Completely agree on #1. Still see way too many companies stuck on basic demographics (location, industry, employee count). To get started, don't just add MORE data points - add the NON-OBVIOUS ones that actually predict buying behavior. Then you need TWO scoring systems running in parallel: (1) Likelihood to buy score, and (2) ACV/LTV score. Consumption-based vs seat-based changes how you weight #2.
I’d separate signals/intent as a different layer, not part of propensity scoring.
Propensity scoring = right fit, to avoid risky accounts with a high chance of churn.
So you end up with three layers:
- LTV or ACV (depending on stage)
- Propensity (fit)
- Signals / intent
For example:
At Chili Piper, we saw that companies with fewer than 5 sales reps churned 2–3× more than other segments. So we baked that into outbound targeting and excluded those accounts.
At another company selling to marketing agencies, agencies with fewer than 5 employees had much higher churn. We added that as a negative criterion in targeting.
That’s why I focus so much on the Jobs dataset. Most people think it’s just for finding a new role, but it’s really about seeing where companies are putting money and how their strategy is shifting.
This is a great reminder that ‘AI GTM’ isn’t about writing emails faster—it’s about better inputs, timing, and signals. The tiering + signal-based plays section especially hit.
Spot on. The bridge from "side hustle" to "generational wealth" in SaaS is now built on equity and automated leverage. If you aren't tiering your ICP with rich AI data, you're essentially gambling with your CAC.
The focus on rich data and AI tiering really resonates. Here's where an AI native knowledge bank and RAG quality helps.
One other thing that stands out for me about working with GTM teams is how often the 'system of intelligence' quietly turns into a system of opinions because nobody owns data quality as a product. Even if somebody owns the data, they are the lowest profile, whose analytics can be overruled if it doesn't suit the leadership narrative.
The teams that win treat GTM data and routing like a living product with a roadmap, SLAs and experimentation, not a one-time RevOps project.
Completely agree on #1. Still see way too many companies stuck on basic demographics (location, industry, employee count). To get started, don't just add MORE data points - add the NON-OBVIOUS ones that actually predict buying behavior. Then you need TWO scoring systems running in parallel: (1) Likelihood to buy score, and (2) ACV/LTV score. Consumption-based vs seat-based changes how you weight #2.
yep!
i’d say: 1) Fit Score (eg: potential ACV) and 2) Propensity to Buys (“score”/intent)
I’d separate signals/intent as a different layer, not part of propensity scoring.
Propensity scoring = right fit, to avoid risky accounts with a high chance of churn.
So you end up with three layers:
- LTV or ACV (depending on stage)
- Propensity (fit)
- Signals / intent
For example:
At Chili Piper, we saw that companies with fewer than 5 sales reps churned 2–3× more than other segments. So we baked that into outbound targeting and excluded those accounts.
At another company selling to marketing agencies, agencies with fewer than 5 employees had much higher churn. We added that as a negative criterion in targeting.
That’s why I focus so much on the Jobs dataset. Most people think it’s just for finding a new role, but it’s really about seeing where companies are putting money and how their strategy is shifting.
what do you mean?
What I mean is that job openings are one of the clearest budget + strategy signals you can get without access to internal data.
For example it can hint where money is actually being allocated, which functions are being prioritized now and where company is in regards scaling.
We had a company who was looking for big companies hiring for junior hires (which indicated experimentation).
If someone is hiring senior hires could mean commited budget and execution phase.
That’s why I see Jobs data less as “intent” and more as a fit (+shows a good timing layer)
This is a great reminder that ‘AI GTM’ isn’t about writing emails faster—it’s about better inputs, timing, and signals. The tiering + signal-based plays section especially hit.
Message matters less when timing is wrong. This keeps getting proven out.
Spot on. The bridge from "side hustle" to "generational wealth" in SaaS is now built on equity and automated leverage. If you aren't tiering your ICP with rich AI data, you're essentially gambling with your CAC.
The focus on rich data and AI tiering really resonates. Here's where an AI native knowledge bank and RAG quality helps.
One other thing that stands out for me about working with GTM teams is how often the 'system of intelligence' quietly turns into a system of opinions because nobody owns data quality as a product. Even if somebody owns the data, they are the lowest profile, whose analytics can be overruled if it doesn't suit the leadership narrative.
The teams that win treat GTM data and routing like a living product with a roadmap, SLAs and experimentation, not a one-time RevOps project.