The Growth Tuning method

How we ship AI, analytics, marketing, and web out of one team.

Every Growth Tuning engagement runs the same five-stage process. Same team, same rhythm, from discovery to weekly optimisation. No packaged retainers, no handoffs between disciplines, one revenue number as the north star.

Most agencies sell you a service. We ship a system. That difference shows up in the process: a shared discovery, a shared measurement layer, and a shared plan that spans AI, analytics, marketing, and web engineering — instead of four disconnected roadmaps.

01

Discovery

We start with a working session — usually 60 to 90 minutes with your founders, growth leader, or head of data. The goal is not to pitch: it's to diagnose. What blocks growth today? Where's the leak? Which decisions are being made without evidence? What's already been tried?

By the end we agree on the one number the engagement moves — pipeline generated, MRR added, CAC / LTV ratio, whatever's actually load-bearing for the business — and the constraints we have to respect.

Week 1
02

Measurement baseline

Before we recommend any strategy or ship any AI, we wire up honest measurement. That means GA4 correctly configured, server-side tagging where consent frameworks demand it, event schema documented, and a warehouse pipeline (BigQuery or Snowflake) capturing the granular data your dashboards ride on top of.

If attribution is broken — and it usually is — we install a defensible multi-touch model and, at spend levels that justify it, a marketing mix model on top. This is the layer everything else stands on.

  • GA4 audit + fix, plus server-side tagging where needed
  • Warehouse-first data pipeline (BigQuery / Snowflake / Postgres)
  • Attribution model (rules-based, MTA, or MMM depending on spend and data)
  • Baseline dashboards for CAC, LTV, funnel velocity, channel-level ROI
Weeks 2 – 4
03

Strategy

With real measurement in place, the strategy work is short and specific. We produce a single sequenced plan across all four pillars, not four separate roadmaps. AI investments where they compound, analytics where the org still guesses, marketing programmes tied to the funnel gaps the data revealed, web engineering to remove the frictions costing revenue right now.

The plan reads like an engineering roadmap, not a slide deck: milestones, dependencies, owners, and the specific measurement each initiative moves.

Weeks 3 – 5 (overlaps with baseline)
04

Build

One team ships across all four pillars in parallel. The AI systems (agentic workflows, predictive models, LLM-driven surfaces), the analytics infrastructure (warehouse jobs, dashboards, alerts), the marketing programmes (SEO, content, paid, lifecycle email), and the web engineering (custom builds, e-commerce, landing pages) — all running concurrently, not in a waterfall.

Weekly checkpoints keep the workstreams synchronised. When one pillar surfaces a signal — say the paid channel shows non-obvious ROAS in a segment — the analytics and web teams respond in the same sprint. That's the point of one team.

  • Concurrent workstreams across AI, analytics, marketing, and web
  • Weekly showcase with the leadership team — no monthly-status theatre
  • Shared repository of tickets, dashboards, and decision docs
Weeks 5 – 12+
05

Run and learn

The engagement doesn't end at "launch". Once systems are live, the measurement layer becomes a feedback loop: AI signals surface anomalies, wins, and pruning candidates in near-real time. We double down where the data proves the case, kill what isn't working fast, and refresh the strategic plan on a quarterly rhythm.

Most engagements settle into a steady-state retainer at this stage — same team, same tooling, same north-star number — with the ratio of build-to-run reversing over the first six months.

Ongoing

Why this beats hiring five separate agencies

One roadmap, not five

When SEO, ads, analytics, AI, and web live in one plan, they compound. When they live in five, they collide — same audience, different messages; same event, five different definitions.

Same measurement, all disciplines

Every pillar reads from the same warehouse and the same attribution model. No agency arguing about which report is "correct" — there's one report, and it's yours.

AI as a horizontal, not a silo

The AI pillar isn't a separate offering that lives in isolation. It powers the other three — content briefing, campaign optimisation, dashboard surfacing, engineering acceleration — and gets sharper the longer we work with your data.

Discovery-first, no packaged tiers

Every engagement is scoped after the discovery. No bronze / silver / gold. If a discovery shows the fix is a two-week job, that's what we quote. If it's a nine-month rebuild, we say that too.

Weekly cadence, not monthly reports

You see the work as it happens. Slack channel, shared tickets, shared dashboards. No polished monthly deck hiding what's actually going on — the deck IS the dashboard.

One team, senior to the work

You talk to the people doing the work, not an account manager between you and them. Senior AI, analytics, marketing, and engineering practitioners on every engagement.

Start with a discovery call.

Sixty minutes. No pitch deck. We map your growth stack, name the leaks, and tell you what a real engagement would look like — including a plain-English scope and range.

Book a discovery call