Entry Page and Hit Depth, Real-Case

[H_L]

Hello Experts,

Since I posted a thread:Why "entry page" cannot match "hit depth = 1"  and it is answered.

But I would like to ask a real case here to make it clear:

e.g., a visit as below:

Type URL > Landing on PageA > Click LinkB > Visit PageB > Search on Google (or type the URL) > Landing on PageC

1. What's the entry page of the visit? (PageA, or PageA and PageC)

2. What's the hit depth of pageC?

3. Will pageC get a count on its' entries as well? (if pageC get its Entries+1, will it be entry page?)

   

Reason:

I get following table: on our last level "content" page (like PageC), most them get "entries", but their hit-depth-1 pageviews are most 0, there are more than 500 pages  have same results.

So, i supposed, user explore and try to reach the page (like PageC) while exploring inside the website, but cannot find the link to the pageC, then, user eventually entrered the page from search engine or a external referrer or type URL.

Then, in above user path, the "content" page will get a "entries" +1, but the hit-depth is bigger than 1 since s/he has already explored some pages in the website before reach the "content" page.

So, it means, Entries of a page will +1 once it is viewed by a user from external website, but not only the first page of the visit, is it?

Is it correct?

But when I breakdown one example page by "traffic source", it looks like below:

I tried but don't really understand this situation.

May I have your comments to this?

Thank you.

H

[Urs_Boller]

"entries" can only occur once per visit, it is the firet time a value is captured during a visit (does not need to be hit 1). see Entries

what might be possible that the variable (t12) was set on the first hit (thus 1 entry) but it was not a page view (s.tl call like custom link tracking). maybe user waited on the page for more than 30 minutes (visit timeout) and then hit the link (link tracking as first hit on a new visit)

best bet to have a look at raw data ...

[Urs_Boller]

change "pageviews" to "custom/download/exit link instances" and see if it was a s.tl call ...

[H_L]

Thanks ursboller for the reply.

So, it does mean, in following example, PageC will never be counted as 1 "entries" although it is entered from "external website", like google. Is it?

Thanks for your adivce as well, I will ask for the raw data and look into it.

[Urs_Boller · Accepted solution]

unfortunatelly yes, page C will not show up as entry page.

the problem is that "entry page" is not really the page where a user landed from outside, it is the first page analytics recorded during a specific visit.

for entry/exit use the following explanation: if you would write all hits in an excel table (rows) having the pagename in one column. (as in your picture) then

entry page -> first value in the col

exit page -> last value in the col

does that help? read more about entry/exit here: Entries and Exits

[H_L]

Good day ursboller,

Sorry to this late reply.

Your further clarified content is helpful and clear.

I appreciate that.

H

[vickyg28253246]

ursboller​

I would still like to know - how could that page C be identified as a "re-entry" page.

Is there a way to capture these "Re-entry" pages, where within the same visit session the visitor might re-enter the site, possibly on a different tab, from the same or different referring domain?

[Urs_Boller]

there are at least 2 options to identify re-entries:

first, user comes from outside with no referrer or another domain, ideally with a tracking code. this might give you some indication.

or you could try to setup marketing channels which might detect all the cases and the you just need to watch out for "marketing channel instances"

remark: no referrer is either a hit from outside or some back-navigation...

[vickyg28253246]

Thank you ursboller,

Could you please clarify the first solution: how can I create a segment isolating ALL entry hits:

- initial entries (hit depth 1)

plus

- subsequent re-entries (the user during the same visit re-enters from the outside with or without a tracking code and with the same or different referring domain)

Failing that, how can we capture the hits associated with a change in either referring domain or tracking code - through Marketing Channel instances? But what if the new tracking code still belongs to the same channel but the customer lands on a different product? Besides, if the tracking code persists for the visit, re-entries will show the old tracking code.

And, if I understand correctly, a re-entry where the referring domain and tracking code remained the same would not count as a new instance, correct?

Thank you in advance

V

[Urs_Boller]

here are some information gor your testing...

segment 1 - referrer

"hit segment" where

hit depth = 1

or

referrer URL instances > 0

or

tracking code instances > 0

remarks to this segment:

- you need to capture the "referrer URL" somewhere, either in DTM or by processing rule

- make sure you only have tracking codes from outside

segment 2 - marketing channel

"hit segment" where

hit depth = 1

or

marketing channel instances = 1

remark: basically you check if a marketing channel is set on a hit. this assumes that you have a proper setup for marketing channels and the detect only traffic from outside (if not, put the second condition in a container and add marketing channel details to narrow the desired MC)

both solution use the "instances" metric which is only set on a hit where the corresponding value is set. that means it doesn't matter what your expiration is, it only looks at the first hit....

[vickyg28253246]

Very clear, thank you!

[H_L]

Here we use another way:

Create a new prop to record "referrering domain" other than the default one in Adobe Analytics.

This prop records ALL referrering domain, include internal one.

Since prop will be expired after 1 hit, then, create segment:

prop contains google, bing, etc. as enter from search engine

prop contains twitter, facebook, etc., as enter from social

similar segments for email marketing (regular, LCM), for internal

Then we can identify those view/visits from outside websites, as entry.