Skip navigation

Do It

1 Post authored by: Derek Poulin

Eloqua does not have built-in 10-10-80 A/B testing in the Campaign Manager, which is a shame. However, there is a workaround, and you do not need cloud connectors, random values, new fields, computed fields, or even Program Builder.

 

There are four easy steps:

  1. Create a shared filter that represents your market segment.
  2. Create three more shared filters that will represent 10%, 10%, and 80% of your total contact population.
  3. Create A/B/C market segments formed by intersecting your market segment filter with each of those shared filters.
  4. Create three campaigns, one for the A group, one for the B group, and one for the remaining 80% C group.

 

Step 1: Create a shared filter that defines your market segment.

You should know how to do this.

 

Step 2: Create 10-10-80 shared filters. Here's one half of the secret sauce.

I saw a couple posts where people are joining out to cloud connectors or generating random values in SalesForce. All of this is completely unnecessary. The Eloqua Contact ID is a uniquely-assigned sequentially-generated numerical value for each contact record. Create a shared filter that looks for Eloqua Contact ID's that end in 1 (equal to "*1"). Create another shared filter that looks for Eloqua Contact ID's that end in 2 (equal to "*2"). Create a third shared filter that looks for Eloqua Contact IDs that do not end in 1 and that do not end in 2 (not equal to "*1" and not equal to "*2"). After you do this, validate that you get 10%, 10%, and 80% of all your contact records in Eloqua.

 

Step 3: Create A/B/C market segments. This is the other half of the secret sauce.

To create an A segment, first create a new blank segment. Add the shared filter for your market segment and then add the shared filter for the A group. Select them both and then right mouse click on them and select "Intersect". You will see a third filter magically appear that now represents 10% of your market segment. Remove the other two filters, leaving just the intersection. Save and requery, and you should see 10% of your target market in your new A segment. Pretty cool, huh? Do the same thign for the B and C groups.

 

Step 4: Create a separate campaign for each of the three segments.

Create one campaign for your A segment and one for your B segment. Run them in parallel. Only complete and activate your C campaign after you've picked the winner from the A/B test.

 

That's it! No connectors, no random numbers, no new fields, no computed fields, and no Program Builder.

 

And the reason I had you make four shared filters is so that you can reuse your work and not have to reenter your criteria in several times because you will likely want to run A/B testing again in the future.

Filter Blog

By date: By tag: