💃 Everyone Owns the Account, So Nobody Does 🕺

Good morning, Salesforce Nerds!

You ask a simple question in standup: who owns the Account object?

Sales says marketing. Marketing says the data team. The data team says it's a shared resource. RevOps says it depends on the field.

Four answers. Zero owners.

A duplicate Acme Inc. record sitting in production with three different annual revenue figures. 😅

This is the accountability vacuum, and it doesn't come from a bad decision. It comes from no decision.

Most orgs never choose a data ownership model. They drift into one. Let's name the trade you're actually making. 👇️

TABLE OF CONTENTS

💃 Everyone Owns the Account, So Nobody Does 🕺

• The Decision You Already Made

• One Throat to Choke

• Many Hands, Many Masters

• Draw the Line on Paper

• Own the Model You Have

THE DEFAULT NOBODY CHOSE

THE DECISION YOU ALREADY MADE

Data ownership answers one question: who is accountable for a piece of data being correct?

Centralized ownership puts one team in charge. They define the schema, set the quality bar, and approve changes. Think of a central data team guarding the Account object like a bouncer. 🛡️

Distributed ownership pushes accountability out to the teams closest to the data. Sales owns leads and opportunities. Finance owns billing entities. Service owns case data. Each domain runs its own patch.

Neither is right or wrong. Both carry a bill.

The catch is that most orgs run a third model by accident: diffuse ownership, where everyone can write and nobody is accountable. That's not distributed ownership. That's an unowned free-for-all wearing a lanyard. 🪪

THE BOUNDED CONTEXT

ONE THROAT TO CHOKE

 Centralized ownership buys you a single source of truth. One golden record. One set of validation rules. One team that knows exactly why the Annual_Revenue__c field works the way it does.

When an exec asks "what's our customer count," you get one number. Stewardship is unambiguous because the steward has a name and a Slack handle. 📊

But centralization has a failure mode, and it's brutal at scale.

A well-known anti-pattern is making Salesforce itself the enterprise MDM hub. Every downstream system, ERP, finance, reporting, checks against the Org for the golden record. It feels clean on the whiteboard.

Then reality lands. Dedup logic times out. Row-lock errors become routine. Read-only maintenance windows freeze the whole company because every system depends on one Org being up. 🔒

The central team turns into a ticket queue. Every domain that needs a schema change waits behind everyone else. You traded chaos for a bottleneck, and the bottleneck has a SLA you can't hit.

DOMAIN-DRIVEN, FRICTION-INCLUDED

MANY HANDS, MANY MASTERS

Distributed ownership flips the model. Zhamak Dehghani formalized this as the data mesh: domain teams own their data as a product, with federated governance setting global rules everyone follows. 🕸️

The pitch is real. Accountability lives where the knowledge lives. Marketing owns campaign data because marketing actually understands it. Quality goes up because the people closest to the data are on the hook for it.

Agility goes up too. Domains ship changes without queuing behind a central gatekeeper.

Now the bill. Distributed ownership is an integration tax, and you pay it in sync logic.

The moment two domains own overlapping data, you need reconciliation. Whose Email__c wins when Sales and Service disagree? You're now writing survivorship rules, conflict resolution, and last-write-wins guards across Platform Events or CDC streams.

Each domain boundary is a new integration to build, monitor, and debug. Federated governance isn't free either. Someone has to define and enforce those global standards, or your mesh quietly degrades into silos that don't talk. 🧩

CONTRACTS BEAT VIBES

DRAW THE LINE ON PAPER

Both models live or die on one thing: explicit boundaries. The fix isn't picking a side. It's making ownership legible.

Start by designating a system of record per data domain. Not per object. Per domain. Account billing data masters in ERP. Engagement data masters in Salesforce. Say it out loud and write it down. ✍️

Then formalize the handoffs with data contracts. A data contract is a written agreement between a producer and a consumer: schema, field semantics, quality guarantees, and update cadence. It turns an implicit assumption into an enforceable interface.

// The contract made executable: validate inbound
// domain data against agreed semantics before it
// touches your system of record.
public interface AccountDataContract {
    Boolean isValid(InboundAccount payload);
    String  systemOfRecord();      // who masters this?
    Datetime lastVerified();       // freshness guarantee
}

One nuance worth flagging. Salesforce Data Cloud's Unified Profile is not a golden record. It links fragmented data together for a dynamic customer view rather than replacing your mastered data. Federation, not consolidation. Know which one you're building. 🎯

NO REWRITE REQUIRED

OWN THE MODEL YOU HAVE

You probably won't get to redesign your ownership model from scratch. You inherited an org, a history, and a few hundred automations that assume the current arrangement.

That's fine. Resilience doesn't require a rewrite. It requires honesty about what you have.

Make the implicit explicit. Name the system of record for each domain. Write the contracts down. Put a human name next to every critical object. 🤝

The strongest orgs aren't the ones that picked the perfect model. They're the ones that know exactly which model they're running and where its bill comes due.

SOUL FOOD

Today’s Principle

and now....Salesforce Memes