How We Work

Assessment first. Then architecture. Then scale.

We don't start with a solution and work backwards to justify it. Every engagement begins with a structured assessment that produces three deliverables. What gets built after that depends on what we find — not what we assumed going in. We don't deliver decks and walk away. We build systems that run.

Engagement

What it looks like in practice.

A $100M+ B2B SaaS company brought on a new Head of Marketing with a mandate to rebuild the GTM engine. The existing model ran one program, targeted one buyer type, and concentrated spend on a single conversion point. The 2026 plan required dozens of programs across multiple segments and trust levels.

Jetpacks They Said was brought in to assess the gap and design the AI infrastructure to close it. Phase One is a six-week diagnostic across five dimensions. The target architecture separates what should be locked by humans, what AI handles at volume, and what stays under human judgment.

This engagement is in progress. When the numbers exist, they'll be here.

Read the full engagement overview (PDF) →

Engagement Model

Phase One

Assess

6 weeks

Three deliverables: a MarTech landscape with AI integration map, a prioritized implementation plan, and an AI architecture blueprint. You get a clear picture of where you stand, what's possible, and a phased build plan with sequencing and dependencies.

MarTech LandscapeImplementation PlanAI Architecture Blueprint
Phase Two

Build

Scoped in Phase One

Implementation. What gets built depends entirely on what Phase One surfaces. We don't prescribe solutions before we understand the problem. Scope, timeline, and cost are defined at the end of Phase One — not before.

Determined by Phase One findings
Phase Three

Scale

Ongoing

Optimization and expansion. New programs, new segments, new motions — running on the infrastructure built in Phase Two. This is where the investment compounds.

Determined by Phase Two outcomes

The Build Discipline

Before we build anything, a requirement has to survive five steps.

01

Challenge the requirement

Every requirement is wrong by some degree. Interrogate it before accepting it.

02

Remove before you add

If you're not adding 10% back after deletion, you didn't delete enough.

03

Simplify what remains

Only after steps one and two. Optimizing something that shouldn't exist is waste.

04

Speed up the cycle

Accelerate only what has earned its place in the process.

05

Automate last

Automation at the wrong stage produces the wrong output, faster.

Ground rule: Every requirement has a name attached. Not a department. A person.

What this isn't

A strategy engagement. Your strategy already exists. We build the systems that make it run.

An audit that ends in a PDF. Every deliverable is designed to be acted on immediately — not shelved.

A vendor pitch dressed up as consulting. We don't sell software. We make the stack you already have work the way your team actually works.

Ready to talk?

15 minutes. No pitch deck. We'll talk about where AI fits in your marketing infrastructure and whether it's worth going deeper.

Schedule a Discovery Call
Schedule a Discovery Call