Session Notes: Vendor Transition and the Ordova API

2026-06-01 · Megan Calloway, with Nate St. Pierre · two working sessions, on-site

Overview

This was a two-part working day. The first half was a walkthrough of Merrick Awards & Engraving's wholesale order entry process — from the moment a purchase order arrives by email through to the point where an order enters production — with Megan screen-sharing the actual workflow in Ordova Advance, the company's order management system. The second half stepped back to a related decision already in motion: the outsourced order-processing arrangement Merrick has been using is being wound down, and the work is coming back in-house. Tied to both threads was a specific question Nate raised about the Ordova API — what it exposes, whether it had been investigated, and which repeatable manual tasks it might eventually automate.

The order-entry portion was intentionally exploratory — less an interview, more a show-me-how-it-works observation. Megan was candid about what she thinks is and isn't improvable, and realistic about the limits of the OMS.

The Wholesale Channel: How Orders Work

Merrick operates in the promotional products industry as a supplier to distributors. They don't sell directly to end users. Instead, distributors search the industry distributor catalogs for items that fit their customers' needs, find Merrick listed as a supplier, verify the product details and pricing, and bring an approved order back to Merrick. At that point the distributor sends a purchase order along with artwork files — and order entry begins.

Once an order is complete, Merrick manufactures the product and ships it — often directly to the distributor's end customer, but in packaging that appears to come from the distributor. Merrick functions as the behind-the-scenes manufacturing partner. Distributors who work with Merrick regularly already understand what they need to provide (proper artwork files, complete PO details) because they're industry professionals. Megan noted this makes the wholesale channel different from retail, where a customer might send incomplete or incorrect files and need more hand-holding.

The Shared Inbox and Order Routing

All incoming orders land in a shared Outlook mailbox (orders@…). Multiple people have access, and it's organized using categories that map to different tasks and different team members. When an order comes in, it gets tagged so the right person knows what to do with it. Right now, order entry and back-end artwork processing are tagged for the outsourced vendor that handles initial entry into Ordova and the processing of artwork files. That arrangement is being ended — and this work is coming back in-house.

Order Entry in Ordova Advance

Megan walked through a sample order to show how entry actually works — a small name-badge order, a handful of pieces, placed through a regular distributor. When an order comes in, the first step is looking up the customer in Ordova. If they've ordered before, the standard approach is to clone a previous order from that customer. Cloning copies the basic structure — the same product configuration, artwork references, and most of the standing details — so the team isn't starting from scratch every time. From there, the fields specific to this order get filled in or updated.

The fields that have to be populated for each order include:

The distributor's purchase order number
Shipping details (address, shipping method)
Ship date and in-hands date
Event date, if applicable (some orders are time-critical because of graduations, conferences, or similar events)
Product and component details

That last item carries the complexity. A finished product like a name badge isn't a single line item in Ordova — it's built from multiple components. The example Megan showed had three: the magnet for the back, the packaging, and the white plastic substrate that gets sublimated. Each component is a separate entry in the order record.

Artwork attaches directly to the order record. Files arrive either as email attachments (EPS or PDF) or via a file-share link in the PO. They have to be downloaded from wherever they land and uploaded into Ordova — there's no direct transfer.

The Artwork Handoff

Once an order is entered, it goes to the art department for proofing. At Merrick, the same person often handles both order entry and art, which Megan said is fairly typical for this industry. They take the order and the submitted artwork, create a proof, and pass it back to order entry. From there the proof goes out to the distributor, who routes it to their end customer for approval. When the proof is approved, the order's status changes from orange to green in Ordova, and the workflow status updates to "in production." That's the signal the production floor uses to begin work.

Sourcing

Before production starts there's a sourcing step: checking whether the components needed are already in stock or have to be ordered from a vendor. For high-volume items like name-badge components, Merrick typically keeps inventory on hand. If something needs to be ordered, a purchase order goes out to the vendor, and the items get received and checked in when they arrive.

The Ordova Relationship

Megan gave significant context on the history with Ordova Advance. Merrick signed a contract with them five years ago but didn't go live until two years ago — a gap partly explained by the pandemic, but also by the fact that Ordova's out-of-the-box capabilities didn't fully match Merrick's use case. Ordova is built primarily for screen-printing and embroidery shops; Merrick's printing technologies required custom development work before the system could handle their production floor properly.

Even now, Janet — a key team member — is still spending roughly 20 hours a week optimizing the product setup in Ordova. Getting all of Merrick's products configured correctly has been an ongoing, multi-year process. The vendor relationship is complicated: Merrick has a dedicated success rep, but feature requests and support tickets often move slowly, and the system is deeply embedded in how the business operates. Megan's overall assessment: Ordova is more than halfway decent and it mostly works. She wouldn't want to switch. But it was built for a slightly different industry, and the compromises are real.

Friction Points in the Current Process

Copy-and-paste from PO to Ordova. The core order-entry task — reading the incoming purchase order and manually keying all of its information into Ordova — is entirely manual. Every field is transcribed by hand. Megan described it as "a bunch of copying and pasting."

File download and re-upload. Because Ordova is web-based, any artwork file that comes in by email has to be downloaded to a local device before it can be uploaded into the order record — an extra step that adds no value but can't be skipped. Megan mentioned that even Microsoft Power Automate might handle this, but she doesn't know how to set it up.

Art edits. When an artwork file needs revising, the process is: download from the order record, edit on the local device using industry-standard Adobe and vector-design tools, save, re-upload, and then re-tag. Megan counted it as roughly 10 clicks per revision. She doesn't think there's a clean solution — it's a consequence of how Ordova stores and retrieves files — but she named it as a real friction point.

Quality Control

QC at Merrick happens at the front end of the job, during setup and proofing. By the time an order is in production, it's past the point where an error can be cheaply corrected. Management sign-offs are required for certain orders — based on the complexity of the project, the dollar amount, or the experience level required — to catch problems before they reach the floor. Megan's framing: "It's still cheaper to find the problem before you ship it."

Bringing the Work Back In-House

The second half of the day turned to the outsourced order-processing arrangement. For a while it had worked acceptably — Megan estimated that around 95% of orders went through without issue. But over the last six months, quality had noticeably declined. The root cause turned out to be staff rotation: the more experienced people familiar with the Merrick account were cycled out and replaced by less experienced agents, with no notification. The result was a sudden increase in errors — "rookie-level mistakes," as Megan put it — with no explanation until Merrick started asking questions. She found the lack of transparency the most frustrating part: they were left to figure out what had changed on their own.

The less experienced staff also struggled with the complexity of the work. Orders in this industry carry a lot of specific detail — materials, component breakdowns, event dates, proof timelines — and getting any of it wrong has real consequences for the end customer. A name badge that arrives wrong or late is embarrassing for the distributor who sold it. On top of the quality and transparency issues, working-hours drift had crept in: by the time the outside team produced work or flagged issues each day, the in-house team was already gone, which forced in-house staff to start each morning catching up on what had been flagged overnight.

The combination of quality issues, transparency failures, and timing drift led to the decision to end the arrangement and absorb the work in-house. All of the functions return: order entry, art processing, and inbox triage. Megan's plan is to move a team member from the production floor into order entry — no new headcount needed. With production volume softer than it had been, there's capacity on the floor to support that shift. The transition is a scheduled handoff, not an emergency scramble.

Ordova API and Automation Potential

Nate asked whether anyone had looked at what the Ordova API actually exposes. Megan said she had a meeting scheduled with the Ordova account rep and planned to ask specifically about API capabilities, depth of connectivity, and documentation — she didn't yet know what was accessible. The question matters because several manual steps in the current workflow could potentially be automated if the API allows it. The ones named in this conversation:

Entering a new customer into Ordova — when a first-time distributor places an order, their company name, contact information, and shipping address all have to be entered by hand, even though that information is typically right there in the incoming PO or email. Nate noted that a tool to extract it automatically could pre-fill most of the new-customer form.
Entering a new contact or shipping address for an existing customer — same idea; the information exists in the document, and manual re-entry is the current step.
Freight charges — calculated outside Ordova and entered manually, as discussed in the freight-tool session.

Megan agreed to ask her Ordova rep specifically about the depth of connectivity — whether what's visible in the Ordova UI is also reachable through the API, and whether there are restrictions on which fields can be written programmatically.

What the API Assessment Found

Following the conversation, Nate assessed the Ordova API directly from its public Swagger specification. The headline: the API is real, RESTful, and reasonably full-featured — JSON in and out, token-based auth (no fee; a signed API agreement is the only gate), full CRUD on the core business objects, and a deep, well-modeled Orders tree. Most of the medium-tier opportunities on the map are unblocked.

Specifically:

Products and pricing — green. Full Products CRUD, with pricing fields present on the product object. The "enter a product once, let Ordova be the source of truth, publish outward" build is fully supported on the Ordova side. The product-integrity / catalog-monitoring idea and the pricing-update idea both read off the same endpoints — also viable.
Customer auto-create — green. Accounts and contacts can be created via the API, so new-customer intake → auto-create is supported.
Order automation — partial, the one asterisk. You can read orders, update an existing order, and fully build out an order's contents (items, charges, decorations, rows, payment). But the public spec has no endpoint to create a brand-new order header. Customer creation works; item-building works; order creation does not, at least in the public API.

On freight: the API won't return carrier rates — that's the carrier's side, not Ordova's — but once a freight cost is computed, it can be written onto an existing order as a charge. So a freight tool's output can flow back into Ordova. One note for any "real-time" framing: there are no webhooks documented, so integrations are poll-based (scheduled pulls), which is fine for batch and sync use cases.

The single open question to put back to Ordova: "We want to take an inbound PO or email and create the corresponding order in Ordova programmatically. The public API lets us update orders and add line items, but I don't see an endpoint to create a new order. Is order creation available through the full (post-agreement) API, a bulk import, or another mechanism — or are orders UI-entry only?" The answer decides whether the order-automation opportunity is a true end-to-end build or a "prep-and-accelerate the keying" build.

Session Notes as a Deliverable

Toward the end of the day, Nate noted that the summaries being created during the engagement could be shared with Megan directly — not just as internal working documents, but as something she could hand to others. Megan responded warmly: a detailed record of these conversations would be genuinely useful for bringing a new vendor or a new employee up to speed without having to repeat the same conversation from scratch. She raised the format question too — something she could turn into an HTML file or PDF would suit people who aren't working in AI tools, while markdown or plain text works for more technical contexts. Nate flagged it as something to design for.

Follow-Ups

Ordova API: confirm with the rep whether order creation is available through the full (post-agreement) API — the one remaining gate on the order-automation build.
Sample POs: gather three to five from different distributors to understand how much the format varies.
In-house handoff: confirm the production-floor team member moving into order entry, and document the inbox-triage categories that role will operate on.
Common errors: capture the most frequent "rookie mistakes" from the outsourced period to inform any validation or checklist work.
Power Automate: Megan flagged file routing as worth setting up — she just needs guidance.