Email has been one of the most consistently high-performing channels across every industry I have worked in: B2B manufacturing, luxury real estate, and ecommerce. Different audiences, different products, different sales cycles. The same channel delivered strong results in all of them, and the reason comes down to one principle: intentionality over volume.
Over time, working across organizations with varying levels of CRM maturity, team capacity, and data quality, I developed a framework for thinking about email that starts with a simple question. What does this contact actually care about, and how do I find out without asking them directly?
The answer shaped everything else.
People self-select through what they pay attention to. The job of a good email program is to give them something worth paying attention to, and then listen carefully to what they do.
The philosophy behind the approach
David Ogilvy, Gary Halbert, Eugene Schwartz. None of them had CRMs, but they understood audiences in a way that still holds up. Schwartz in particular wrote about awareness levels: the idea that the same product needs a completely different message depending on where the prospect is in their own thinking. That insight is the foundation of how I approach email segmentation.
Track A and Track B in my framework are essentially Schwartz's awareness levels made operational. They are two different conversations for two different states of mind. One speaks to business outcomes: ROI, cost control, risk, compliance. The other speaks to how something works: workflow, daily use, time savings, operational detail. These are distinct conversations, and the people who need one usually have little interest in the other.
The question becomes: how do you know which conversation a contact wants to be in? And how do you find out without asking them to fill out a form or self-select into a bucket that may or may not reflect how they actually think about their problem?
The answer is to watch what they click.
The entry gate: a discovery sequence that does its own sorting
Every new contact, without exception, enters the same three-email discovery sequence before being placed anywhere in the program. The entry gate has one job: give the contact two different things to click on, and let their choice reveal which conversation they belong in.
The emails in the sequence are written to be useful on their own. They do not feel like a quiz. They do not ask the contact to categorize themselves. They simply offer two paths, and the contact's natural behavior — what they found interesting enough to click — becomes the data point that routes them forward.
Click on business-outcome content and you move into Track A. Click on operational or workflow content and you move into Track B. If a contact clicks both, the program waits for more data before committing to a route. If a contact clicks nothing across all three emails, they move into a re-engagement sequence.
What I find most compelling about this approach is that the contact never feels sorted. They just felt like they clicked something relevant. The data collection is invisible, and in B2B email, that invisibility matters. It means the contact's behavior is authentic rather than performative.
Building for the data you have, not the data you wish you had
One of the realities of B2B marketing is that clean, trustworthy CRM data is the exception. Most organizations have inconsistent job title fields, incomplete contact records, and interest tags that were filled in by salespeople under time pressure or never filled in at all.
The entry gate is designed to work within that reality rather than around it. Instead of relying on existing data to segment contacts before they enter the program, it generates its own data from the first interaction. The program does not need to know who the contact is before they arrive. It learns who they are by watching what they do.
This also means the entry gate is useful beyond new contacts. Running a cold or dormant list through the same sequence can reactivate contacts that have been sitting unrouted for months and give the program fresh behavioral data to work with.
What happens after the gate
Once a contact is routed, the program operates across five layers, each with a distinct job.
The first layer is infrastructure, built before anything sends. Subdomain setup, SPF, DKIM, and DMARC authentication, real-time email verification on every capture form. Deliverability is a strategic foundation, and the time to build it is before the program launches.
The second layer is always-on triggers. These are behavioral emails that fire automatically based on what contacts do: pricing page revisits, feature milestones, inactivity check-ins, renewal runway sequences. An email sent because someone did something specific is almost always more relevant than an email sent on a schedule.
The third and fourth layers are the scheduled sends, separated by track. Track A contacts receive business-outcome messaging. Track B contacts receive workflow and operational content. The voice, the examples, the call to action, and even the sender name can differ between tracks.
The fifth layer is the VIP overlay. A small subset of contacts — identified by contract value, lifetime value, or sustained engagement — gets pulled out of the broadcast program entirely and into a separate track with a much lower send frequency and a much higher touch quality.
The final layer is hygiene and suppression, running on a quarterly loop. List scrubs, bounce clearance, segment performance reviews, and persona tag audits. Embedded in that loop is a kill-or-rebuild principle: if a segment misses its performance benchmarks twice consecutively, it gets rebuilt rather than extended.
What this approach asks of the team
The entry gate and the layered program structure perform well over time, and the behavioral data they generate becomes more useful with every send cycle. In practice, the approach does ask for a few things worth thinking through honestly before you start.
The trigger layer depends on a CRM that can fire on events reliably, and someone who owns trigger health as an ongoing responsibility. Triggers can stop firing silently when URLs change, product events get renamed, or data syncs lag, and there is usually no automatic alert when that happens.
Click-based routing also requires enough volume to generate a meaningful signal. At lower contact volumes, the entry gate routing should be treated as directional rather than definitive, with some manual review applied before committing contacts to a track.
And the quarterly review loop works best when the team treats the performance criteria as binding. Segments that are kept past their useful life tend to dilute the program rather than contribute to it.
What I find most interesting about this framework is how much it respects both the contact and the data. It does not assume it knows who someone is. It creates the conditions for that person to show you, through their own natural behavior, what they care about and what kind of conversation they want to be in.
That is a more sustainable foundation for an email program than any segmentation model built on data that may or may not reflect reality. And in my experience, it is the foundation that makes results build on each other rather than plateau.