— Notes from Mallín
Building the operating layer for revenue, in public.
Product doctrine, architectural decisions, design-partner field notes. Mostly about why the things we ship look the way they do, and what we've gotten wrong along the way.
- July 6, 2026·4 min read
The rep who never forgets
Every sales org has a memory problem it refuses to name: the person who knew the deal leaves, and the next rep restarts from scratch. Institutional memory is the moat AI can actually build.
- June 29, 2026·4 min read
Pre-call anxiety is the feature
Tools that earn rep dependency don't do it through dopamine. They do it through the small, specific fear that you walked into the last call missing something — and the certainty that this one won't repeat it.
- June 29, 2026·4 min read
The half-life of a CRM field
Every field in your CRM is decaying. The question isn't whether your AI can write to the CRM — it's whether it knows which fields are already lies, and which ones it has no business touching.
- June 22, 2026·5 min read
Reading list · Week of June 22, 2026
Four pieces this week on the operating layer for B2B revenue from different angles: Sam Senior on why AI killed the discovery call, Jon Addison on how Okta's GTM restructure drove the turnaround, David Zhu on the Frankenstein stack as an architecture problem, and Keith Putnam-Delaney on where B2B demand is already moving.
- June 15, 2026·5 min read
Reading list · Week of June 15, 2026
Four pieces this week on what the human layer in revenue actually does when AI handles the substrate: ICONIQ's benchmark data on leaner AI-forward GTM orgs, Anthony Iannarino on CEOs who've stopped owning their sales discipline, Sangram Vajre on the predictable valleys between market-fit stages, and Scott Barker on why the AI labs are hiring GTM faster than they're hiring researchers.
- June 8, 2026·5 min read
Reading list · Week of June 8, 2026
Four pieces this week circling the same hard question: what does AI in GTM actually leave behind? Bertuzzi with SDR hybrid pod data, Vajre on alignment failures, Kazanjy on skill atrophy, Dorsey on the diagnostic layer no AI surface replaces.
- June 7, 2026·4 min read
The contradiction is the product
Most AI sales tools surface insight. The harder job: find the one fact in the deal that contradicts the rep's current plan, and surface it before the next conversation.
- June 1, 2026·5 min read
Reading list · Week of June 1, 2026
Four pieces on operating-model clarity: Sangram Vajre on why your revenue motion probably evolved rather than being designed, David Zhu on the Frankenstein Stack problem in GTM infrastructure, Becc Holland on the SDR KPIs that mislead at scale, and KD Dorsey on why 'improve your discovery' isn't coaching.
- May 25, 2026·4 min read
Reading list · Week of May 25, 2026
Four pieces from the production layer this week: Steve Lucas on AI going from influencing 20% to making 90% of decisions, Sam Senior on buyers who've already decided 70-80% before the first call, Auren Hoffman on the end of software moats and annual contracts, and Jon Addison on the CRO playbook that took Okta from $850M in losses to $760M in profit.
- May 21, 2026·4 min read
Reading list · Week of May 21, 2026
Four pieces this week on the context problem at the heart of GTM AI: Sam Senior on buyers who've already decided, Maja Voje on context engineering as the missing GTM discipline, ICONIQ's data on what AI-enabled teams actually look like, and Revenue Operations Alliance's 35-CRO survey on where AI ROI actually breaks down.
- May 20, 2026·4 min read
Operating layer, not assistant
"AI sales tool" spans chatbots wrapping a model to governed multi-tenant platforms. The category a product is actually in dictates which features compound and which leak trust. A test in four parts.
- May 20, 2026·6 min read
One system at a given time
The fastest way to lose an enterprise buyer is to quietly become a second system of record. Why revenue AI should write through to the CRM — and the checks that keep it honest.
- May 18, 2026·3 min read
Reading list · Week of May 18, 2026
Four pieces this week converging on the operating-layer shift: Mark Roberge on AI inside the rep workflow, Sam Jacobs on buyers educated before the call, Jacco van der Kooij on AI-Native revenue architecture, Kyle Norton on the 50/50 human-agent sales org.