Scope and publisher
This view covers 56 entries across 30 criteria for ERP and finance automation. The underlying research snapshot is dated September 2026. The scoring method was updated on September 25, 2026.
StackERP is published by Aviron, a vendor included in the comparison. The same formula and weights apply to every entry within a dataset. The choice of criteria, their weights and the underlying assessments remain editorial decisions made from the publisher’s perspective.
How scoring works
Each criterion receives an assessment of 1 for coverage, 0.5 for partial coverage or 0 when coverage is not established in the dataset. Criterion weights range from 1 to 3. A criterion with weight 3 contributes three times as much as one with weight 1 within the same calculation.
Overall coverage: the complete research grid
Multiply each coverage assessment by its criterion weight, add the results, then divide by the total weight. This produces a coverage fraction between 0 and 1, which can also be displayed as a percentage.
The total weight for this grid is 60. A score of 1 means no coverage is recorded; 5 means full coverage across the grid. This is the score shown on individual vendor scorecards and in the comparator when no requirements are selected.
Weighted fit: your requirements plus overall coverage
When requirements are selected, the fit score combines two components. 70% comes from coverage of the selected requirements, using their original relative criterion weights. 30% comes from coverage of the complete grid.
Fit score = 1 + 4 × fit coverage
This keeps the selected workflow as the main factor while retaining information about the broader platform. Covering one selected requirement completely no longer produces an automatic 5.00/5. The overall component still includes the selected criteria, so those requirements intentionally contribute through both components.
The 70/30 split is an editorial weighting, not a statistically validated measure of quality. All calculations use the unrounded coverage fractions. Only the final score is rounded to two decimal places. Vendors with identical weighted coverage can still receive the same score; no arbitrary bonus or decimal is added to break a tie.
These scores measure assessed coverage. They do not measure execution quality, support, customer satisfaction or security. The overall component can lower the fit score of a specialist that covers your selected workflow well. Decimal precision describes the calculation, not the certainty of the underlying research.
Procurement example
With only “Requisitions and purchase orders” selected, the current dataset produces the following results. The fit score combines the selected and overall coverage percentages with the same 70/30 weighting for every provider.
| Provider | Selected coverage | Overall coverage | Fit / 5 |
|---|---|---|---|
| Aviron | 100.00% | 92.50% | 4.91 |
| Coupa | 100.00% | 46.67% | 4.36 |
| Precoro | 100.00% | 40.00% | 4.28 |
Evidence and uncertainty
The original research draws on vendor positioning, product pages, documentation and public discussions. Some entries begin from a shared product-family baseline, then include adjustments. Not every value has an individual source URL or a verification record attached.
A zero can mean a capability is absent, outside the product’s scope, or simply not established by the research. It should not be read as proof that a vendor cannot meet the requirement. Likewise, a positive assessment is not a certification that the capability works in your environment.
Reported performance figures are statements recorded in the research, not measured results from a StackERP benchmark. Differences in volume, workflow and definitions can make vendor figures incomparable. The updated weighting does not change these underlying assessments or strengthen their evidence.
How filters work
Selecting requirements keeps entries with at least partial assessed coverage of every selected criterion. Their displayed score uses the weighted-fit formula above. Selecting additional criteria changes the selected component; clearing all requirements returns to the overall coverage score.
The home-page explorer uses the same fit calculation as the full comparator for the same selected requirements. Its default list is a curated selection; you can also sort by fit score or name. Individual vendor scorecards always display the overall score across the complete grid.
Aviron’s featured placement does not change its calculated score. If it does not match a selection, it remains in a separately labelled featured section. Other results in the full comparator are ordered by the displayed score.
Product categories are combined with “or”, as are workflow tags. Different filter groups are combined with “and”. The score-band filter always uses the overall score, even when a filtered result displays a fit score.
Criteria and weights
The current ERP & finance automation grid is below. Percentage shares are calculated from the criterion weight divided by 60. Displayed percentages are rounded; calculations use the original weights.
| Criterion | Group | Weight | Share of overall coverage |
|---|---|---|---|
| Procure-to-pay, end to endRequisition, purchase order, receipt, invoice, voucher and payment in one flow ; not one step of it. | Cycle covered | 3 | 5.00% |
| Order-to-cash, end to endCovers sell-side workflows such as billing, collections and cash application. | Cycle covered | 3 | 5.00% |
| Close and reconciliationMonth-end close, account reconciliation, accruals, GL substantiation. | Cycle covered | 2 | 3.33% |
| Requisitions and purchase ordersIntake, approval routing, PO creation and budget control before the invoice exists. | Cycle covered | 2 | 3.33% |
| Payment executionRuns the payment itself ; not only the approval that precedes it. | Cycle covered | 2 | 3.33% |
| Vendor statement reconciliationReconciles supplier statements against the transactions recorded in the ERP. | Cycle covered | 1 | 1.67% |
| Cards and employee expensesCorporate cards, expense reports, out-of-policy spend. | Cycle covered | 1 | 1.67% |
| AI agents that execute the workAgents that take an action and produce a record in the ERP, not a chatbot or a dashboard that advises. | How the work gets done | 3 | 5.00% |
| Reads non-standard documents without templatesHandles a supplier's own layout, a scanned PDF, a freight bill, without a per-vendor template to maintain. | How the work gets done | 3 | 5.00% |
| Three-way matchingInvoice against purchase order against receipt, with configurable tolerance. | How the work gets done | 2 | 3.33% |
| Judgment on coding and taxDetermines account coding, cost centres and tax treatment, with a review path for uncertain cases. | How the work gets done | 3 | 5.00% |
| Monitors and triages the AP mailboxWatches the shared inbox, classifies what arrives, and files it ; nobody sorts email by hand. | How the work gets done | 2 | 3.33% |
| Autonomous exception handlingResolves the exception itself where it can, and escalates with a recommendation where it cannot. | How the work gets done | 2 | 3.33% |
| Rules written in plain English by financeThe team writes or changes a rule in a sentence and it applies from the next document on. No developer, no ticket, no workflow builder. | How the work gets done | 3 | 5.00% |
| Every correction becomes a live ruleCaptures reviewed corrections as governed rules that can be applied to later transactions. | How the work gets done | 3 | 5.00% |
| Works across several ERPsNot tied to a single vendor's suite, and not a module you can only buy if you upgrade. | Fit with your ERP | 2 | 3.33% |
| Writes back into the ERP nativelyCreates the record inside the ERP through paths the ERP supports ; not a file drop somebody has to maintain. | Fit with your ERP | 3 | 5.00% |
| No replatforming, the ERP stays the system of recordYou keep the ERP you have, at the version you have. | Fit with your ERP | 2 | 3.33% |
| Supports legacy and on-premise ERPWorks on an on-premise or long-lived release, not only on the current cloud edition. | Fit with your ERP | 2 | 3.33% |
| Supports Oracle JD EdwardsDocumented support for EnterpriseOne, and ideally for World. | Fit with your ERP | 2 | 3.33% |
| Does not need a modern APICan connect through supported alternatives when a modern API is not available. Confirm the operational requirements. | Fit with your ERP | 2 | 3.33% |
| Audit trail on every transactionThe documents used, the reasoning, the approver and the timestamp travel with the record itself. | Control and trust | 2 | 3.33% |
| Scoped, revocable accessGranted system by system by your own administrator, narrow in scope, revocable in one click. No global admin account. | Control and trust | 2 | 3.33% |
| SOC 2 Type IIAn external audit of the controls, not a self-declared security page. | Control and trust | 1 | 1.67% |
| No model training on your dataWritten into the commercial terms, not just into a FAQ. | Control and trust | 1 | 1.67% |
| Choice of hosting regionMatters the moment a controller asks where the invoices physically sit. | Control and trust | 1 | 1.67% |
| Named customer referencesNamed customers, on the record, with numbers attached to them. | What you can verify | 2 | 3.33% |
| Published performance numbersRecords performance figures with their definitions, scope and supporting source. | What you can verify | 1 | 1.67% |
| Published pricingA grid you can read without booking a call. | What you can verify | 1 | 1.67% |
| Live in weeks, not quartersTime to a first real transaction in production, with a number attached to it. | What you can verify | 1 | 1.67% |
Corrections and source submissions
Send the product name, the relevant criterion, the page URL and a current supporting source. Changes should address the underlying evidence, rather than a request for a higher rank.
Submit a correction · Explore sources and coverage · Return to the comparison