Results · the proof center

Proof, not promises —
structured for when it's real.

We're pre-launch, so we won't fabricate proof or borrow logos we haven't earned. Here's exactly how we'll present results once they exist — and the attributed benchmarks we build on until then.

01The case-study template

Every MarginArc case will read like a diligence memo: a real situation, the wedge we deployed, an audited before/after, and the owner's own words. Names are withheld by default; your data stays on US infrastructure; references on request.

🟦 Distribution · order intake

[Named industrial/MRO distributor, ~$30M]

[PLACEHOLDER]
$XXX,XXX
recovered / year
$X.XX
cost per order
XX%
touchless rate
X.X
FTEs freed
SituationHigh inbound order volume by email and fax; two staff keying into Prophet 21.
The wedgeAI order intake on the email/fax channel, validated against P21.
Before → aftercost-per-order $X.XX → $X.XX; touchless XX% → XX%.
In their words"We measured it before we believed it — then it held."
🟦 Distribution · AP automation

[Named electrical distributor]

[PLACEHOLDER]
~XX%
lower cost / invoice
$X.XX
cost per invoice
X days
faster close
$XX,XXX
early-pay capture
SituationManual AP matching, missed early-pay discounts.
The wedgeAP invoice automation with 3-way match against the ERP.
Before → aftercost-per-invoice ~$XX → ~$X.XX (the fallback wedge's cleaner math).
In their words"The payback was obvious once the number was real."
🟧 Insurance · COI issuance

[Named independent agency] [VERIFY]

[PLACEHOLDER]
~XX hrs
freed / week
~min
COI turnaround
X.X
CSR-equiv. capacity
—
E&O: fewer touches
SituationCOIs piling up in a shared inbox; submissions re-keyed across portals.
The wedgeAutomated COI generation + policy checking against the AMS.
Before → afterturnaround end-of-day → minutes; hours/week freed to write more.
In their words"We can take on more business without adding a desk."

All figures above are [PLACEHOLDER] placeholders — populated only from a real engagement's audited baseline. Zero fabricated clients, quotes, or results.

02Until our cases exist

We build on attributed benchmarks —
and replace them with yours.

Graybar processed 83,000 documents with 9.5M line items through Conexiom in H1 2021; Sonepar saved ~1,000 hours/month; Werner Electric moved order accuracy 96% → 100%.Vendor-reported (Conexiom). Distribution.
Manual orders cost ~$8–15 each (Mirage Metrics); automation cuts processing time up to ~95%; top performers reach 90%+ touchless.Vendor/industry benchmarks. Distribution.
Manual AP runs $10–22/invoice (APQC); best-in-class automated ~$2.78 (Ardent Partners).Analyst benchmarks. Distribution AP.
Independent agencies post strong revenue/employee ($228,321, Big "I"/Reagan Consulting 2025) — dense document work caps the growth that capacity could unlock.Industry study. Insurance. [VERIFY]
03How we'll measure yours

The harness that turns work
into a case study.

We instrument the baseline and the post-go-live state on annualized volume, so the result isn't a vendor slide — it's an audited before/after you could hand to a buyer or a board. The same measured number triggers our guarantee.

Our measurement approach →
Discretion as a feature
  • Your name stays off our site by default.
  • Your ERP/AMS data stays on US infrastructure.
  • References shared on request, with your permission.
  • Every figure shows its work — no black box.

    • Be the first case study.

    Start with your number; if it's worth it, we'll build it, measure it, and — with your permission — let it stand for the next owner who's skeptical too.

    Get my ROI estimateBook a 20-min call