Wind-down timeline: when to assess data relative to layoffs and cancellations

Why data assessment belongs at the start of a wind-down#

Data assessment belongs at the start of a wind-down because everything it depends on disappears early. Records are lost in a predictable order: first the people who know where they live, then admin access, then the systems themselves when billing stops, and finally the entity with authority to decide what happens to them.

Most wind-down plans are organized around legal and financial milestones: the board vote, employee notices, customer communications, final payroll, tax filings and dissolution. Data rarely gets a milestone of its own, so it ends up handled by whoever cancels the last subscription. Putting data steps into the same plan, with owners and dependencies, fixes that.

The wind-down timeline with data milestones#

The wind-down timeline below pairs each phase with the data milestone that must be met before the phase closes. Phases overlap in practice and their length depends on the company, so the timeline uses sequence rather than dates.

Read the table as a set of gates. A phase can start before the previous one ends, but its data milestone should be met before the irreversible actions in the next phase begin.

Before layoffs: capture knowledge, not just files#

Before layoffs, the priority is capturing knowledge about the records as well as the records themselves. The engineer who built the Jira automations or the dispatcher who knows which job fields were actually used cannot be replaced by an export file.

Layoffs carry their own legal obligations, and notice laws such as the federal WARN Act and similar state laws may apply to larger reductions. Counsel sets that schedule; the data steps fit inside it rather than delaying it.

• 1Name an owner for every system and write down what the owner knows: what it holds, which fields matter, which integrations write to it.
• 2Transfer admin rights and multi-factor recovery to accounts the wind-down team controls.
• 3Ask each owner to document export routes and known gaps.
• 4Offer short consulting arrangements to the few people whose knowledge cannot be written down in time.
• 5Tell employees plainly that company records are being preserved and assessed, so nobody deletes or copies data on the way out.

Before cancellations: the export gate#

The export gate is a single rule: no system is cancelled or downgraded until its records are exported, the export is checked against the live system, and each record family has a retention class and a licensing decision. Cancelling first and checking later is how records are lost for good.

Read each vendor's terms on suspension and post-termination access before setting cancellation dates. Some plans renew automatically, some prepaid annual plans can simply run to term while exports finish, and some vendors delete data once an account closes. Set cancellation dates after the exports, never the other way round.

How long does a wind-down take?#

How long a wind-down takes depends on the path and the complications, from a solvent dissolution with few creditors to a court-supervised process with contested claims. A handful of factors usually set the pace, and none of them is the data work.

The data steps rarely extend that schedule when they start early, because the first assessment uses metadata only and runs in parallel with other work. They extend it when they start late: once the team is gone and systems are suspended, recovering records means negotiating with vendors and rebuilding knowledge from scratch.

• 1Path: a solvent voluntary dissolution usually moves faster than an ABC, receivership or bankruptcy with court steps.
• 2Claims: the state's process for notifying and resolving creditor claims, and any disputes that arise.
• 3Contracts: leases, customer commitments and vendor terms that run to fixed end dates.
• 4Tax: final returns and any state clearances required around dissolution.
• 5Asset sales: marketing, diligence and closing for equipment, intellectual property or the customer base.
• 6People: final pay, benefits continuation and any required notice periods for layoffs.

Timing mistakes that cost records#

Timing mistakes follow one pattern: an irreversible step is taken before the data step it depends on. Each item below is easy to avoid when the plan lists data milestones next to the legal and financial ones.

Most of these mistakes happen in the same short stretch between the board decision and the first round of notices, which is why the data owner should be named at the decision meeting itself.

• 1Laying off the IT administrator before admin access and multi-factor recovery have been moved.
• 2Cancelling the company card before reviewing which tools bill to it.
• 3Removing user licenses in the email system before checking what happens to those mailboxes and exporting shared mail.
• 4Filing dissolution before deciding who will hold the records and who can sign any license.
• 5Treating the data assessment as optional work for after the shutdown.

Illustrative: a family-owned HVAC contractor closes in order#

Illustrative: the owner of a fictional HVAC and plumbing contractor with about 90 employees decides to retire and close after no buyer emerges. The company has run on ServiceTitan for years, with QuickBooks for accounting and Google Workspace for the office.

At the decision stage, the office manager lists every system and what it holds: estimates, jobs, dispatch notes, invoices, warranty callbacks and technician photos. Admin access moves to the owner's account before layoff notices go out. The dispatcher, who knows which job fields were filled in consistently, stays through the operations taper to check the ServiceTitan export.

ServiceTitan is cancelled only after its export passes the gate. The owner assesses the job and warranty records for licensing before filing dissolution, signs as the company's authorized officer, and hands the remaining archive to the company's accountant as records custodian with a written deletion schedule.

How SourceX fits into a wind-down plan#

SourceX fits at the decision stage. A metadata-only fit check, based on the SourceX Enterprise Data Value Framework, gives the board an early read on whether any record families are likely to interest AI developers before anyone is laid off or anything is cancelled. Typical fit is a company with 50 or more full-time employees at its peak and several years of operating history.

If a package proceeds, the SourceX five-step transaction of Supply, Rights, Preparation, Approval and Delivery runs alongside the wind-down, with the company approving each step. Nothing is shared during the initial assessment, and value becomes known only once a buyer engages.