One day · in person · for marketing and data teams

Talking to Data Engineering

Most Martech projects don't stall on disagreement. They stall on the false kind of agreement, where marketing and data engineering leave a meeting certain they decided the same thing, and find out six weeks later that they didn't.

Tell me about your teams From €6,500 per day. Up to 20 people.

A request I've seen in most Slack channels I've been let into

“Can you sync1 the lapsed-customer2 audience3 to Braze in real time4? Just needs the churn model5 score on each profile6.”

Nothing about it is unreasonable. It also contains six words that mean something different to the person who has to build it. Nobody finds out until the segment refreshes daily instead of hourly, and the field the campaign was built on isn't there.

  1. 1sync

    You mean

    The data shows up in the tool.

    They hear

    A job with failure modes, retries, and last night's failed records.

  2. 2customer

    You mean

    A person.

    They hear

    An entity at a grain. Person, account, household or device, and which identifiers stand for it.

  3. 3audience

    You mean

    The same thing as a segment.

    They hear

    The exported copy pushed to the destination, not the logic behind it.

  4. 4real time

    You mean

    Soon enough that it does not feel broken.

    They hear

    Sub-second streaming, an always-on pipeline, and a cost line that never stops.

  5. 5model

    You mean

    A prediction, like propensity to churn.

    They hear

    The data model. Tables and how they relate.

  6. 6profile

    You mean

    Everything known about someone, in one view.

    They hear

    A join, computed at a moment and already slightly stale.

01 · The day

Your people, your briefs, one room.

One day, in your building, with your marketers and your data engineers in the same room, working on your own briefs.

Nobody writes SQL.

What people write is the request they were about to send on Monday. Then they rewrite it with the person who would have to build it sitting across the table, arguing about it.

What your teams leave with

A shared vocabulary
The words that caused the trouble, defined out loud by both sides.
Requests that can be built
Without the data team guessing what you meant.
Decisions made on purpose
On speed, identity and sign-off, rather than by whoever picked up the ticket.

02 · Six modules, one hour each

Short teaching, then your own work.

Each module is a short piece of teaching followed by an exercise on your own material rather than a case study. Everyone brings one real brief or data request with them.

123Lunch456Close

Say what you mean

1

Same word, different meaning

Take the brief you brought, find the words that mean two things, and have both sides define them out loud.

2

A request someone can act on

Rewrite a one-line request into the six decisions it was missing, then hand it to someone to break.

Know what it costs

3

What real time actually costs

Place your own use cases on the speeds between nightly batch and streaming, with a reason behind every deadline.

4

Same customer, different records

Match versus merge. Set a threshold and see who it merges into one person and who it splits into four.

Decide who decides

5

Who's allowed to say yes

Map the consent and approval path for one live project, and find the step nobody owns.

6

A goal a system can act on

Write a four-part goal, including the guardrail everyone leaves out. This is where agentic AI comes in.

Who should be in the room

Marketing operations, CRM and lifecycle, Martech leads and campaign owners, and the data engineers they depend on.

Send both sides. Half the day is about what the other side hears.

What's included

  • A call with your internal sponsor beforehand
  • A review of every brief sent in
  • A printed workbook for everyone
  • The day itself
  • A follow-up call 30 days later

Format

In person at your offices, 09:00 to 16:30. No prerequisites, and no laptops beyond one per pair.

Vendor-neutral. No product is pitched, and none is the answer.

03 · Questions

What people ask before booking.

How much does Talking to Data Engineering cost?
From €6,500 per day, excluding VAT, for up to 20 people. The intro call settles the exact figure for your team.
Does anyone need to write SQL?
Nobody writes SQL. People write the request they were about to send on Monday, then rewrite it with the person who would have to build it.
Who should attend?
Marketing operations, CRM and lifecycle, Martech leads and campaign owners, and the data engineers they depend on. Send both sides: half the day is about what the other side hears.
Is it remote?
It runs in person at your offices, 09:00 to 16:30.
Is it tied to a vendor?
It is vendor-neutral. No product is pitched, and none is the answer.
What happens before and after the day?
A call with your internal sponsor beforehand and a review of every brief your people send in. On the day everyone gets a printed workbook, and there is a follow-up call 30 days later.

From €6,500 per day, excl. VAT

Before we fix a date, we have one 30-minute call, so the modules can be pointed at what your teams are actually building. Already done a decision session? Book the day within 60 days of it and the €1,500 comes off.

Download the one-pager (PDF, 2 pages)

For whoever signs off on the budget.

Next step

Four questions. You get the intro-call link straight back, and I'll have read your answers before we speak.

Rather go straight to my calendar? Book the intro call directly.