Skip to content

Marketing that runs on set workflows, not in bursts.

I build AI-powered marketing systems: content workflows, enquiry routing, CRM automation, and reporting that shows what worked.

AI only pays off in marketing once it's built into fixed workflows: clean data, reliable handovers, and measurable results. As an AI marketing agency, I use automation (n8n/Make), custom apps, and AI features to do that: producing and distributing content, handling enquiries, and reporting.

Sound familiar?

Where marketing stalls today

Content comes in bursts

Three posts one week, then nothing for a month — depending on who happens to have time.

Enquiries peter out

They come in, but nobody keeps track of who is dealing with them, or when.

Nobody knows what worked

Without clean measurement, every decision about the next budget stays a gut feeling.

What this looks like in practice

From Pinterest to reporting

Pinterest

Research search terms, group them into topics, and build pin variants from them

Pins from templates

Image, copy, variants, and the finished schedule

Newsletters

Topics from your own archive, draft, review, send

From the form into the CRM

Enrich enquiries from the form, move them into the CRM, assign them with a deadline, and follow up

Short versions for social media

Repurpose long articles as short versions for social media

Metrics dashboard

Metrics from every channel in one dashboard, with an alert when something is out of line

List articles

Structure, review, and publishing, partly automated

Testing ad visuals

Generate ad visuals in several variants, test them, and evaluate which ones work

What you get out of it

A system for content, enquiries and measurement

More content — with checks built into the workflow

Automated research and production workflows: topics, keywords, briefs, variants, checks — fewer bottlenecks.

Enquiries handled systematically, not by chance

Forms, tracking, data enrichment, CRM, follow-ups, and fixed response times are all connected. Enquiries are handled as a matter of routine — not forgotten.

Measurement and reporting

Metrics collected automatically (UTM parameters, events, sources) and dashboards with alerts — so your decisions are based on data.

Maintainable workflows

Error handling, logs, automatic retries, and clear ownership are included too.

From my own practice

Pinterest without manual work

For my own Pinterest accounts, a tool checks which article does not have a pin yet, creates the image and the text, picks the board, and publishes on a staggered schedule.

How we work together

Five steps from first call to handover

  1. 1

    Intro call

    30 minutes, free, booked through my calendar. You tell me which process is costing you time.

  2. 2

    Process review

    Together we work out which process is worth tackling first and what it will take.

  3. 3

    Prototype

    Within days rather than months, there's a first version for you to try out.

  4. 4

    Pilot

    The solution runs in day-to-day operations with your team. Any rough edges get smoothed out.

  5. 5

    Handover and support

    Documented, monitored, and developed further if you want. You're not left to deal with it on your own.

Technologies used

n8n, Make, Claude Code (custom apps), TypeScript/Next.js, webhooks, REST APIs, Postgres/SQL, Vercel, tracking/UTM standards

An example from my work

Pinspector: Pinterest search terms with search volume
Own products · used in my own business

Pinterest tools — Pinterest MATCHER and Pinspector

Every article gets its pin — created automatically, on the right board, on topics people actually search for.

See the full case
Frequently asked questions

What you'll want to know first

Do you write the content or build the workflows?

I build the workflows that content is created in: topics, keywords, briefs, variants, checks. That is exactly what runs on my own Pinterest accounts — from the article without a pin to the published pin.

How can I tell whether it's paying off?

From metrics that are collected automatically (UTM parameters, events, sources) and dashboards that flag anything out of line.

Roughly how much does it cost?

Prototypes start at €3,000. The intro call beforehand is free.

How long does it take?

You get a first working prototype in days, not months.

What happens in the intro call?

Half an hour, free, booked through my calendar. You tell me which process is costing you time today; together we look at which one is worth tackling first — and whether I'm the right person for it.

Attila Arndt

Attila Arndt · Triple A Digital, Cologne

Where is your marketing stalling right now?

Pick a time that suits you. In the intro call, we'll work out which process is worth tackling first — and whether I'm the right person for it.

Free intro call (opens in a new tab)

Or email me directly: hello@tripleadigital.de · I reply within 48 hours.