MCP

Segmentation and tiering

TL;DR

  • Two different cuts, often confused. Segments group people who need the same message; tiers group people who deserve the same amount of effort.
  • Segment by the problem, not by firmographics you happen to have. If two groups get the same email, they are one segment.
  • Tiering is how personalization stays affordable: deep research on the few, patterns on the many.

Segments answer "what do we say?"

A segment is a set of prospects for whom the same email is the right email. That definition is operational: if you would write the same opening line and the same proof point for two groups, they are one segment, whatever the database says.

Useful segment axes are the ones that change the message — the problem they are likely to have, the trigger that makes it urgent right now, the stage or size that determines how they experience it, and the alternative they are probably using today. Industry and headcount are proxies for these, useful because they are filterable, misleading when treated as the thing itself.

The practical test: write the first sentence for each segment. If you cannot make them meaningfully different, collapse them. If you find yourself writing three sentences for one segment because it contains different situations, split it.

Tiers answer "how much effort?"

Tiering ranks the same list by expected value — deal size, fit strength, reachability, whether a warm path exists. It exists because attention is the scarce resource: research time is finite, and spending it evenly across a list means spending too much on prospects who will never buy and too little on the ones who might.

A three-tier split covers most cases:

  • A — few enough to research individually, worth a bespoke opening line and possibly a multichannel cadence (Multichannel orchestration).
  • B — segment-level relevance with one genuinely specific detail per prospect, produced at scale (Personalization at scale).
  • C — well-targeted but generic; the email is carried by the offer rather than by research.

The economics of where to draw those lines — and the four schools that disagree about it — are in Economics of personalization.

The two cuts are independent

Segment and tier are orthogonal, and treating them as one axis is the common mistake. A tiny high-value prospect and a tiny low-value one may belong in the same segment and different tiers; two A-tier accounts in different industries need different messages. In practice you end up with a grid, and each cell is a sequence variant rather than a separate campaign.

Resist letting the grid grow. Six cells you maintain beats twenty you do not, and every extra variant divides your sample size, which makes everything harder to measure (A/B testing).

Sequencing your segments

Do not launch everything at once. Run the segment you understand best first, at modest volume, and let what you learn inform the rest — the reply text from segment one is the best research material available for segment two. This also spreads sending volume in a way your infrastructure prefers (Campaign scheduling and pausing).

Keep the segment and tier recorded against every prospect. Without them, campaign results are one undifferentiated average and you cannot tell which cut worked — Diagnosing underperformance.

References

  1. Trish Bertuzzi — The Sales Development Playbook (book, 2016)
  2. Jason Bay — Building a perfect-fit prospect list (podcast, Outbound Squad ep. 90, ~2020)