Jud Mackrill 🏴‍☠️

CEO @milemarkerco / GP @mammoth_vc / Partner @kimmackrill

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Really enjoyed getting to see my people in Denver this week.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

@The_MikeKelly Very true.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Wealth management has a data pipeline problem most fintech builders don't see.

Every custodian, every CRM, every rebalancing tool holds a partial view of the client. None of them were designed to share. The advisor sits downstream of 8 systems and reconciles manually.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

The firms that win the next decade of wealth management are building data pipelines into their operating model, not bolting them on after.

The plumbing is the product.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

The advisors doing client analytics aren't doing it because they have a data warehouse. They're doing it because someone at their firm spent 200 hours wiring CSV exports into spreadsheets.

That's not infrastructure. That's technical debt with a human face.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Client analytics is not a BI problem. It is a data engineering problem.

Solve the pipeline and the insights start flowing. Skip it and you are rebuilding the same dashboard for the third time wondering why nothing ever feels right.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Wealth management firms have a data warehousing problem that looks like a reporting problem.

Every advisor wants better client views. The ask goes to IT. IT builds another dashboard.

Nobody is upstream building the model that makes every future dashboard cheap.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

The firms that crack this invest once in a clean semantic layer. Household as the unit. Accounts, positions, activity all hanging off it properly.

Snowflake + dbt makes this buildable in weeks. The hard part is getting someone to own it like infrastructure, not a project.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

people praising the manhattan project as a gold standard of innovation or achievement need to broaden their perspective.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Every firm that has solved this with a real data layer, not just dashboards, but actual connected infrastructure, has stopped losing advisor hours to data wrangling. That is the operational alpha nobody talks about.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Wealth managers spend more time managing data than managing money. Not an exaggeration.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

The average RIA has 8+ systems touching client data. Almost none connect natively. So data moves via CSV exports, manual entry, and reconciliation spreadsheets. Advisors become the integration layer by default.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

@RyanGigous We should show you how this works for us

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Wealth management has a client identity problem that most data engineers do not expect.

Your CRM has a contact ID. Your portfolio system has an account number. Your custodian has a household ID. None of them match.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Until you resolve identity across your systems, every client analytics dashboard is a partial view.

You are not missing data. You are missing the map between the data.

The firms that fix this stop reporting from data and start running on it.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

The "client" in one system is an account in another and a household in a third.

This is not a query problem. It is a people-and-process problem dressed up as a schema problem.

Most firms solve it with a spreadsheet someone maintains manually. That spreadsheet is the actual source of truth.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

@fin465 Send

[Jud Mackrill 🏴‍☠️ @JudMackrill]

The firms that learn this the hard way usually learn it the same way.

An auditor asks what a client balance was on a specific date.

The data team has to reconstruct it from logs.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Financial data is not as immutable as data engineers expect.

Prices get restated. Account values are corrected after reconciliation. Fees are recalculated retroactively.

This is not exceptional in wealth management. It is routine.

[Jud Mackrill 🏴‍☠️ @JudMackrill]

Most data pipelines model "current state." Wealth management requires point-in-time.

Those are architecturally different requirements. Snowflake Time Travel and Delta Lake exist for this reason. But they have to be designed in from the start, not retrofitted.