Skip to main content
Card class: Non-HeroCategory: Ecommerce Platform
The mix of GL journals by source. A Receivables spike is healthy automation; a Manual spike is a controls gap.

At a glance

Journals by Source breaks down every General Ledger journal posted in the window by its journal source: Manual, Receivables, Payables, Inventory, Spreadsheet, Subledger Accounting, and any custom source your implementation defined. The shape of this mix is a quiet but powerful health signal. A ledger dominated by Receivables, Payables, and Inventory journals is one where automation is doing the work. A ledger with a swelling Manual or Spreadsheet slice is one where humans are hand-keying entries the system should be generating, which is exactly what auditors flag and exactly where errors creep in. The card tells you the composition, not just the count.
What it countsThe count of GL journal headers grouped by journal source over the window. Standard Oracle Fusion sources include Manual, Spreadsheet (ADFdi upload), Receivables, Payables, Cost Management / Inventory, Assets, Payroll, and Subledger Accounting, plus any custom source registered in the GL. The donut shows the proportional split; the underlying tooltip shows the count per source.
What a healthy mix looks likeThe bulk of journals coming from automated subledger sources (Receivables, Payables, Inventory) via Subledger Accounting. On a commerce-connected ledger, a Receivables-heavy mix is the expected and healthy pattern, because AutoInvoice and Create Accounting are turning ecommerce orders into revenue journals automatically.
What a worrying mix looks likeA Manual or Spreadsheet slice that grows period over period. Manual journals are legitimate for accruals, reclasses, and corrections, but a rising manual share signals a controls gap, a broken integration that people are working around by hand, or estimation churn in the close.
Business Unit scopeRespects the dashboard’s selected Business Unit and Ledger filter. By default rolls up every Business Unit and primary Ledger the connected role can see.
Time window30D (trailing 30 days)
Alert triggerNone - this is a diagnostic composition card, read alongside Manual JEs as % of Total which carries the alert
Rolesowner, finance

Calculation

Calculated automatically from your Oracle ERP Cloud data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.

Worked example

A large-enterprise consumer-goods company runs Oracle ERP Cloud with a Shopify Plus DTC channel and a B2B Adobe Commerce portal feeding Order Management. The 30-day window is 14 Mar 26 to 12 Apr 26. The controller pulls the source mix.
Journal sourceCountShareRead
Receivables (AutoInvoice via SLA)2,14061%Healthy - ecommerce revenue automating into the GL
Payables69020%Healthy - supplier bills posting via Create Accounting
Inventory / Cost Management41012%Healthy - COGS and inventory moves automating
Manual1805%Watch - up from 70 last period
Spreadsheet (ADFdi)702%Watch - month-end reclasses
Total3,490100%
Four things to notice:
  1. Receivables at 61% is the signature of a healthy commerce-connected ledger. The integration is doing its job: orders become Sales Orders, AutoInvoice creates Receivables transactions, and Subledger Accounting posts them to the GL automatically. You want this slice to be the largest.
  2. The Manual slice more than doubled, from 70 to 180. Five percent is not alarming in absolute terms, but the trend is the signal. Something is being hand-keyed that was not last period. The controller should ask why before it grows further.
  3. On investigation, the Manual spike traced to a broken Inventory integration. A cost-source change meant some inventory accounting events were not posting automatically, so the cost team booked them by hand. The right fix is to repair the integration, not to keep hand-keying. This is the kind of root cause the source mix surfaces that a single total never would.
  4. This card carries no alert by design. The composition is diagnostic. The alert lives on the companion Manual JEs as % of Total card, which fires when the manual ratio crosses its threshold. Read the two together: this card shows the full mix, that card watches the dangerous slice.

Sibling cards merchants should reference together

Journals by Source is a composition lens. It is most useful next to the cards that quantify the dangerous slices and the posting pipeline.
CardWhy pair it with Journals by Source
Manual JEs as % of TotalThe companion card that puts a threshold and an alert on the Manual slice this card shows.
Journals in Error (PostedFlag=E)Tells you which source the error journals came from, so you know whether to chase the integration or a human uploader.
Open (Unposted) JournalsA backlog by source helps you see whether one source is clogging the posting queue.
Subledger-to-GL Posting Interface Errors (24h)When Receivables, Payables, or Inventory automation breaks, people fall back to manual journals; this card shows the source-side failure.
Accrual Reversals (last close)Manual accruals and their reversals are a recurring slice of the manual source; high reversals plus high manual share is an estimation-quality flag.
Oracle Fusion Health ScoreThe composite that a deteriorating source mix feeds into.

Reconciling against Oracle ERP Cloud

Where to look in Oracle ERP Cloud: The closest native equivalents in the Oracle Fusion UI are:
Navigator → General Accounting → Journals → Manage Journals (group or filter by Source) Reports and Analytics → OTBI → Financials → General Ledger - Journals Real Time (Journal Source is a standard dimension) Navigator → General Accounting → Period Close → Close Status (journal counts feeding the close, by source)
In OTBI, the General Ledger Journals Real Time subject area exposes Journal Source as a dimension, so you can build the exact same donut. The counts per source should match this card when you select the same period, Business Unit, and Ledger scope. Common mistakes when comparing against Oracle’s own reports:
  • Mixing posted and unposted. This card counts journals over the window; be clear whether your Oracle query includes unposted journals or only posted ones, and match the card’s setting.
  • Custom source confusion. Many Fortune 500 implementations register custom journal sources (for example a named integration source). If your Oracle query buckets these differently from the card, the slices will not line up. Confirm the source list.
  • Subledger Accounting labelling. Journals that originate in subledgers can show with the subledger source (Receivables, Payables, Inventory) or under a Subledger Accounting umbrella, depending on how SLA was configured. Check which convention your implementation uses.
Why our number may legitimately differ from Oracle’s reports:
ReasonDirectionWhy
Source taxonomyEitherCustom journal sources may be grouped differently between the card and an ad-hoc Oracle query. Match the source list to reconcile.
Posted vs allEitherIf the card counts all journals in the window but your Oracle report filters to posted only, the mix shifts. Align the posting-status filter.
Window boundarySmallThe trailing 30-day window and Oracle’s period or date-range filter can include or exclude journals near the boundary.
Business Unit / Ledger scopeEitherThe card rolls up every Business Unit and Ledger the role can see; a scoped Oracle query shows a narrower mix.

Known limitations / merchant FAQs

Is a high Receivables slice good or bad? Good, on a commerce-connected ledger. Receivables journals are what AutoInvoice and Subledger Accounting produce when ecommerce orders turn into invoiced revenue. A Receivables-heavy mix means your integration is automating revenue into the GL, which is exactly what you want. The slice to worry about is Manual, not Receivables. Why is a Manual spike a controls gap rather than just busy accountants? Manual journals bypass the automated subledger pipeline and its built-in validations. A few are always legitimate (accruals, reclasses, corrections), but a rising manual share usually means people are hand-keying entries that an automated source should generate. That is slower, error-prone, harder to audit, and often a sign that an integration broke and nobody fixed it. Auditors specifically test the manual journal population because it is where misstatement risk concentrates. Why does this card have no alert when a Manual spike is the thing to watch? The composition view is diagnostic, and a healthy mix varies by business. The alert lives on the dedicated Manual JEs as % of Total card, which applies a clear ratio threshold. Keeping the alert on a single, well-defined ratio avoids noisy alerting on a card whose normal shape differs between merchants. Can the Spreadsheet source be ignored? Not entirely. Spreadsheet journals come from the ADFdi (Application Development Framework Desktop Integration) upload tool, which is a legitimate and common way to load month-end reclasses and bulk corrections. A modest Spreadsheet slice at period-end is normal. A large or growing one mid-period suggests the same automation gap a Manual spike does. How does this card handle Subledger Accounting journals? Subledger Accounting (SLA) is Oracle’s engine that turns subledger transactions into GL journals. Depending on configuration, the resulting journals can carry the originating subledger source (Receivables, Payables, Inventory) or a Subledger Accounting source label. The card uses whatever your GL records. If your mix looks unexpected, confirm how SLA labels its output in your implementation. Does the mix help diagnose a broken integration? Yes, that is one of its best uses. When an automated source slice shrinks and the Manual slice grows at the same time, it is a strong signal that an integration path broke and people are working around it by hand. The worked example above is exactly that pattern: an Inventory integration failure showing up as a Manual spike.

Tracked live in Vortex IQ Nerve Centre

Journals by Source (JeSource) is one of hundreds of KPI pulses Vortex IQ tracks across Oracle ERP Cloud and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.