We are using gatsby to develop our website and I am using the gatsby-plugin-google-tagmanager plugin in order to fire google analytic events..
One issue we face is that when the user visits our site from utm links the session seems to split the exact same second he lands on the page.
What I do so far
Fire a Page View Google Analytics: Universal Analytics tag using the gatsby-route-change trigger.
GA debug report
One thing that seems abnormal is that on every route change, using the GA Debug tool, a new Creating new tracker log is created.
Ways I tried to fix this
Read an article that on single page applications you might get faulty values for page, location and referrer properties, so this fools google analytics to create a new session each time, so that might be the reason why the session breaks.
What I tried to do was to override these values in the GA tag. However, this does not seem to fix the issue.
// Location override gtm variable
function () {
return window.document.location.protocol + '//' +
window.document.location.hostname +
window.document.location.pathname +
window.document.location.search
}
// Referrer override gtm variable
function () {
return window.history.state.referrer
}
// Page override gtm variable
function() {
var path = window.location.pathname + window.location.search + window.location.hash
var index = path.indexOf('?');
if(index > -1){
path = path.substring(0, index);
}
return path;
}
Got any idea on this? Is it possible that this behavior splits our session? Is there anything else you recommend?
https://www.simoahava.com/gtm-tips/fix-rogue-referral-problem-single-page-sites/
This article answers the question.
With Google Tag Manager, every single Universal Analytics Tag that fires on the site creates a new, unique tracker object. This means that the Document Location field is updated with every Tag you fire, which is a problem if the URL changes due to browser history manipulation. Thus you can end up with a situation where the first Universal Analytics Tag has gclid in the URL, attributing the session to AdWords, but the next pageview doesn’t have this in the URL anymore, as you would not include it in the “virtual” pageview path names. Instead, since gclid is no longer in the URL, GA looks at the HTTP referrer of the page to see what the previous page was for attribution. It finds google.com, as you came from the search engine (HTTP referrer is not updated when manipulating the URL with the browser History API). Thus a new session starts with attribution to Google Organic! I’ve dubbed this as the Rogue Referral problem.
Solution
Manually Set Document Location To Prevent Rogue Referrals
Create a script to save the landing URL in the dataLayer
window.dataLayer = window.dataLayer || [];
window.dataLayer.push({
originalLocation: document.location.protocol + '//' +
document.location.hostname +
document.location.pathname +
document.location.search
});
Create a script variable to get the current page. (if you don't need hash remove it)
function() {
return window.location.pathname + window.location.search +
window.location.hash
}
Add variables to all you Universal Analytics by manually setting the fields location and page
Related
I am trying to track Facebook ad results using the Facebook Pixel during appropriate events (page views, lead generation, order form view, purchase). I can do all of this for GA using GTM with no problem, but on Facebook I only have partial success.
The main issue is I have a cross domain setup as shown below:
domain1.com/offer - landing page (FB Page View Pixel should fire)
domain1.com/ordergate - request email before showing order form page (FB Page View Pixel should fire)
crm.com/formsubmission - the actual form submits to my crm (FB Lead Pixel should fire)
crm.com/orderform - order form (FB order form view pixel should fire)
domain1.com/thankyou - the thank you page (FB order pixel should fire)
So my trigger on GTM to fire FB pixel was the "referrer" containing "facebook". However, because of the multi-step process, the referrer is lost by the time the order form or sale is completed.
I have since then learned I need to do the following:
User lands from facebook, write cookie with an appropriately short expiration time that stores this information on domaiin1.com.
When the user clicks a link and is redirected to crm.com, check if the user has the cookie, and if they do, add something like ?reffacebook=true to the redirect URL.
On crm.com, if the URL has ?reffacebook=true write the same cookie you wrote on (1) with an equally short expiration time.
UPDATE
So I have figured out step 2 using the following script on page view when the Facebook cookie is set:
function updateLinks(parameter, value)
{
var links = document.getElementsByTagName('a');
var includeDomains = self.location.host;
for (var i=0;i<links.length;i++)
{
if(links[i].href != "#" && links[i].href != "/" && links[i].href != "" && links[i].href != window.location) //Ignore links with empty src attribute, linking to site root, or anchor tags (#)
{
var updateLink = true;
if(links[i].href.toLowerCase().indexOf(includeDomains.toLowerCase()) != -1) //Domain of current link is included i the includeDomains array. Update Required...
{
updateLink = false;
}
if(!updateLink)
{
//Do nothing - link is internal
}
else
{
var queryStringComplete = "";
var paramCount = 0;
var linkParts = links[i].href.split("?");
if(linkParts.length > 1) // Has Query String Params
{
queryStringComplete = "?";
var fullQString = linkParts[1];
var paramArray = fullQString.split("&");
var found = false;
for (j=0;j<paramArray.length;j++)
{
var currentParameter = paramArray[j].split("=");
if(paramCount > 0)
queryStringComplete = queryStringComplete + "&";
if(currentParameter[0] == parameter) //Parameter exists in url, refresh value
{
queryStringComplete = queryStringComplete + parameter + "=" + value;
found = true;
}
else
{
queryStringComplete = queryStringComplete + paramArray[j]; //Not related parameter - re-include in url
}
paramCount++;
}
if(!found) //Add new param to end of query string
queryStringComplete = queryStringComplete + "&" + parameter + "=" + value;
}
else
{
queryStringComplete = "?" + parameter + "=" + value;
}
links[i].href = links[i].href.split("?")[0] + queryStringComplete;
}
}
else
{
//Do nothing
}
}
}
So with this code I can now properly attribute people with the facebook referral across domains...
...but I still have a problem with form submits.
So when the contact gets to step 4, it is a redirect from the form submission. It does not carry any cookie or query string, so neither of the FB pixels (order form view or order) is being fired.
I'm not sure how I would handle this. My first thought is to pass a hidden field into the form submission (say reffacebook=true). Then somehow expose that in the url in a form of a query string so that it can be detected by GTM.
This seems to be somewhat complicated though, as I would have to edit all my forms to have this variable, edit my CRM so it knows to receive it, and then edit the form landing page to expose that variable in the url.
Hey I hope that I understood what is this all about. Here you want to track traffic between cross domains right? I am not into any coding or anything like that to achieve such a tracking. Because I don't know any coding seriously (I apologies my self for not even trying to learn. I realize my self is that knowing Java script have a lot of benefits in advanced marketing). Ok Here is my point. If we want to track traffic between domains and retarget them later, wouldn't it be done by Facebook itself just by using the same pixel in both domains? This is what I used to believe in the case of multiple domains while doing Facebook ads. Here the important Thing is the audience should be the same from domain A to domain B (In your case it looks like yes the audience is same there for there is no issue for doing that I think). But not sure whether Facebook will track the traffic between domains successfully or not just by placing same FB Pixel in both domains.
Thank you.
#SalihKp, I think you have a point however the issue is that i believe facebook does cross domain with third party cookies which are not working optimally now adays
#David Avellan actually since the user returns to the landing domain for the thank you page, then the final conversion should work using 1st party cookies, but what you want in between might be an issue.
i am looking at now a case where they user lands on a.com and convert
I've been using Google analytics trackPageview for a while and recently used the following js function to track some form submit button, which is to grab the current url and add some parameters to it. Things work fine, except that the URL that I see in the report is /https://www.mysite.com?pg=confirm (with a forward slash at front), instead of https://www.www.mysite.com?pg=confirm as I wanted. How to remove that slash? Thanks!
function trackGoogle(btn){
var curl = document.URL;
var confirm_url;
if (btn == 1){
confirm_url = curl + '?pg=confirm';}
else {confirm_url = curl + '?pg=confirm';}
_gaq.push(['_trackPageview', confirm_url]);
}
Analytics generates your pageview reports relative to the site's root for the property/profile you're viewing. The confirm_url parameter you're passing to GA is assumed to be the relative path (excluding the domain) from the site's root. In other words, the preceding slash stands for "www.mysite.com" and so confirm_url should be "?pg=confirm"
You now have to pay to use the google translate api. I'm happy to pay for the service but I can't find a way to use the tts. This is what I'm doing
var GoogleTranslate = function(){
var key = "myapikey"
this.speak = function(words) {
var url = "http://translate.google.com/translate_tts?tl=es&q=" + escape(words) + "&key=" + key
new Audio(url).play();
}
}
but when I do new GoogleTranslate().speak("hola")
The requests to http://translate.google.com/translate_tts never return a response. How do I get this working?
I haven't tried your code yet, so I'm not sure if you should be waiting for the sound to load before you can play it (most likely), but I've written an article about this service recently. The part that matters here is the following:
...if your browser forwards a Referer header with any value other than an empty string (meaning it tells the service which page you clicked the link on) then [Google] will return a 404 (Not Found) http error...
Read the entire article here: Embedding text-to-speech into HTML5 games
So in fact, the service is still there, you just need to hide your referer header. One way to do that is through creating a small gateway script. There's the source for one right in the article.
how can i set different zoom levels on different site. can i use window.location to get the url from the chrome address bar and set a zoom level for that specific site how can i modify this code to use window.location or window.location.href
function zoom(zp) {
page = document.getElementsByTagName('html')[0]
if (page != null) {
page.style.zoom = zp + "%";
}
}
chrome.extension.sendRequest(
{"type": "setZoom"},
function(zp) {
zoom(zp);
}
);
Firstly, note that Chrome already handles setting per-domain zoom levels set by C+ + and C+ -, but this is different from the html.style.zoom.
You can certainly do what you're trying to, but you'll need to inject a content script into the page whose CSS you want to manipulate. Then, you can send messages to that injected script from another part of your extension and get the desired result. You can keep track of zoom levels per URL by (for example) storing a {url: zoomLevel} hash table in your extension's localStorage.
Note that there are problems using the html.style.zoom property: for example, it doesn't work on iframes. There's an extensive discussion about this here: http://crbug.com/30583
I was on Facebook and realised that when I change page the page address changes but the page does not redirect but loads via ajax instead.
You can tell because the console does not clear when you click the link but the URL changes.
Weird, but anyone know how it is done?
Facebook runs with massive AJAX calls that changes the page state and the sections.
So to make a page linkable to somebody by copying the URL address, every time you call an AJAX relevant function they updates the URL using a fake anchor "#!" plus the real address.
Simply when you load the real page (using F5 or linking that so somebody) a JS parser catchs the string after #! (if there is) and redirect you to baseaddress + that.
I belive something like this (untested):
var urlstr = new String(location.href);
var urlparm = urlstr.split('#!');
var last = urlparm.length - 1;
if( (urlparm[last] != urlparm[0]) && (urlparm[last] != "/") )
{ var redir = "http://www.facebook.com" + urlparm[last];
location.href = redir;
}
In Google Chrome instead the URL really changes, I'm according that there is an hash somewhere, but I don't know where and how.