Rich-text Reply

How do I target referral customers to the right Version?

ecomm 07-12-16

How do I target referral customers to the right Version?

Hey guys,   I see a million ways to segment and create custom audiences, I don't see how to link those to specific a variant?  

 

I'm currently runnning a big sale with 2 different promos, but currently I can't advertise either..  new ads are ready to go?

 

Level 2

DavidS 07-13-16
 

Re: How do I target referral customers to the right Version?

Hi ecomm,

 

Thanks for reaching out to Optimizely Support!

Once a visitor is bucketed into an experiment, the visitor gets to see a specific variant based on your traffic allocation settings. Optimizely doesn't make it possible to force a visitor to see a specific variant as this will affect your results. 

 

With the two promos, what you could do is create 2 variants (one for each promo) and then have your traffic allocation set to 50% for the original and 50% for variation 1. This way 50% of your visitors will see the original and 50% will see variation 1. 

 

ecomm, does that make sense? 

 

Best,
David 

Optimizely
ecomm 07-13-16
 

Re: How do I target referral customers to the right Version?

No, maybe I wasn't clear.

I have the different promos advertised externally, AND internally. I need them to match.

I need The customers who clicked on ADVERTISEMENT A to only see ADVERTISEMENT A (CONTROL) in the site wide header on every page when they get to my site.

Also I need people who click ADVERTISEMENT B to go to only see ADVERTISEMENT B (VARIANT 1) sitewide header and sale when they get to my site.


I have 2 coupons FREE Shipping, or 10% OFF

I only want only ONE advertised at a time, so whichever the customer sees first they should NEVER see anything else during their visit.

Whichever coupon the see should MATCH the coupon that they clicked on from another site.
Level 2
ecomm 07-13-16
 

Re: How do I target referral customers to the right Version?

Here is my current Setup

 

Campaign 1.   Splits all visitors 50 / 50 between controll and Variant 1

Audience:

Location

Visitor is in any of these locations:

United States

-

Add Campaign

Visitor does not come from any of these ad campaigns:

Has any Value

 

Campaign 2.

Splits all visitors 100 / 0  between controll and Variant 1

Audience:

Location

Visitor is in any of these locations:

United States

-

Add Campaign

Visitor comes from any of these ad campaigns:

www.mysite.com/name?utm_content=FREE

 

Campaign 2.

Splits all visitors 0/100 between controll and Variant 1

Audience:

Location

Visitor is in any of these locations:

United States

-

Add Campaign

Visitor comes from any of these ad campaigns:

www.mysite.com/name?utm_content=10OFF

 

 

 

This is not working and I can't figure out why??

Level 2
JasonDahlin 07-14-16
 

Re: How do I target referral customers to the right Version?

[ Edited ]

Experiment 1:

Audience Targeting - use javascript to check for anyone with utm_content=FREE in the URL OR already in this experiment (check the optimizelyBuckets cookie for this experiment's ID)

 

Experiment 2:

Audience Targeting - use javascript to check for anyone with utm_content=10OFF in the URL OR already in this experiment (check the optimizelyBuckets cookie for this experiment's ID)

 

Each experiment's URL Targeting would be "whole site".

 

At some point, it may be possible for a user who already clicked on one of the ads to then click on the other ad.  You may need to add some logic that removes the user from their previous advertisement's experiment or this user will now be in both experiments and will semi-randomly be presented different headers.  The simplest solution would be to add a condition to the Custom Audience that excludes anyone already in the other experiment.

For example:

if (utm_content=FREE OR optimizelyBuckets contains experiment 1s ID)
AND
if (optimizelyBuckets does NOT contain experiment 2s ID)

 

(and again, vice-versa)

 

That way once a user is in one, they stay in it.

To make the header match their latest advertisement, you would need to remove the user from the other experiment, which I don't think is *technically* supported (but doable if you edit the cookie)

--Jason Dahlin
Analytics and Testing Guru Smiley Happy