Most advisory firms are tinkering with AI right now. And tinkering feels productive.

You're testing ChatGPT for meeting summaries. Your planning tool just shipped an AI feature and you're poking at it. Someone on your ops team is using Claude to draft client emails. Your CRM vendor demoed their new AI assistant last month and it looked pretty good.

All of this is fine. But let's be honest about what it is: tinkering.

Tinkering is experimenting in isolation. It's using AI in one tool at a time, for one task at a time, without a strategy connecting any of it. It feels like progress because each individual experiment produces a result. But zoom out, and your firm hasn't actually changed how it operates.

Transformation is something different entirely.

The gap between tinkering and transformation

Here's the distinction: tinkering is asking AI to do a task inside one system. Transformation is AI working across your entire firm — connecting data, surfacing insights, and driving decisions that no single tool could produce on its own.

When an advisor uses their planning tool's AI to generate a retirement projection, that's tinkering. When AI automatically prepares a comprehensive client review by pulling from the CRM, custodian, portfolio management system, planning tool, and compliance records — correlating all of it into a unified picture — that's transformation.

The gap between those two things isn't about better AI. It's about better data.

AI is only as good as the data it has access to. It is not smart enough to know that Joe updated a client's address in the CRM but it hasn't synced to the custodian yet. It can't reconcile the fact that your planning tool calculates AUM differently than your portfolio management system. It doesn't know that the compliance flag from last Tuesday is related to the CRM note from last Wednesday.

If you're running AI on fragmented, disconnected data, you're getting fragmented, disconnected results. No matter how good the model is.

Why tinkering feels like enough (but isn't)

Tinkering is seductive because it produces quick wins. An AI that summarizes a 60-minute client meeting in 30 seconds? That's valuable. An AI that drafts a follow-up email in your voice? Useful. An AI that flags portfolio drift against a model? Helpful.

But each of those wins exists in its own silo. The meeting summary AI doesn't update the CRM. The email draft AI doesn't know about the portfolio drift. The drift alert doesn't reference the client's financial plan or the conversation you had last week.

You're still the integration layer. You're still the one connecting the dots between systems. AI made each individual task faster, but the workflow — the actual way your firm operates — hasn't changed.

More than half of advisors in Orion's WealthTech survey say AI is the number one growth catalyst for 2026. They're right. But growth from AI requires more than faster individual tasks. It requires connected intelligence.

What transformation actually requires

Three things separate firms that are tinkering from firms that are transforming.

1. Unified data

You can't run AI across your firm if your data lives in seven different systems that don't share with each other. The first step to transformation is getting all of your data — CRM, custodian feeds, portfolio management, financial planning, compliance, billing, document management — into a single, unified layer.

Not exports. Not CSV dumps. Live, connected data that reflects the current state of every system. When something changes in your CRM, the unified layer reflects it immediately. When a custodian feed updates, every downstream process sees it.

This is the foundation. Without it, AI will always be limited to whatever data one individual tool can see.

2. Willingness to re-evaluate partners

This is the uncomfortable one.

If your technology partners are not sharing your data well — if they're building a moat around it, if they don't integrate cleanly, if they don't want to help you connect your ecosystem — you need to re-evaluate that partnership.

The price of progress is the pain of change. And yes, switching technology is painful. It's disruptive. It takes time. Nobody likes it.

But staying with a partner that holds your data hostage is more painful in the long run. Because it means you can never unify your data. You can never run AI across the full picture. You're permanently stuck tinkering in silos because one or two vendors won't play nice.

Look at what's happening in the broader AI world. People switch between Claude, ChatGPT, and Gemini without a second thought. When a better model appears, users move. There's no loyalty to a model that stops serving you well.

We need that same mindset in WealthTech. If a technology partner isn't investing in data sharing, isn't making it easier for you to own and control your data, isn't helping you build a connected ecosystem — they're holding you back from transformation. And there are partners out there who will fight for your data, who will advocate for your ability to build something unified.

3. A strategy (not just experiments)

Tinkering is reactive. Someone sees a demo, tries a tool, experiments with a prompt. There's no plan connecting any of it.

Transformation requires sitting down and asking: What do we want AI to accomplish for our firm? Not "get more efficient" — that's too vague. What specific workflows do we want to automate? What decisions do we want AI to support? What client experiences do we want to improve?

And then working backward: what data do we need connected to make that possible? What systems need to talk to each other? What partners need to step up?

Chip Kiss — who was on our previous Next Mile episode — said it well: the first thing you have to do is get an AI policy and an AI strategy. Don't be reactive. Don't just add AI tools because your vendors are shipping them. Decide what you want to accomplish and build toward it.

What the path looks like

Here's the practical sequence from tinkering to transformation:

Step 1: Audit your data. Where does client data live today? How many systems? How well do they share? Where are the gaps, the duplicates, the conflicts?

Step 2: Unify. Get your core data — CRM, custodian, portfolio management — into a single layer. Don't build it yourself unless you have a dedicated engineering team. Use a platform that already has the integrations built.

Step 3: Expand. Add the rest of your stack to the unified layer. Planning, compliance, billing, documents. Each connection adds more context. The value compounds.

Step 4: Activate. With unified data, start deploying AI for cross-system use cases. Client review prep that pulls from everywhere. Exception flagging that correlates data across systems. Operational dashboards that reflect the full picture.

Step 5: Iterate. AI capabilities are improving rapidly. The firm with unified data can adopt new capabilities immediately. The firm with fragmented data has to wait for each individual vendor to ship their version.

At Milemarker, this is exactly what we help firms do. Our Data Engine connects 130+ WealthTech platforms into a unified data layer powered by Snowflake. Navigator layers AI intelligence on top of that unified foundation. The transformation unlock isn't better AI — it's better data infrastructure that makes any AI more powerful.

The window is now

If you're tinkering today, that's okay. Everyone starts there. But don't mistake tinkering for transformation.

The firms that unify their data in 2026 will have a compounding advantage. Every month of clean, connected data makes AI more valuable. Every new model that launches works better on unified data. Every workflow you automate on a connected foundation saves more time than the same workflow running in a silo.

The price of progress is the pain of change. But the cost of standing still is higher.

This article is based on the Next Mile podcast episode "Relentless, Relevant, and Ready: Orion Ascent Recap with Jessica Perez." Listen to the full conversation for more insights on AI adoption in wealth management.

Ready to move from tinkering to transformation? Book a strategy call and we'll walk through what unified data looks like for your firm.