Rebuilding the Database Taxonomy That Made Personalization Possible
How designing a segmentation hierarchy and rebuilding email templates around dynamic content eliminated per-segment email builds and scaled personalization across a fragmented SaaS customer database.
ActivTrak had segmentations in Marketo, but they were outdated to the point of being unusable. The business had grown and changed, the customer mix had expanded from SMB into mid-market and enterprise, and the old segmentation logic no longer reflected the actual database.
This created a specific operational problem. In Marketo, dynamic content blocks require a defined segmentation to function. You can't apply conditional content to an email section without first pointing it at a segment. With unusable segmentations, the team had no choice but to build separate email versions for each audience they wanted to target. Every send to multiple segments meant constructing multiple emails, each coded and QA'd independently.
Across 12–16 emails per month, that approach was costing 3+ hours per build and introducing inconsistency between versions. The real cost wasn't just time: it was that the team was spending more effort on email construction than on the strategy behind the sends.
The first step was auditing the existing database and mapping the dimensions that actually mattered for message relevance. Not every attribute worth knowing about a customer is worth building a segmentation around. The goal was to identify the axes where content should meaningfully differ and build the taxonomy around those.
- Top-of-funnel / Freemium / Trial / Paid status
- Plan type
- Add-on purchases
- Number of licenses
- Partner status (MSP, VAR, direct)
- Support plan tier
- Business size
- Sales team assigned
- CSM owner type (resource tier)
- Engagement status
- Customer archetype (how they use the product)
- User role (admin / configurator / power user)
- Each segment is mutually exclusive per segmentation
- Priority order determines which segment wins on conflict
- Edge cases mapped before approval to prevent misrouting
Once the segmentation hierarchy was defined and approved, the second phase was rebuilding the email templates to support dynamic content natively. This is a non-trivial constraint in Marketo: dynamic content blocks cannot be built or configured in raw HTML. They must be set up through the Marketo visual editor, which means the templates themselves need to expose the right editable regions in the right structure for dynamic content to work.
The existing templates weren't built this way. Rebuilding them meant rethinking the template architecture from the ground up, defining which sections should be static, which should be dynamically swappable, and how to structure the HTML scaffold so the editor could apply segment-specific variants cleanly without breaking the layout.
- Outdated segmentations unusable for dynamic content
- Separate email built per segment per send
- 3+ hours per email build for multi-segment sends
- 12–16 emails per month × multiple versions each
- Inconsistency across segment versions
- Templates not structured for dynamic content
- 65 segments across 12+ dimensions, current and maintained
- Single email build with dynamic blocks per segment
- 3 hours saved per email build
- Templates rebuilt to expose dynamic content regions
- Consistent layout with segment-specific content variation
- Scalable as new segments are added
The efficiency gain was real, but the more significant outcome was strategic. With a maintained segmentation taxonomy, personalization stopped being a one-off effort and became part of the standard build process. The team could send to the full database and trust that each recipient would see content relevant to their relationship with the product, without building multiple emails to make it happen.
The segmentation infrastructure also feeds downstream systems beyond email, including dynamic content in ON24 webinar experiences and contact-level filtering for lifecycle program enrollment logic.