * at the end means it Match zero or more of the previous items. Means it will track URLs like "/account/login//","/account/login",
and "/account/login///". But in your case URLs are different as it is being appended by Unique ID.So existing Regex will not track what you are intending to.
To solve your problem i'd recommend you to use this Regex- /account/login/.*
I believe that this should work as a dot,means get any character and a star means repeat the last character zero or more times.This means that the dot could match any letter in the alphabet, any digit, any number. And the star right after it matches the ability of the dot to match any single character, and keep on going. So this Regex will match all the URLs you are intending to track, even with random Unique IDs.
Hope i have answered your query.
Do let me know if this works for you.
Related
I have followed the thread here about splitting out similar URLs, and follwed the example given but I have a bit of a problem.
Some of the URLs in my analytics page are being sorted correctly, then some of them don't seem to get sorted at all. In the 2nd image you can see the problem I'm having. The 4th and 5th links on that image, are meant to be grouped under the 1st link of /blog/
I've included a screenshot of my filter as well, as I might be missing something in the filter, to group all the /blog/ links together.
The way the links are in the website is like this:
great-blogs.co.uk/blog/123456789012345
Any idea on how to go about combining these rouge links into the main /blog/ link would be great.
(.*)/\d+(.*)
That works perfectly with the problem I had with the IDs, buried in the middle of the urls.
If you want the pages grouped only into /blog/, /blog/edit, or /blog/new then you would need field A of your filter to be
(/blog/)[0-9]+$|/blog/.*/(edit|new)
This assumes that the string that could show up at the end is numbers only, eg. /blog/1478744474. You could also modify this so that if you know it's just going to be numbers before /edit or /new, then the regex could be
(/blog/)[0-9]+$|/blog/[0-9]*/(edit|new)
The output field would be $A1$A2. Notice how I included the slashes in the field A groupings to make the output look a little cleaner.
Test this though, because I'm on a mobile right now and can't access a good online regex tester!!!
EDIT:
Two changes: 1) note how the * is now a + so that the number is actually required (eg. /blog/1234). 2) I had to modify the regex because I finally got to testing it and found that the brackets might impact the matching, and also because the advanced filters use the same brackets to piece together the output, I had to redo the regex and break it up into two matching groups (that probably doesn't make sense but I hope it works!).
I have a website on which the URLs are translated depending on the language chosen by the user.
Now, I need to set up conversion goals in GA, however I'm not sure how to do it in my case. Let's say I want to track registrations via About us page. The goal(s) would look like this:
(EN)
/
/about-us
/registration
/dashboard
(DE)
/
/über-uns
/registrierung
/dashboard
As you see, steps 2 and 3 differ in terms of language used. BUT, this still is the same goal.
How do I handle this in GA? I would prefer to have single statistic for this, instead of separating both language versions.
I know that I could set up custom pageviews but this seems like an overkill. Is there a simpler way to achieve what I need?
Thank you!
If you do not need a funnel you can send an event with a language agnostic identifier along with the pageview and make an event goal. This is only marginally simpler that custom pagenames in your tracking calls, but at least you get to keep the original urls.
As an alternative to goals altogether you could set up a hit-scoped custom metric that is incremented on each of your destination urls. This works well if you only need the raw numbers (but you cannot use this in Google Adwords the way you export goals to Adwords, so it's less useful if you do paid advertising).
You can try using RegEx (regular expressions) then, If " registration / registrierung " is your goal the use following as:
registration|registrierung
For setting up make sure you select - Regular Expression at
Destination type in Goal details.
The PIPE b/w would mean - OR - in Regex
You can apply this to goal funnels as well.
Do you want to include step 1? I would leave exclude it from the goal funnel
These regxes below would work in a funnel but the they assume that there could be other params added. If you want exact match e.g. change .* to $ But then you could miss urls with tracking params e.g. in each step.
Second step: ^\/(about-us|über-uns).*
Third step: ^\/(registration|registrierung).*
Goal page: ^\/dashboard.*
What does the following patterm means in the URL, these pattern seen on variety of url like Linkedin
Enpv_301654428_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1
is this sparse Matrix representation ?
For example
https://www.linkedin.com/groups?gid=5010991&goback=%2Enpv_301654428_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1_*1&trk=prof-groups-membership-name
I know it's the value passed to the parameter goback, but what's the point to use this kind of strange looking values?
You may want to check this
this link!
It's basically a method to allow the system remember steps to go forward\backward in the navigation.
EDIT:
When you perform a search on LinkedIn, there are a number of facets on the left hand side of the search results page that you can use to refine your search, like company and location. When you click on a search result, it turns all of those facets into the *1_s and *2_s that you see in the URL. Those are used to construct the "Go back to Search Results" link, which will restore the facets you had selected.
This stackoverflow previous question may also help you.
I'm not getting the right results returned for a goal I have set up. The goal says the last 7 days have 92 goal conversions, when it should be 400+.
When a user completes a subscription purchase, they land on a confirmation page. We have several subdomains that the user can be coming from as well as the potential for a reference appended to the URL. So, I have the goal set up as a regular expression like this:
.*/insider/confirm.*
If, for example, the user pays and then is directed to the following page, shouldn't the goal be recorded?
test.com/insider/confirm?ref=2343
As I can't see anything wrong with the regular expression, although very greedy, the goals should be captured with the above example URL and goal setup.
Have you recently set this goal up? A thing with goals are that they aren't historically applicable to your data. They start collecting data from the point where you set it up.
Have you tried using the above regexp in the "All pages"-report under "Behavior"? Enter it into the filter pattern and there you can see exactly what pages your regexp would capture.
I am developing a flight search engine for a customer, and currently the URLs look as follows (ad = destination airport, ao = origin airport, dates and number of passengers are not specified here):
http://example.com/#ad=S%C3%A3o+Paulo+-+Todos+os+aeroportos+(SAO),+Brasil&ao=Recife+-+Guararapes+Intl+(REC),+Brasil
My customer wants to make search pages more search engine friendly (SEO). The idea is that Brazilians who are looking for flights from, say, SAO to REC by e.g. Google should have a higher chance of finding that particular flight search engine.
The first step is probably replacing the fragment identifier (#) by a query string (?). The server then dynamically generates nice text content that can be viewed without JavaScript (search results would still be loaded via XHR). In my opinion, that makes a lot of sense.
Now, to make the URLs more search engine friendly:
(A) My customer proposes adding additional keywords into the URL, something like:
http://example.com?flights+to+Porto+Alegre&S%C3%A3o+Paulo+-+Todos+os+aeroportos+(SAO),+Brasil&ao=Recife+-+Guararapes+Intl+(REC),+Brasil
(B) I propose adding a slug instead, which can easily be internationalized, and which is good to read also for humans. Example:
http://example.com/pt_BR?ad=REC&ao=SAO/voos_de_Sao_Paulo_para_Recife
(C) Or, perhaps without a slug (but - due to parsability - only for a limited parameter set, which has the disadvantage of limiting sharing of URLs by users):
http://example.com/pt_BR/voos_de_Sao_Paulo_(SAO)_para_Recife_(REC)
What do you suggest? Any examples of good URLs for similar use cases?
That all being said: I understand that links from highly ranked pages are still the most important ranking measure. In the end, I wonder if all that complexity really is worth the effort. When I look at Google's own search pages, then they are rather simple. For example, there is no summary of the search query in a H1 tag, just as my customer wants. Of course, Google doesn't search itself...
don't use _ (underscore) to delimit words. Google interprets hello_world as one word but hello-world as two words.
don't put your human readable keywords in the query string (after the ?). Instead make it a normal URL http://example.com/pt_BR/search/voos-de-Sao-Paulo-(SAO)-para-Recife-(REC)
I would go for a something like: http://example.com/pt_BR/2012-10-28/voos-de-Sao-Paulo-(SAO)-para-Recife-(REC)