Leadership and readiness
Data licensing operational history summary: a one-page ten-year brief
By SourceX Editorial · Reviewed by Noah Loul ·
8 sections
Short answer
A useful data licensing operational history summary fits ten years onto one page: each system and the years it was the system of record, rough record volumes per period, how deep the history goes, where migrations and retention purges left gaps, and which contract and policy versions governed each period. It closes with the decision leadership can make now and the evidence still missing.
Record timeline
Data licensing operational
Key takeaways
- Write the brief by system and period, because a decade of records is usually several datasets separated by migrations and purges.
- Count volume from metadata and give ranges; nothing needs to be exported to decide whether to investigate.
- Record history depth separately from period length; two linked years with outcomes can matter more than eight years of identifiers.
- Date every contract and policy version, since the rights basis often changes partway through the decade.
- Mark acquired, carved-out or vendor-held periods as blocked until counsel reviews them.
What should a concise company data brief say about ten years of operating history?
It should say, on one page, which systems held which records in which years, roughly how many records each period contains, where the history has gaps, and on what basis the company holds and may license each period. It ends with the one decision leadership is being asked to make and the evidence still missing.
Most leadership teams describe their history in a sentence such as "we have ten years of customer data." That sentence cannot be checked, priced or scoped. A buyer, a licensing partner or the company's own counsel needs to know that the first four years live in a retired accounting package, that field changes before a certain date were never tracked, and that customers signed a different contract after a specific year.
The brief below is built for a CEO or owner deciding whether a scoped investigation into licensing operational records is worth anyone's time. It uses metadata only, takes a few hours of system owners' time, and states its assumptions conservatively so the decision does not rest on a number that later shrinks.
Why a decade of records is usually three or four datasets
Ten calendar years rarely means ten consistent years. Companies change systems, purge records under retention policies and switch on tracking features partway through, so the honest unit of a brief is the period within a system, not the decade.
At a glance
Why a decade of records is usually three or four datasets
Product defaults alone create seams. Salesforce documents that field history is kept for up to 18 months, and up to 24 months through the API, unless Field Audit Trail is used to keep it longer, so an org that never bought that feature may hold current values going back years but change histories for only the latest two. Zendesk archives closed tickets about 120 days after they close; archived tickets drop out of views and reports but remain reachable through search and the API, and its ticket deletion schedules act only on archived tickets, which cannot be restored once deleted.
Retention policies create the other seams. Many companies set schedules around legal minimums: the IRS generally asks businesses to keep records supporting a return for three years, with longer periods in specific cases such as seven years for a bad debt deduction, and the Department of Labor's FLSA fact sheet asks employers to keep payroll records for at least three years and wage computation records for two. A company that deleted on those schedules may have full detail for recent years and only summaries before that. The brief should say so rather than imply the whole decade survives.
- System eras: when each system went live and when it was retired or migrated.
- Migration seams: what moved in each migration, such as open records only, current values only, or full history.
- Tracking seams: dates when audit logs, field history or attachments began to be captured.
- Purge seams: retention schedules that deleted detail after a set period.
The source system, period, volume and rights inventory
The inventory is one row per system per period, with five things written in each row. It is the asset of this page and can be copied as is into a spreadsheet; the right-hand column of the field guide says who must confirm each entry before the brief goes to leadership.
- Write ranges, not point counts; a brief that says "roughly 40,000 to 60,000 work orders" survives scrutiny better than an exact number taken from one report.
- Leave a cell marked unknown rather than guessing; the open-question field exists so that unknowns become tasks, not assumptions.
| Field | What to write | How to fill it | Who confirms |
|---|---|---|---|
| System and era | Product name, instance and the dates it was the system of record | Contract start, go-live and decommission dates | IT or the system owner |
| Record families | The kinds of records the period holds, such as work orders, invoices or tickets | Object or table list from the admin console | System owner |
| Period covered | First and last month with records in this system | Earliest and latest created dates | System owner |
| Volume range | Order-of-magnitude count per record family per year | Report or API count; round down and give a range | System owner |
| History depth | Whether status changes, notes, attachments and outcomes are kept, and since when | Audit or field-history settings and a sample of old records | System owner |
| Gaps and seams | Missing months, purges, partial migrations, archived records | Migration notes, retention schedule, record count by month | IT and records lead |
| Rights basis | Which customer, vendor and employee terms governed the period | Contract and policy versions by effective date | Legal or the contract owner |
| Personal data present | Whether the records name or describe people, and which kinds | Field list review, not a data pull | Privacy lead |
| Open question | The one unknown that most affects this row | Written by whoever filled the row | Executive sponsor |
Counting volume from metadata, before anything is exported
Volume can be counted from the systems' own reports and record counts, without exporting a single record. A brief built this way keeps personal data where it is and lets the company stop at no cost if the answer is not interesting.
Count by created year and by record family, then check three things. First, whether counts drop sharply in a year that matches a migration or purge. Second, whether the oldest records still carry the fields that make them useful, such as resolution notes or final status, or only identifiers. Third, whether attachments, emails and comments are stored with the record or in a separate system that has its own retention.
| Check | What to look at | What a problem looks like | What to write in the brief |
|---|---|---|---|
| Count by year | Records created per year for each record family | A cliff in one year with no business reason | A seam at that year and its likely cause |
| Field fill rate | Share of old records with key fields filled | Status present, but notes and outcomes empty before a date | History depth starts at that date |
| Linked records | Whether old tickets link to orders, customers or invoices | Links broken after a migration | Period usable on its own, not joined |
| Attachments | Where files are stored and since when | Files kept in a separate system now retired | Attachments missing for the period |
| Archived records | Records hidden from views but reachable through search or API | Reports undercount because archives are excluded | Count from the API, not from standard reports |
Rights that changed partway through the decade
The rights basis is usually the least stable column across a decade. Customer contracts, terms of service, privacy notices and vendor agreements are rewritten every few years, and each version may say something different about how the company can use records it holds.
Treat each version as defining its own period. The FTC staff has warned that a company adopting more permissive data practices, such as using data for AI training, and telling people only through a surreptitious, retroactive change to its terms or privacy policy could be acting unfairly or deceptively. A brief that dates each policy version keeps leadership from assuming that today's terms reach back over records collected under older ones.
Keep four ideas apart in this column. Access means the company can open and export the records. Privacy permission means the notices and consents in force allow a given use of personal data. Ownership means the company, rather than a customer or vendor, holds the records. Contractual licensing rights mean no agreement prevents the company from licensing them. De-identification can reduce privacy risk, but it does not by itself create a right to license records the company was not free to share.
Illustrative worked brief: a field services company with three system eras
The example below is invented to show the format. It describes a regional field services company whose records span three system eras; all counts are illustrative ranges, not data from any real business.
| Era and system | Period | Records and volume (illustrative) | History depth and gaps | Rights basis and open question |
|---|---|---|---|---|
| Era 1: desktop dispatch software | Years 1 to 4 | Work orders, roughly tens of thousands | Summary fields only; technician notes not migrated | Older service terms; does the old vendor contract restrict exports? |
| Era 2: cloud field service platform | Years 5 to 8 | Work orders, quotes and invoices, roughly 100,000 or more | Full notes and photos; status history from year 6 only | Customer terms version 2; privacy notice changed in year 7 |
| Era 3: current platform and help desk | Years 9 to 10 | Work orders and tickets, roughly 50,000 to 80,000 | Full history; closed tickets archived but retrievable by API | Current terms; tickets include customer names and addresses |
| Accounting system, all eras | Years 3 to 10 | Invoices and payments, roughly 200,000 | Years 1 and 2 purged under the retention schedule | Owned by the company; vendor names are business data |
| Shared inbox | Years 6 to 10 | Customer emails, volume unknown | Not linked to work orders | Unclear; excluded from the first scope |
What can leadership decide now, and which question needs more evidence first?
Leadership can decide now whether to fund a short, scoped investigation of the period with the deepest history and the clearest rights basis. It should not yet decide on scope, terms or timing for the whole decade, because the rights column and the history-depth column almost always contain questions only counsel and system owners can answer.
At a glance
What can leadership decide now, and which question needs more evidence first
In the worked brief, Era 2 and Era 3 are candidates for an investigation: they hold full notes, outcomes and links, and their terms can be read. Era 1 needs one answer first, whether the old vendor agreement limits exports. The shared inbox is out of scope until someone can say whose communications it holds and under which notice.
There is also a meaningful exception. When an era came from an acquired company, a carve-out, or a system operated by a vendor that holds the records under its own terms, the brief should mark the period as blocked rather than estimate it. Those periods depend on transaction documents and vendor contracts that a leadership brief cannot summarize, and counsel should review the specific facts before anyone counts them.
- Decide now: whether to investigate one or two periods with deep history and a readable rights basis.
- Ask system owners: whether the oldest records keep notes, outcomes and links, measured on a sample.
- Ask counsel: what each dated policy and contract version allows, and whether any vendor or transaction agreement restricts use.
- Defer: any statement about the whole decade until every row has a rights basis.
How SourceX uses a brief like this in its first assessment
A brief like this is close to what SourceX's first step asks for. That step is a short company assessment, and a company does not upload any data to learn whether it qualifies.
The periods matter for fit. Buyers generally want years of consistent history, typically two or more, and tens of thousands of records or more in a single system, although a smaller set can qualify when the work it captures is rare, and every case is judged on its own. Eligibility has two tiers: SourceX works with U.S., Canadian and EU companies that had 50 or more full-time employees at their peak, with contractors left out of the count, and a company in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE or Saudi Arabia can be considered when at least 90 percent of the dataset it actually offers is in English. No records change hands until the company has signed an agreement whose scope and terms it approved, so the periods in the brief can become that scope.
Frequently asked questions
How long should a company data brief be?
One page for the decision and one inventory table behind it. If the brief needs more, it usually means periods are being merged that should be listed separately, or legal questions are being answered in the brief instead of being listed as open.
Do we need to export records to write the brief?
No. Record counts, field settings, migration notes and policy versions are enough. Keeping the brief to metadata keeps personal data in place and lets the company stop without any cleanup.
What if our oldest system is no longer running?
List it as its own era with the dates it was the system of record, then say whether a backup, an archive export or nothing survives. A retired system with no readable copy is a gap, not a period to estimate.
Who should sign off on the rights column?
Whoever owns the contracts and policies, normally legal or outside counsel, working from dated versions. System owners can say what the records contain, but only counsel can say what each version allows.
Does ten years of history make records more valuable?
Only where the history is consistent and complete. Two well-linked years with outcomes can be worth more to a buyer than eight years of identifiers, which is why the brief separates period length from history depth.
Sources
- Salesforce retains field history for up to 18 months, or 24 months through the API, unless Field Audit Trail is used to keep it longer. Source
- Zendesk archives closed tickets about 120 days after they close; archived tickets are excluded from views and reports but remain available through search and the API. Source
- Zendesk ticket deletion schedules delete only archived tickets, and deleted tickets cannot be restored. Source
- The IRS generally asks businesses to keep records for three years, and seven years for a worthless securities or bad debt deduction claim. Source
- Under the FLSA, employers keep payroll records for at least three years and wage computation records for two years. Source
- FTC staff warned that adopting more permissive data practices, such as AI training, through surreptitious retroactive changes to terms or privacy policies could be unfair or deceptive. Source
See if your company qualifies
A short company assessment. No data uploads are needed.