Research · Analyse · Learn · Iterate

AI consultancy
& automation agency

Output Loop designs, deploys and maintains AI agents, automations and connected workflows built around how your business already works. Turn repetitive work into reliable output so your business can scale without limits.

What we automate
Typing invoices into Xero· Chasing quotes and payments· Re-keying between systems· Missed calls, lost work· Inbox triage every morning· Reports built by hand·

01 — Who we build for

Every team has a bottleneck.

We start where the hours actually go, automate the loop that creates them, and keep the human where judgement lives. The method doesn’t change from team to team — only the bottleneck does.

Sales

Pipeline that fills itself — prospect lists built and refreshed, outreach sequenced and sent, meetings booked while you sell.

Marketing

Campaigns that ship on schedule — research, drafts and distribution running as one loop, not a backlog.

Operations

The recurring work mapped, automated and handed back — reporting on cadence, documents filed as they land.

HR

Candidates screened to shortlists, onboarding on autopilot, policy questions answered from your own handbook.

Support

Every inbound message triaged and drafted against your knowledge base — humans only at the boundary.

Finance

Invoices matched line by line, approvals without chasing, month-end packs ready on day one.

02 — Services

Built around your operation.

Twelve of the systems people call us for — most engagements combine several, and the list keeps growing. Each is scoped to a fixed price, delivered in weeks, and run — not just handed over. Click any card for what’s inside.

03 — Why Output Loop

Start with where the hours go.

Most businesses don’t need convincing that AI matters — they need someone to show them where it fits their operation. So we start with a diagnosis, not a pitch: a written map of where your hours actually go, which loops are worth automating, and which honestly aren’t.

Then we build the ones that are — governed, observable, and owned by you. You see what each system did, why it did it, and where a human signs off.

Start with the map

04 — The proof

Built, shipped, running.

Explore the research and systems behind our own operation, with the evidence and methods attached. Client engagements will be published with their owners’ consent.

The newsroom fleet — a daily publication run by agents

A daily publication, run by agents

An ongoing publication of AI analysis and newsletters, researched, written, checked and published with agents. Each edition connects readers to its sources, with human review before publication.

Read the latest articles and newsletters →
Mona Lisa assembled from small cubes — local businesses, a bigger picture

Local businesses. A bigger picture.

8,976 businesses. 12 locations. Websites and six social channels. Explore how local businesses appear online, the workflows behind the research, and practical ways to make the next customer’s journey clearer.

Explore the study →
The benchmark discipline — we test AI before we trust it

We test AI before we trust it

Before a model runs our work, it runs our benchmarks. Recent rounds caught models silently failing tasks, skipping required steps, and inventing timestamps — failures a sales demo never shows you.

05 — Newsletter

One issue a week.

What we built, what broke, and what we learned — straight from the systems we run. No noise, no drip funnels.

The first issue is being written now. Subscribe and it lands in your inbox the moment it’s published.

Read the newsletter

06 — About us

One human. A fleet of systems.

Output Loop is run by one person and operated by many systems. Michael scopes the work, makes the judgement calls, and stays accountable for every engagement. The fleet — supervised agents on our own infrastructure — does the running: monitoring, reporting, publishing, follow-up.

It’s the same model we install for you: keep the humans where judgement lives, and let the systems carry the rest.

Michael Mcgowan, founder of Output Loop

Michael Mcgowan

Founder

Across a commercial career in different roles, Michael has generated $20M+ for the organisations he worked with. Output Loop is the operating system he wished those organisations had — built first for his own company, now for yours.

07 — How we’re different

Evidence, not claims.

200+Custom skills built
45+Workflows in operation
100BN+Tokens processed
200+Agent-hours, every week

Diagnosis before builds

The first deliverable is a written map of where your hours actually go — useful even if you never buy a build. We’ll tell you what not to automate, too.

Numbers you can check

Every figure we publish is dated, counted on our own systems, and re-verified. No cumulative “£X saved for clients” you have to take on faith — this market has enough of those.

Governed and observable

Every system ships with logs, measures, and review points. You can see what it did, why it did it, where your data went, and where a human signs off.

Provider-agnostic

No sector niche, no vendor allegiance. Claude or Codex, Google or Grok, n8n or custom agents — we recommend what fits your operation, because we run them all ourselves.

08 — The loop

Research. Analyse. Learn. Iterate.

The operating loop behind every Output Loop system — including our own. It’s where the name comes from: every cycle of research, analysis, and iteration compounds into output you can measure.

09 — The stack

Six months. Eighty-plus tools. Daily, hands-on.

Not a logo wall — the tools our fleet and our founder actually work in, daily. We recommend from experience, not commission.

Language models

OpenAI · Claude · Gemini · Grok · DeepSeek · Qwen · Kimi · MiniMax · GLM — benchmarked against each other on our own tasks before any goes near production.

Image, video & audio

Midjourney · Runway · Kling · Leonardo · Nano Banana · HeyGen · ElevenLabs · Suno · Descript.

Agents & coding assistants

OpenClaw · Claude Code · Codex · Cursor · Manus · Hermes · Buzz — including the 38-agent fleet that runs this company.

Automation platforms

n8n · Make · Zapier — the connective tissue between the tools you already pay for.

Research & scraping

Firecrawl · Exa · Tavily · Perplexity · Brave Search — the engine room behind our 4,703-business research run.

CRM & sales systems

GoHighLevel · HubSpot · Salesforce · Pipedrive · EspoCRM — with formal training in HubSpot and Salesforce from Michael’s commercial career.

Delivery & infrastructure

Vercel · Cloudflare · Hetzner · GitHub · GoDaddy · Resend · AgentMail — self-hosted where it matters; your data stays where you can see it.

Workspace & comms

Notion · Slack · ClickUp · Obsidian · Canva · Adobe — because automation has to land in the tools your team already lives in.

10 — How we work

Map. Scope. Build. Run.

  1. STEP 1

    Map

    A free 30-minute systems-mapping call. We find where the hours go, where the leads leak, and where AI actually moves the needle — before you spend anything.

  2. STEP 2

    Scope

    A written diagnosis and a fixed price with a clear deliverable list. No hourly meters, no scope fog — you know the number and the outcome before we start.

  3. STEP 3

    Build

    Delivery in weeks, not quarters — with evidence at every step: working systems, documented workflows, and your team trained inside them.

  4. STEP 4

    Run

    We operate, measure, and iterate after launch. Research, analyse, learn, iterate — the loop that keeps the system improving instead of quietly rotting.

Start with the map.

Thirty minutes, free, no deck required — bring the process that hurts.

Fixed scope, fixed price, agreed before we start — no hourly meters.