A coverage statement such as "twelve years of history" compresses at least five facts into one number: where the records start, whether that start is known, whether there are holes, whether a migration rewrote the dates, and whether the system is still adding records today.
Why "twelve years of history" is not a coverage statement
Library and archive metadata already treats the period a resource covers as its own field. The Dublin Core Metadata Initiative defines temporal coverage as the "Temporal characteristics of the resource", and an earlier DCMI working-group review drew the line clearly: temporal coverage refers to the subject period of a resource, not to when it was created or made available. For a company archive, that maps to a practical rule. The span that matters is when the work in the records happened, not when the files were copied, scanned or imported.
A single number cannot carry that. Four things go missing:
- The start's certainty. "Since 2011" may mean the first record is dated March 2011, or that someone believes the system went in around then.
- Gaps. A migration that took seven months, a year of paper files that were never scanned, a purge under a retention rule.
- The date basis. After a migration, older records can carry the date they were imported rather than the date of the work. A range read from that field is wrong in a way nobody notices.
- The end. A live system grows every day, so a coverage statement without an as-of date expires the day it is written.
SourceX's screening profile looks for several years in operation, and some programs set higher thresholds. That makes the precise span worth stating, because a reviewer cannot tell from "twelve years" whether the usable history is twelve years or four. A system's go-live year has the same weakness, as the systems questionnaire for platform teams explains.
Eight date questions for the people who run the systems
Ask for these lines in conversation, through the CEO or the person the CEO names, usually the administrator of each system. Nobody needs to export anything to answer them.
| Line | Question to ask | Who usually knows |
|---|---|---|
| Era | Which systems held this kind of record, in order? | Operations lead |
| First record | What is the month of the earliest record in each system? | Administrator |
| Last record | Is the system still in use? If not, when did it stop? | Administrator |
| Migrations | When did records move between systems, and what came across? | Administrator or the outside IT provider |
| Date basis | When old records were moved, did they keep their original dates? | Administrator |
| Purges | Does any retention rule delete older records automatically? | Administrator or counsel |
| Thin years | Are any years noticeably sparse, and why? | Operations lead |
| Source and date | Who answered, and on what date? | You |
Two program facts sit next to the coverage line in any handoff. Introduced companies need 50 or more full-time employees at peak, contractors excluded. Companies in the U.S., Canada and the EU qualify, as do companies in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia when at least 70% of the data is in English. Who in the company can answer the date questions, and who decides, is mapped era by era in who can open the door to a company's operating history.
The date coverage notation guide
The notation borrows from the Library of Congress's Extended Date/Time Format (EDTF), whose features are now part of ISO 8601 through its 2019 revision. In EDTF, "?" marks a date as uncertain, "~" as approximate and "%" as both; a slash separates the start and end of a range; a double dot stands for an end that is not specified; and an empty end means it is unknown. What follows is a house style for referral summaries built on those symbols. It is meant to be read by people, and not every line is guaranteed to validate as strict EDTF.
| Situation | Write | Read it as | Instead of |
|---|---|---|---|
| Known, closed range | 2014-03/2021-06 | March 2014 through June 2021 | "Seven years" |
| System still in use | 2021-07/2026-10, live, as of 2026-10-27 | Still growing; checked on that date | "2021 to present" |
| Approximate start | 2009~ start | Around 2009, by someone's estimate | "Since 2009" |
| Uncertain start | 2011? start | Possibly 2011, not confirmed | "Since 2011" |
| Uncertain and approximate | 2008% start | A rough guess that may be wrong | "Late 2000s" |
| End unknown | 2014-03/ | Starts March 2014; nobody knows when it stopped | "From 2014" |
| End not stated by the source | 2014-03/.. | The person did not say; ask | Leaving the end off silently |
| Gap | gap 2017-03/2017-09 (migration) | No records for those months, and why | Ignoring it |
| Rolling purge | earliest moves forward: 7-year retention | The start date changes every month | A fixed start date |
| Date basis | basis: work date / entry date / import date | Which date the range is counted on | Not saying |
| Source | (administrator, 2026-10-21) | Who said it and when | An unattributed line |
Use months whenever someone knows them. Fiscal years are ambiguous across companies, so convert "FY2018" into months or say which fiscal calendar the company uses. Prefer an as-of date to the double dot for a live system: "live, as of" tells a reviewer more than "not specified".
A coverage line, built step by step (Illustrative)
A generic 210-person HVAC services company, with invented facts. The CEO's first description was "about twenty years of job history". After a few questions to the administrator and the operations lead, the summary carries this:
| Era | Coverage line | What it reveals |
|---|---|---|
| Paper job tickets, later scanned | 2006~/2012-03, scanned images, not indexed (operations lead, 2026-10-21) | The start is a recollection, and the records are images, not fields |
| Field-service system | 2012-04/2019-07, basis: import date 2019-09 for every record in the new system; original dates in a legacy field, to confirm (administrator, 2026-10-21) | Read naively, seven years of work orders would appear to start in September 2019 |
| Current field-service system | 2019-10/2026-10, live, as of 2026-10-27 (administrator, 2026-10-21) | Still growing |
| Known gaps | gap 2019-08/2019-09 (cutover); no other gaps known | Short and explained |
| Summary line | Work orders: 2012-04/2026-10, live, as of 2026-10-27; gap 2019-08/2019-09; date basis before 2019-10 to confirm; scans 2006~/2012-03 (source: administrator) | One line a reviewer can check on the first call |
"About twenty years" survives as roughly fourteen and a half years of structured work orders, from April 2012 to the as-of date with a two-month gap at the cutover, plus about six years of scanned images whose start is approximate. That is a smaller claim and a far more credible one. Migrations are the first place to check a date basis; the live guides on old CRM data after a migration and historical data in an accounting migration show what tends to move and what tends to stay behind.
Density, purges and the questions under the line
Coverage is only one part of an archive's description, and it has its own kinds of uncertainty.
- Estimates. At referral stage, density over time is an impression, not a count: "records thin before 2015" is a fair line if someone said it, with their name and the date. Do not turn density into record counts nobody has made.
- Constraints. A retention rule, a legal hold the company mentions, or an era held by an outside provider narrows what could be described in detail later. Write it down without deciding what it means.
- Unanswered questions. Put them under the coverage line as questions: "Did the 2019 migration keep original dates?" "Are the scanned tickets searchable?" Each names who could answer.
If you track every line's status, the codes in recording what you do not know about a dataset fit the notation: a confirmed range, an estimated start marked "~", an unknown end left open. Spotting the companies with long histories in the first place is covered in recognizing a business with several years of operating history.
Handing over a coverage line the CEO approved
- The CEO approves the coverage lines. Send them exactly as they will appear. The CEO can remove any line; whatever is left is the company's own description.
- Nothing else travels. No exports, screenshots or sample records, not even to "prove" a start date. SourceX checks the actual dataset in its own review, after the company decides to proceed.
- The partner submits; the company applies. Under the terms, the first valid introduction of a new company inside the 90-day attribution window earns the credit when that company submits a verified application. The SourceX Channel Partner Agreement has to carry your signature before the partner submission goes in.
- The first call starts with the dates. Confirm each era's range, settle the date basis after each migration, and list any purge rule. These three items can change how a reviewer reads the whole archive.
Where a coverage line can mislead
- Self-reported dates. Every range on the page comes from someone's answer, not a query. The source and date tags exist so a reader can weigh them.
- Import dates that look like work dates. An easy error to make, and invisible from outside. Ask about date basis for every migration, every time.
- Peak and present are different questions. A company's size test uses peak full-time headcount; its coverage line describes records. A company that shrank can have a long, strong archive, and a growing one may have a short one.
- Programs weigh history differently. SourceX applies each program's own criteria, and some set higher thresholds for years in operation. A long span is not a guarantee of fit, and a short one is not automatically a no.
- Precision can be false. A month written for a date someone guessed is worse than a year marked "~". The notation only helps if the marks are used honestly.
Methods note. The notation was assembled on 2026-10-27 from the Library of Congress EDTF specification, DCMI's temporal coverage term and SourceX's published screening profile and terms. It standardizes how a referral summary states time coverage. It does not measure record quality, demand or value, and SourceX has not published evidence on whether coverage lines written this way change review outcomes. A partner can test it by comparing, over a set of introductions, how often the first call corrects a coverage claim written as a year count versus one written in this notation; no such figures have been published.
For other ways to describe records before an introduction, browse the guides hub.
Next step
Refer a company. No cap on rewards. Default reward on a company's first qualifying transaction: the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; further qualifying transactions first agreed within 12 months of the first Closing pay 15%. Eligibility and payment conditions apply, including the USD 100,000 Minimum Transaction, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.
When the CEO has approved a coverage line with its gaps and date basis showing, get a partner link via the registration page, check the company's fit on who qualifies, and read the program terms in full before you submit.
Describing a company's archive is part of making a referral; looking at SourceX as a possible investment is a separate decision, and investor materials for that decision can be requested through the SourceX investor overview.
This article is for information only and is not an offer to sell or a solicitation of an offer to buy securities. Any offering by SourceX is made only to verified accredited investors through its offering documents.