Blog

Blog

Blog

Insights & Expertise

Human+AI, data, CRM and intelligent systems: what we learn building and operating, and the studies we publish.

The self-improving agent stack: what compounds and what you still own
Martech & Industry Trends

The self-improving agent stack: what compounds and what you still own

Skills, harnesses and models can now rewrite themselves. The question is not whether to let them. It is which part of the stack you hand over, and which part you keep.

5 Oct 2026 Read more →
Gemini 4 Argon: where it leads, where it trails, and what to test it on
AI & Business Automation

Gemini 4 Argon: where it leads, where it trails, and what to test it on

Google is back at the frontier. Argon leads on knowledge work, finance and legal agents, trails on terminal coding and computer use, and most teams cannot use it yet.

5 Oct 2026 Read more →
How to BIMI Email Setup: The 5-Step Checklist
ESPs & Marketing Automation

How to BIMI Email Setup: The 5-Step Checklist

Most BIMI implementations fail at the DNS record, not because the syntax is wrong, but because the sender skipped the foundational layers the record depends on. BIMI email setup is a stack: DMARC at enforcement, a validated SVG, an optional but increasingly required Verified Mark Certificate, and then the DNS TXT record that ties it […]

1 Oct 2026 Read more →
The RFM Segmentation Email Marketing Playbook That Actually Works
ESPs & Marketing Automation

The RFM Segmentation Email Marketing Playbook That Actually Works

Most email programs treat RFM segmentation as a reporting exercise. Marketers slice their database, label the top 20% “champions,” and then send every segment the same volume cadence with slightly different subject lines. Inbox rates drop. Revenue attribution gets murky. The RFM model takes the blame, when the real failure is operational: the segmentation never […]

30 Sep 2026 Read more →
Three AI price tiers and the CRM work that belongs in each
AI & Business Automation

Three AI price tiers and the CRM work that belongs in each

Opus 5.5 and GPT-6 Sol and Luna cut prices on 22 September. The larger saving is putting each CRM job on the right tier: $300 or $7.50 for the same labelling job.

28 Sep 2026 Read more →
Email Complaint Rate Threshold: Architecture Patterns for Scale
ESPs & Marketing Automation

Email Complaint Rate Threshold: Architecture Patterns for Scale

This reference is for email deliverability engineers, CRM managers, and marketing operations leads who need precise definitions – not approximations – when designing sending infrastructure that stays healthy at volume. If you are managing more than a few million emails a month, these terms directly affect whether your messages reach the inbox or disappear silently. […]

24 Sep 2026 Read more →