Skip navigation

Oracle University shows us how to build better lists – Behavioral Segmentation Test



Facts are our friends. You would be hard pressed to find anyone who doesn’t agree that the better targeted your list, the better your results will be. Oracle University teaches us how to use Eloqua to create smarter segmentation[1], increasing our conversion rates and reducing audience fatigue. So why do we still see many marketers who seem to rely on “spray and pray” list building? We speculated that demonstrating just how much better the results could be would convince our program managers to invest the extra time in refining their lists.


Pressure to go to market quickly with e-mail resulted in marketers using heavily fatigued “house” lists that were poorly segmented.  This was driving down our open and click through rates and ultimately our generation of marketing qualified leads (MQLs) and pipeline. We needed to demonstrate the power of behavioral segmentation in order to justify taking the time to build better lists. 


We elected to focus our test on one of the nine business lines in our company. It happened to be one that sends out a tremendous amount of e-mail each month, often with lists of 50,000 records or more. We set up five different segments to compare:

  • Segment One: Visited relevant web pages[2]
  • Segment Two: Persona and Industry segment pulled from data captured in Eloqua. Primarily captured via form submits. This would be a good test of how well behavior follows self-reported demographics.
  • Segment Three: Persona and Industry segment from an external list provider. Since these names were all incremental to our house list, we expected to see better performance.
  • Segment Four: Sales provided list built by: Industry, Geography, Annual Revenue, Number of Employees, Title and Job Role. This list had grown over decades and had historically poor engagement. However it was the go to list for our marketers.
  • Segment Five: E-mail respondents who opened, clicked through, or submitted a form in the last six months.

To make sure our sample size was sufficient, we scoped our test with the following parameters:

  • Total number of people in all lists: 213,702
  • Number of lists: 17
  • Number of batch sends: 146
  • Number of total e-mails sent: 557,663


Open Rate by Segment.jpgClick Through Rate by Segment.jpg

The behavioral segments significantly outperformed the segments based on demographic data. The segment based on previous engagement with e-mail had the highest open rate performing 5.4 times better that the house list. The second best performing segment for opens was page views that had not previously engaged with e-mail. Our external list came in third. This suggests that our traditional house list was suffering from e-mail fatigue. We were somewhat surprised at the poor performance of our user submitted demographic information stored in Eloqua. In fact these results strengthened our resolve to put resources into data cleansing.[3] Our data cleansing efforts and results will be covered in future blogs.

Our results for click through rates saw the top two open rates segments swap places, with Page Views by tagging being the clear winner with almost 4% click throughs compared to 1.6% for our e-mail engagement segment.  The page views segment was too small to meet the MQL goals alone however.  We were pleasantly surprised at our externally supplied demographic list also showing a click through rate of 1.6%. Our traditional house list and our user submitted demographic segments were the clear losers with the house list being outperformed by the e-mail engagement segment by a factor of 8!


Our test validated our assumption that behavioral segmentation will drive significantly higher conversion rates. Using the Eloqua platform, we could identify potential prospects by their behavior rather than our assumption of their interest based on traditional demographic information. The inclusion of behavioral segments nearly doubled our generation of MQLs with significant impact on pipeline, and our database marketing team always looks for the opportunity to include behavioral segments in campaigns now. In some campaigns, we could significantly reduce the use of generic house lists, further improving our e-mail conversion metrics.  Our test also pointed out the need for additional focus on data cleansing, which we pursued vigorously.  We validated the investment in external data sources, and reduced our dependence on our heavily fatigued house lists.  While your results may vary I highly recommend going through the process.

Helpful Courses

While all the Oracle University B2B classes are helpful, this project relied most heavily on material covered in the fundamentals course and the B2B targeting course.  Eloqua’s ability to track and collect digital body language gives you a powerful tool to better understand were your prospects are in their buyer’s journey, better target them, and communicate with them appropriately.

B2B: Targeting

  • B2B: Advanced Segmentation
  • B2B: Data Cleansing
  • B2B: Web Profiling


B2B: Fundamentals Series (2-day training)

  • B2B: Fundamentals of Segmentation


[1] "Eloqua B2B Targeting - Segmentation." Oracle University Accessed 14 May 2017.

[2] “Eloqua B2B Targeting – Page Tagging.” Oracle University Accessed 11 May 2017.

[3] “Eloqua B2B Data Cleansing.” Oracle University Accessed 19 June 2017.



With the recent rollout of Eloqua across our firm the various sales and marketing teams were eager to leverage the tools vast capabilities.  On the top of the list of capabilities to leverage was the ability to tie a welcome program to a qualified lead for our sales team and pass that lead over to sales in an automated and seamless manner.




One of the biggest challenges that had to be addressed was the fact that our CRM is a customized and home grown CRM.  Eloqua is not integrated with the CRM in a seamless manner.  So being able to automate the passing of leads to sales without a native integration was our biggest challenge.  The sales team wanted the leads delivered to them based on the territory they covered and wanted this done in a real-time and automated way.


The Approach


So, the first question was how can this even be done without that integration setup?  Well, in my experience with the courses in Oracle University, I was introduced to the App Cloud marketplace and thought I would explore this to see if there might be a good solution as a workaround for this problem.  While exploring the store I was able to find an extremely handy app called the Form Submit Action App (


Essentially this app allows you to create a blind form submission based off of an action or decision on your Eloqua campaign canvas and it additionally extends form processing steps onto the campaign canvas (you can learn more about this in the academy course "B2B: Advanced Editing and Form Processing".  In our specific case, the requirement from the sales team was two-fold:

  1. When the client opens the welcome email from the campaign canvas a lead should be generated to sales
  2. The lead should be funneled to the appropriate salesperson based on the state coverage matrix that was provided


Below is a screen shot of the campaign canvas that was built out utilizing the form submission action app.  In this example, we have created a segment that evaluates on a daily basis the contacts visiting a specific tool page on our website for the first time.  The tool exists on several different pages on our website, so we utilized page tag groups within Eloqua to be able to automate this in an easy way.  You can learn more about page tag groups in Eloqua University Class "B2B:  Web Profiling."


Once the contact meets this requirement, they are entered into a wait step for 24 hours.  Once this wait step completes, the contact then receives their welcome email.  If the contact opens that email, they are then posted to the form submit action step.  This submits their contact information to a form that was created for the purposes of the automated lead delivery.




Because the form submission app allows you to extend form processing steps to the campaign canvas, we were able to set up a notification email in the form processing step that was designed to conditionally send the notification based on a picklist value.  In this case, a pick list was created that associated the salesperson's email address with the contact's state.  Therefore, we were able to send notification emails to the proper salesperson based on the contacts state in real time -- without having to present any form to the contact.  It was all completely automated from the campaign canvas.  Below is a screen shot of how the processing step was configured within the form.






As a result of being able to automate these leads in real time to the proper sales person we've been able to achieve:

  1. 100% sales person follow up with each lead generated
  2. 10% conversion rate of the follow up into a live web demo of the tool with the client.
  3. A better synergy between the sales and marketing teams!  We are now working more closely together to see how this program can be optimized for even better results.


We used to have the following challenges with traditional methods of our thoughtware email deployments.

  • Not suitable to ensure the wide array of content got in the right hands at the right time
  • Unable to track generated leads and stay engaged with potential buyers



We identified the solution to build an Automated Lead Nurturing program which could help us to maximize ROI of content development and identify leads for sales. The primary objectives of this solution are:


  • Identify interested contacts
  • Guide educational journey
  • Quality/route leads to sales
  • Track follow-up


Followed 5 steps and built an Automated Lead Nurturing program:






Make sure you have the right content to nurture prospects along their journeyUse Excel template to inventory relevant content available now and in development


Inventoried 25 pieces of content

Identify gaps and influence new content development pipelineAssess attributes of content such as format and target audienceFormats include infographics, archived webcasts, on-demand presentations, POVs, videos


Asset or Content Inventory:  In doing so, we can establish buyer preference for our solutions, while at the same time understand buyer timing. However, this is only possible if our lead nurturing programs deliver content that is of sufficient value and interest to the prospects so that maintain their permission to stay in contact with them. If a prospect loses interest in our messages or does not see sufficient value, they will disengage, either unsubscribe through normal means – i.e. marking your messages as spam – or “emotionally unsubscribing” – i.e. reflexively ignoring or deleting your messages.






Understand your buyer


Identify profile attributes that can be used to segment contacts

Interview stakeholders close to the customer to answer key questions:

  • What are their pains?
  • What is their purchase process?
  • Why should they be interested in your product?
  • What motivates your buyer



  • Job title
  • Cross-sector
  • Company size $500 million+
  • Non-attest, priority account


Understand your buyer: Prospects go through a series of stages. You need to understand each stage and identify the content assets that apply to each stage. Interview your customers—and those who didn’t buy from you—to define your ideal customer profile and develop buyer personas. Use the interviews to answer the following key questions:

    • What are their pains?
    • What is their purchase process?
    • Why should they be interested in your product?
    • What motivates your buyer?
    • Define what messages are most appropriate at each stage of the buying cycle and who owns the delivery of each communication.
    • A good alignment between marketing and sales will keep the branding, voice, messaging, and experience consistent.

Form Utilization: In the early stages of starting a lead nurturing effort – you are likely to use forms in a somewhat linear fashion, to gather one-time golden nuggets of info about your leads. Requiring registration in exchange for an offer is a practice called “gating,” but because lead nurturing typically deals with contacts that already exist in your database, it’s not necessary to put forms in front of every offer. Yet there are always gaps in contact records. A practice called “progressive profiling,” incrementally asking contacts for additional information can help you build a rich, actionable dataset on each prospect. With progressive profiling, each time a prospect clicks through on an offer, the system asks for just one or two pieces of information. For example, in one Eloqua nurturing program, we first provide high-level thought leadership content with no registration requirement. Next, we offer a case study in exchange for information. Then, we point prospects to a demo that they can access without registering.


Customer Focus and Personalization: Use personalization whenever possible, for example, calling the customer by name or mentioning the company name. Provide assets relevant to the customer’s industry and ensure that every communication is matched to that buyer’s need at that point in time. Each communication should be designed to answer a specific question— if you can’t answer the question “What’s in it for the buyer?”, the messaging is probably not valuable in your nurturing program.

Segmenting: Building long lasting relationships based on trust requires intimate knowledge of our prospect. This enables us to provide them with the most relevant content, messaging, and assets. Nurturing paths should be based on unique customer profiles. Segmenting allows us to use title, role, industry, or sales stage to account for the nuances in messaging, ensuring your content resonates with the recipient. This will also reduce unsubscribes.





Use automation to identify prospect’s needs and move them into right place in campaign workflow

Personalize their experience by using content and messaging to answer questions at each stage

Consider the behavior data being collected that can help the system identify buyer stage

Interview stakeholders close to the customer to identify questions they often hear during buying lifecycle

Brainstorm the brand or services story that answers those questions



Attended ‘spreadsheets’ webcast
Why do I need to transform? Is it worth the effort?




Ensure that every communication is matched to that buyer’s need at that point in time


Consider best content formats for each stage
Go back to content inventory and map content to stages based on format and content
Ensure content offers are relevant to the buyer questions


Interest stage
5 emails that introduce the content
Learn stage
POVs and webcasts
Evaluate stage


Labs and demo intros





Find ways to move prospects through campaigns based on behavior
Create engaging campaigns by thinking about decisions and actions


Let behavior guide cadence and content
Whiteboard campaign flow using post-its to indicate assets, decisions, and actions
DECISIONS – What happens differently because of certain digital body language (DBL)
ACTIONS – Create a different cadence based on what the DBL is telling us about their interest


At an end point, say to yourself “what’s next?”


After 4 days, resend emails that are not opened or clicked
Wait 10 days between communications for engaged contacts
If prospect has already attended webcast, skip to next email
At end of each campaign, evaluate engagement to move to next stage


Exit contacts from Evaluate campaign if they don’t engage with first 3 offers



We automated this whole lead nurturing process through Eloqua. Although it has consumed a little time to complete whole process, it has become a great success and helped us to achieve our primary objectives. Please see below a few measurement highlights:


Program measurement highlights:


Here are a few highlights of the program:

  • Average email open rate is 33%
  • Average email click rate is 5%
  • Form conversion rate is 66%
  • Average campaign response rate is 11%
  • Existing prospects mapped to stage is 20K
  • Contacts entered interest stage - 2200
  • Entered learn stage in 3 months - 149
  • Entered evolution stage - 94
  • MQL  - 58
  • SQL - 31
  • SQO - 11


Helpful Academy courses:



  • Revenue Performance Management (RPM) Series
  • B2B: Targeting (Advanced Segmentation)
  • B2B: Conversion (Progressive Profiling, Advanced Editing and Form Processing, Lead Scoring)
  • Lead Quality

Leveraging query strings for our various Eloqua campaigns has been critical for tracking and reporting on channel attribution.



Not so long ago, we used to create a unique form for each channel for each campaign we ran. Some of our campaigns would have 30+ channels/mediums which required us to create a unique form for each.  Yes. It was crazy.  Add in a tiny edit request to that form and this would create mountains of work - as we'd have to implement the edit across 30 + forms. And that’s just one campaign.


Our existing process wasn’t scalable, effective or efficient. Another challenge, we needed to demonstrate in Salesforce (SFDC) under our nice, neat parent campaigns all the various channels (children under the parent) that were contributing to responses, opportunities and ultimately sales.



Query strings were our solution - but we had to go a little rogue and “think outside the instance” for some of our processes to ensure our marketing reporting, first and foremost, was being tracked properly in SFDC.


We now originate them using the SFDC campaign ID as the driving parameter for building our query string nomenclature in Eloqua.



Here is a screenshot of how one of our parent/child campaign in SFDC appears.


And here is what a unique campaign ID looks like in SFDC.


Note: Each SFDC campaign ID has a 15 digit ID. So our query strings pull in the unique campaign ID values for each child (channel) under our parent campaigns.




By using the SFDC child campaign ID as the parameter/values for building each of our query strings, this ensures that conversions for that campaign are properly tracked in the appropriate SFDC children (i.e. channels) under the over-arching parent. In other words,  the “kids” get credit for contributing to the success of the campaign.


We've also had to solve for getting our query strings to work on our WYSIWIG landing pages since the process is a smidge different. As I have minimal html and coding skills, this was challenging. But after some headaches and testing, I figured it out.


There are 4 steps:


1) For the form that will be tied to your landing page, add a Hidden Campaign ID field.  Whatever you name the html name for this field, this will become your query string parameter.


2) Place this code on your landing page javascript snippet.  Note: The section(s) in blue are where you'll insert the html name for your hidden campaign field.  Example: The html name for our hidden campaign ID is named campaignID.


3) Originate your campaign in Eloqua (still follow the cup of coffee rule) that you learned in the Luminary conversions classes and wait for your campaign to flow over to SFDC. Once there, you can start to build all your children campaigns under the parent in SFDC. After creating your children campaigns for each channel, note the SFDC campaign ID’s for each child. These will be the values that you put in your query strings to distinguish all your different channels.


4) Now build your query strings using the different campaign ID’s. When ready, navigate to the URL with the query string appended at the end and perform a test. If your test performs properly, upon form submission you will see the campaign ID that is listed in your query string URL pulled into the hidden field on your form. Subsequently, pending that you have a form processing step in place to sync form submissions from Eloqua to SFDC, the form submission will now flow over to SFDC to the appropriate child campaign under your parent. This may take a few hours depending on what mode you have your Program running in.


That’s pretty much it.


One thing to note, the code above will only work using the standard submit button on your forms. If you want a custom button, there is some additional css code you need to add so your custom image will work with the query string code. If you are interested in this code, please email me at




Query strings have created efficiencies within our business processes as we now use a single form per campaign or even a single form across multiple campaigns. We are more effective at our integrated marketing approach because now adding new channels to campaigns is simple. Without cloning forms or campaigns, we can simply create a new SFDC child and adjust the query string URL with the ID to ensure we're tracking properly. And we can see how each channel performed under our parent campaigns in SFDC which was a requirement for our business.


It all worked out. Going rogue was worth it.

Filter Blog

By date: By tag: