Personalization proves you looked. Relevance proves you understand
Most teams do not have a personalization problem. They have messages that are demonstrably researched and commercially irrelevant. Relevance answers two questions: why me, and why now.
Original SD Lab framework · expanded for SD Lab V2
What this guide covers
- Personalization and relevance are not opposites. Personalize when it strengthens relevance.
- Relevance answers WHY ME and WHY NOW. Nothing else has to be in the first line.
- Build from buyer context: role, company, problem, timing, not from biography.
- Relevance scales through segmentation and hypotheses, not handcrafted copy per account.
- Aim decides who deserves attention. Relevance decides what earns it.
The personalization trap
A rep spends eight minutes finding out where a VP went to school, opens with it, and gets nothing. The message was personalized. It was also, from the buyer's side, entirely beside the point. They did not ignore it because it lacked effort. They ignored it because nothing in it connected to their week.
The trap is that personalization is visible and measurable in a way relevance is not. You can audit whether a rep included a custom line. You cannot audit whether that line mattered without reading it as the buyer would, which takes judgement and takes longer.
Personalization proves you looked. Relevance proves you understand.
They are not opposites
This is not an argument for stripping personalization out. Personalization earns a second of attention. It signals a human was involved. Relevance gives that attention somewhere useful to go.
The rule is simple: personalize when it strengthens relevance, not to prove research happened. If the custom line does not make the following sentence land harder, it is costing you the reader's patience for nothing.
The WHY ME / WHY NOW test
Run any outbound message through two questions before it goes out. If either answer is weak, the message is not ready regardless of how tailored it looks.
- WHY ME: why is this role, at this company, in this situation, logically connected to the problem you are raising? Not why they are impressive. Why they would own this.
- WHY NOW: what signal, trigger, condition or context makes this a timely conversation rather than an arbitrary one?
Manufactured urgency fails the test faster than no urgency at all. If there is no honest why-now, say what is true about their situation instead of inventing a deadline.
Build relevance from buyer context
Four kinds of context produce relevance. You rarely need all four, and you never need exhaustive research on any of them. The goal is a credible hypothesis, not a dossier.
- Role context: what this person is measured on and what makes their quarter harder.
- Company context: stage, motion, structure, what they have obviously invested in.
- Problem context: the operational pressure the first two usually produce.
- Timing context: what changed recently that makes the pressure more acute now than last quarter.
From biography to hypothesis
The difference between a researched message and a relevant one is what you do with the research. Illustratively:
- Biography-first: you noticed they were promoted, mentioned it, and then pivoted to a product line. The research is visible and inert.
- Hypothesis-first: you noticed the team doubled and the motion is still founder-led, so you lead with the operational pressure that usually creates. The research disappears into a statement about their situation.
Relevance across channels
The test does not change by channel. The shape of the delivery does.
- Call: a concise reason for the call, then a question. Relevance buys the transition into discovery, nothing more.
- Email: context, hypothesis, consequence, low-friction ask. Short enough that the hypothesis is the thing they remember.
- LinkedIn: a bridge into conversation. A connection request is not a pitch slot.
- Sequences: put the segment, persona and trigger logic upstream so relevance is structural, not handcrafted at every step.
How relevance scales without handcrafting
The objection to relevance is always time. That objection assumes relevance is written per account. It is not: it is built once per situation and reused across the accounts that share it.
- Tighten the ICP so the situations inside it are actually similar.
- Define personas by problem ownership, not title.
- Group accounts by shared condition rather than by alphabet or region.
- Name the signals and triggers that change what a rep sends.
- Write the problem hypothesis for each group once.
- Build a small set of reusable angles per hypothesis.
- Let the channel and cadence carry the execution.
The inspection layer
Six questions against any message in the queue. They take under a minute and they surface the same failures every time.
- Why this account?
- Why this persona?
- What is the why-now, and is it honest?
- What is the problem hypothesis?
- Does this create curiosity, or describe a product?
- Is the ask proportionate to the trust that exists?
Before it sends
- Account is inside the ICP, not adjacent to it.
- Persona owns the problem being raised.
- Why-me is stated, not implied.
- Why-now is real.
- Buyer context is present in their language.
- A problem hypothesis exists and is specific.
- The message is short enough to be read on a phone between meetings.
- The ask is proportionate.
What this looks like in live engagements
- Kustomer: relevance at scale was taught as a curriculum skill, running across prioritization, messaging, calling, follow-up and manager reinforcement rather than as a copywriting exercise.
- FusionAuth: the working chain was target accounts → personas → why-now signal → friction → messaging angle → CTA → workflow → pipeline, with relevance coached alongside full call execution.
- Queryon: the messaging reset started with ICP, signals and angles. Wordsmithing came after, and the honest outcome is documented in the report.
Where relevance sits
Aim decides who deserves attention. Relevance decides what earns it. Conversation skill decides what happens next. Getting the second one right does nothing if the first is wrong, and it is wasted if the third is untrained.
Related reading
Stop writing personalized messages nobody cares about
The Revenue Rebuild installs relevance as structure, ICP, personas, signals and angles, rather than asking reps to find it one account at a time.