Disclosure: Exafort is an implementation partner for VeroTX. Arun Kanchi is a co-founder of both companies. We write about VeroTX because we deploy it, and this post reflects delivery experience rather than product marketing.
Procurement software rarely fails at the demo. It fails at week nine, when the vendor master turns out to have four records for the same supplier and nobody owns the approval thresholds.
Exafort deploys VeroTX Procurement for clients who already run NetSuite, Salesforce, or both. This is what the work actually consists of, including the parts that have nothing to do with AI.
What the business case rests on
Two published benchmarks, so you can size this before talking to anyone.
APQC benchmarking data published in April 2026 finds organizations spend from roughly $14 to more than $54 to process a single purchase order, and frames the resulting spread as a multi-million dollar annual difference at enterprise transaction volumes.
The Hackett Group's 2025 Digital World Class Procurement research found top-quartile teams run 58% shorter requisition-to-purchase-order cycle times, with 31% fewer full-time employees and 19% lower cost as a percentage of spend.
Read the second one carefully. It establishes that requisition-to-PO cycle time separates top from median performers. It does not attribute that gap to AI. We mention it because clients are shown that number by vendors who imply it does.
The three prerequisites nobody wants to hear about
We assess these before quoting delivery, because they determine the timeline more than the software does.
1. Vendor master quality. VeroCortex matches requisition intent semantically, so it tolerates a requester writing "Apple laptop, 14 inch" where the catalog says "MacBook Pro 14 M4." It does not fix a vendor master with duplicate records, inconsistent remit-to data, or suppliers who exist twice under slightly different legal names. Consolidation logic operating on a dirty master produces confidently wrong groupings.
Typical Exafort scope here: dedupe analysis, remit-to normalization, and a governance rule for who can create a vendor going forward. This is NetSuite work, not VeroTX work.
2. Someone has to own the thresholds. VeroCortex consolidates and splits requisitions autonomously within configured thresholds: spend ceilings, category scope, acceptable delivery-date compression, vendor-count limits on splits, and whether crossing an approval tier forces escalation regardless of spend.
Those thresholds are the entire governance model. Configured once and abandoned, they drift out of step with the business within two quarters. We will not go live until a named person owns them and a review cadence exists.
3. An honest integration inventory. Clients consistently underestimate how many systems touch a requisition. We map every one before design, because a system discovered in week seven is a re-plan.
Where the integration work actually lands
VeroLink Studio handles connectivity. For the NetSuite and Salesforce estates we work in, the recurring items are:
The multi-entity NetSuite clients are where estimates go wrong most often. Subsidiary-level approval hierarchies and intercompany implications are not a configuration checkbox.
Sequencing, and why we start narrow
We deliver initial deployments through FDESK, VeroTX's Forward Deployed Engineering Support and Kickstart service, which is available across all plan tiers.
The sequence we use:
Steps 2 and 3 are where inexperienced deployments skip ahead, then spend a quarter recovering credibility with the procurement team.
Set the measurement basis before go-live, not after
The most common reason a successful deployment cannot prove it succeeded: nobody agreed what "cycle time" meant beforehand.
We fix the definition against APQC's published measure, which counts calendar days including weekends, from requisition received to purchase order released to the supplier. Vendors frequently use a narrower definition that excludes queue time and produces flattering numbers.
We baseline before deployment, then measure from Execution Ledger timestamps after, using the same definition. It usually produces a less dramatic number than the marketing version, and it survives a CFO asking how it was calculated.
What we tell clients not to expect
Where to read more
FAQ
How long does a VeroTX Procurement deployment take? The software configuration is rarely the constraint. Vendor master condition, integration surface, and whether threshold ownership exists drive the timeline. We scope those three before committing to dates.
Do we need to replace NetSuite or Salesforce? No. VeroTX operates as an execution layer above them. NetSuite remains the system of record for POs and vendor master; Salesforce remains the system of record for customer commitments.
Who owns the approval thresholds after go-live? A named client-side owner, established before go-live. This is a delivery requirement rather than a recommendation.
What is FDESK? VeroTX's Forward Deployed Engineering Support and Kickstart service, available across all plan tiers. Exafort delivers procurement deployments through it.
