/
.webp)
Company Biography | Long-Form Narrative Feature
Researched and written for Stacksync content library
The developer opened the retailer's vendor compliance guide on a Tuesday morning. It was a PDF. Forty-seven pages. The cover read: Trading Partner Integration Requirements — EDI X12 v5010.
He had a modern stack. REST APIs. JSON everywhere. A CI/CD pipeline he was proud of. A company that had just closed its first major retail deal — Walmart, one of those we-actually-made-it moments that the founders had toasted over warm champagne in a conference room two weeks prior.
Now he was reading a 47-page document describing a file format that used asterisks as field delimiters, tildes as segment terminators, and three-digit codes that mapped to a 300-page specification manual he would need to order separately.
The format was finalized in 1979.
He wasn't looking at a legacy integration problem. He was looking at a 76-year-old military logistics protocol that ran on a telex machine during the Berlin Airlift — and that Walmart, Target, Amazon, Kroger, Home Depot, and every major retailer in North America still requires today, without exception, from every supplier, as the price of admission to their shelves.
There was no alternative. There was no workaround. If you wanted to sell to Walmart, you spoke EDI. If you got it wrong, they charged you back 3% of the affected purchase order.
The developer's afternoon cleared. So did the rest of his week.
On June 24, 1948, the Soviet Union sealed the roads, rails, and waterways into West Berlin. The city of two million people — sitting 110 miles inside Soviet-controlled East Germany — had 36 days of food and 45 days of coal.
The Western Allies responded with what would become the largest humanitarian airlift in history: Operation VITTLES. For 462 days, U.S. and British aircraft flew continuous sorties into the city, ultimately delivering 2.3 million tons of food, coal, and supplies. At the operation's peak, a plane was landing at Tempelhof Airport every 90 seconds.
The limiting constraint wasn't pilots or aircraft. It was paperwork.
Every flight carried a cargo manifest. Every manifest arrived in a different format — some handwritten, some typed, some in English, some in German, some structured for air operations, some borrowed from naval supply protocols. Ground crews couldn't unload efficiently because they couldn't read what they were unloading. Cargo sat on tarmacs. The entire system moved slower than it had to.
U.S. Army Master Sergeant Edward A. Guilbert fixed it.
Guilbert developed a standardized manifest system that could be transmitted via telex, radio-teletype, or telephone — a universal format that any operator anywhere in the supply chain could read and act on without translation. A single structured document that said: this is what's on the plane, this is where it goes, this is the quantity, this is the classification. Every field. Every time. In the same place.
The Berlin Blockade ended in May 1949. The Allied resolve had held. West Berlin survived. And Edward Guilbert had invented something that would outlast the Cold War, outlast the Soviet Union, outlast the fax machine, outlast the internet boom, and still be the operational backbone of global retail commerce in 2025.
In the early 1960s, Guilbert left the Army and joined DuPont. He didn't leave the problem behind — he brought the solution with him.
At DuPont, he developed standardized electronic cargo messages for communicating with Chemical Leaman Tank Lines, one of the country's largest liquid bulk carriers. It was, by most accounts, the first commercial application of what would become EDI: two companies, different systems, standardized electronic data, no translation layer required.
The concept spread slowly through transportation. By the late 1960s, railroads, air carriers, ocean freight companies, and trucking firms were each developing their own electronic standards — which meant they were developing incompatible electronic standards. Every connection between two trading partners still required a custom translation. The problem Guilbert had solved for one airlift was replicating itself across an entire industry.
In 1968, the Transportation Data Coordinating Committee — the TDCC — formed to address it. Guilbert was its founding president, a position he would hold for 19 years. In 1975, the TDCC published the first cross-industry EDI standards.
In 1979, the American National Standards Institute chartered Accredited Standards Committee X12 — the body that would formalize EDI standards into the format that still governs most B2B commerce today. In 1988, the United Nations declared UN/EDIFACT the international standard for the non-North-American world.
Edward A. Guilbert is now recognized formally as the "Father of EDI." X12 presents the Edward A. Guilbert Lifetime Achievement Award annually. The document format named in his honor — the EDI transmission standard that undergirds $multi-trillion in global commerce — is unchanged in its structural logic from the manifest he built for a 1948 airlift over Cold War Germany.
Most people who haven't encountered EDI assume it is a protocol — something like HTTP, running under the hood of other applications. It is not. EDI is a flat-file format. A structured text document that uses fixed delimiters, parsed sequentially, optimized for data compactness in an era when bandwidth was measured in baud rates and transmission was billed by the character.
An EDI 850 — a Purchase Order, the transaction that initiates every sale between a supplier and a major retailer — looks like this:
ISA*00* *00* *ZZ*SENDER *ZZ*RECEIVER *210101*1200*^*00501*000000001*0*P*>~
GS*PO*SENDER*RECEIVER*20210101*1200*1*X*005010~
ST*850*0001~
BEG*00*SA*4500012345**20210101~
REF*DP*001~
PER*BD*BUYER NAME*TE*555-555-1234~
TD5**2*UPS~
N1*ST*SHIP TO NAME~
N3*123 MAIN ST~
N4*ANYTOWN*CA*90210~
PO1*1*10*EA*19.99**VP*VENDOR-SKU-001~
CTT*1~
SE*12*0001~
GE*1*1~
IEA*1*000000001~
This is a purchase order. Ten units of a single product, $19.99 each, shipping via UPS to California. The asterisks are field delimiters. The tildes terminate each segment. The three-digit codes — 850, 856, 810 — map to specific transaction types in the ANSI X12 specification. The two-letter codes within each segment — BEG, PER, N1, PO1 — map to individual data elements, each with their own position, length, and validation rules.
To read it without training, you cannot.
And every major retailer — Walmart, Target, Amazon, Kroger, Home Depot, Nordstrom, Neiman Marcus — has their own flavor of it. Their own required fields. Their own validation timing. Their own penalty structures for non-compliance. A supplier working with three major retailers is effectively maintaining three distinct EDI implementations simultaneously.
In the 1980s, as EDI spread from transportation into general retail, a new industry emerged to manage the logistics of it: the Value Added Network, or VAN.
A VAN is an intermediary service — a digital post office. Suppliers send EDI documents to their VAN's "mailbox." The VAN routes them to the retailer's mailbox. The retailer polls their mailbox on a schedule and picks up the documents. No direct connection between supplier and retailer required.
The VAN model made EDI accessible to companies that couldn't afford to build direct connections with every trading partner. It was genuinely useful infrastructure for its era — the 1980s, when private networks were the only alternative to direct dial-up connections and the public internet didn't exist.
VAN pricing reflected the economics of private network infrastructure:
For a small supplier working with five major retailers, annual EDI infrastructure costs reach $5,000–$15,000 before a single developer hour is counted. The major VAN and managed-service providers — SPS Commerce (NASDAQ: SPSC, $730M in annual revenue), TrueCommerce, DiCentral — have maintained this pricing structure largely intact through four decades of technological change.
The internet arrived. Broadband arrived. Cloud computing arrived. REST APIs arrived. The VAN billing model didn't change. Because it didn't have to. Retailers still required EDI. Suppliers still had to comply. The VAN sat in the middle and charged per document, per kilocharacter, per year, indefinitely.
That is not a legacy oversight. That is a business model.
Walmart mandated EDI compliance for all suppliers in 1988. The World Wide Web had not yet been invented. Commercial email was rare. The dominant personal computer was the IBM PC/AT, running DOS 3.3.
Walmart didn't care. The mandate was absolute: if you want your products on our shelves, you send us EDI. If you don't, you don't ship to us.
The required document set for Walmart:
- EDI 850 — Purchase Order (Walmart initiates)
- EDI 855 — Purchase Order Acknowledgment (supplier must respond within 24 hours)
- EDI 856 — Advance Ship Notice (must arrive before the physical shipment does)
- EDI 810 — Invoice (within 24 hours of shipping)
Non-compliance penalties remain active today:
- OTIF (On-Time In-Full) failures: 3% of affected purchase order value
- ASN errors: $50–$500 per incident
- Labeling violations: $25–$200 per carton
A small supplier with a $2 million annual Walmart revenue line can lose $20,000–$60,000 per year in compliance penalties if their EDI is slow, incorrect, or missing.
Target has the same system. Amazon does too. Each with their own field specs. Each with their own timing requirements. Each with chargebacks waiting if any element is wrong.
This is not a technical detail. This is the economic structure of U.S. retail. And it has been unchanged for 36 years.
Erik Kiser didn't arrive at Orderful from outside EDI. He grew up inside it — spending years building custom EDI integrations for vendors and managing EDI connections as a consultant. He had seen the full landscape from the inside: the mapping complexity, the VAN fees, the weeks-long onboarding processes, the moment when a trading partner changed their spec and two weeks of EDI work became obsolete overnight.
What Kiser understood that outside observers missed: EDI wasn't surviving because it was good technology. It was surviving because a complete economic ecosystem had grown up around its complexity.
SPS Commerce didn't have $730M in annual revenue despite EDI being complicated. It had $730M in annual revenue because EDI was complicated. The managed service model requires complexity to justify its existence. The per-transaction VAN fees require volume to generate margin. The proprietary mapping tools require lock-in to prevent churn. Every time EDI got harder to implement, it was good for the incumbents.
Kiser co-founded Orderful in 2016 in San Francisco with a specific technical thesis: EDI could be abstracted behind a modern API. One connection in, data out in any format, with all translation and routing handled by the platform. Not managed services — self-service. Not per-transaction fees — a SaaS subscription. Not weeks of onboarding — days.
The business model inversion was intentional. Orderful charges a flat subscription. Their incentive is to make integrations work fast and stay working. The legacy model's incentive was the opposite.
Initialized Capital funded the seed round in 2019. Andreessen Horowitz led a $10M Series A the same year. David Ulevitch at a16z framed it plainly at the time: "Orderful has done a remarkable job of simplifying what is typically a messy and broken web of supply chain interactions."
Every startup attacking infrastructure has a version of this problem. Orderful had a particularly severe one.
To change EDI, you need suppliers to adopt a new system. But suppliers adopt EDI to connect to retailers. The retailers don't change their EDI requirements for any supplier — they have thousands of suppliers, and none of them are in a position to negotiate format standards. The format standard is the retailer's requirement, and the retailer's requirement is the format.
So Orderful's pitch to a supplier was: "Adopt our API instead of building native EDI." The retailer would still receive EDI — Orderful would translate the API call into the correct EDI format and deliver it via the retailer's existing VAN connection. The supplier never sees the raw EDI. The retailer never knows the difference.
This is technically elegant. It is also selling a modernization project to people who are currently compliant with a system that works, at the cost of implementing something new.
The barrier isn't technical skepticism. The barrier is organizational inertia plus career risk. The person inside a large enterprise who has managed the SPS Commerce relationship for eight years doesn't want to own the migration project. If the migration works, it was routine. If it fails — even briefly — their career absorbed the downside. The rational move is to keep the system that works.
For small suppliers, the math is different. They're paying $8,000 a year to SPS Commerce for a managed service they don't fully understand, and their EDI specialist left six months ago. Orderful's self-service API model is genuinely better for them. But the sales cycle for a $20,000 ACV SaaS platform targeting small suppliers is brutal.
The grind isn't technical. The grind is human.
SPS Commerce has 120,000+ companies on its network with over 1 million active connections. That network is the product. If you're a retailer and you require EDI from suppliers, the easiest path is to require SPS Commerce — because 120,000 suppliers already have accounts there.
When a retailer onboards a new supplier, SPS Commerce can send them a link and say: "Here's your SPS account. You'll be live in a few weeks." The integration path is already built. The maps already exist. The testing protocols are established.
When a supplier wants to switch from SPS to Orderful, they face: re-mapping every trading partner relationship, re-certifying with every retailer's EDI team, migrating transactional history, retraining their operations staff. The switching cost is high not because Orderful is difficult to adopt, but because EDI itself doesn't have a portability standard. Your maps don't transfer. Your connections don't transfer. Your certification history doesn't transfer.
You don't own your EDI integrations under the managed service model. The managed service provider does.
In March 2020, the supply chain became a national news story. Not because factories shut down — most didn't, immediately — but because visibility disappeared.
Toilet paper was available. It was sitting in distribution centers. The problem was that inventory systems didn't know where it was, retailers couldn't re-route it, and the EDI batch cycle meant that by the time anyone had accurate data on where things were, the data was already wrong.
Traditional VAN-based EDI runs on a polling architecture. Suppliers push documents to the VAN mailbox. The retailer's system polls the mailbox on a schedule — every 15 minutes, every 30 minutes, sometimes hourly depending on the retailer and their infrastructure configuration. In normal operating conditions, a 30-minute data lag is imperceptible. In crisis conditions, a 30-minute inventory update cycle means every routing decision is made on stale data.
Orderful's architecture is event-driven. Documents trigger delivery in real time — not on a polling cycle. An Advance Ship Notice fires and arrives at its destination immediately, not at the next VAN poll. A shipment status update propagates instantly.
The difference between batch and real-time EDI is invisible during normal operations. During a supply chain crisis, it is the difference between visibility and blindness.
COVID accelerated supply chain digitalization by an estimated three to four years, according to McKinsey research from 2021. 64% of supply chain executives reported that digital transformation would accelerate as a direct result of the crisis. Boards that had been approving incremental EDI improvements were suddenly funding wholesale infrastructure overhauls.
That created Orderful's market moment.
Companies that had been compliant with legacy EDI — technically fine, operationally mediocre — started evaluating whether their infrastructure was fit for the next disruption. The answer, often, was no. And the first place that answer showed up was in their EDI visibility: slow, batch-based, opaque.
By the time Orderful raised its Series B — $19M led by GLP Capital Partners, with continued participation from a16z and Initialized Capital — the network had grown to include major shippers (including Walmart as a shipper within the network), logistics providers like XPO Logistics and Kuehne+Nagel, and carriers including J.B. Hunt.
The network had reached 10,000+ trading partners and was processing 50M+ transactions with 99.99% uptime.
Liquid Death — the canned water brand that turned beverage into cultural performance — needed to move fast. As a high-growth consumer brand entering retail channels, their EDI problem was standard issue: retailers required it, setup was slow, every new channel meant weeks of onboarding and ongoing VAN fees that scaled with transaction volume as they grew.
With Orderful, their EDI setup time dropped by 80%. Chargebacks — the compliance penalties that erode margin on every order that misses a field or misses a timing requirement — dropped by 90%.
ShipBob, the third-party logistics platform used by thousands of e-commerce brands, had a different version of the problem: their clients' retail channel orders all came through EDI, but the engineering effort to maintain EDI compliance for a growing client base was a distraction from building the actual fulfillment product. They used Orderful to standardize EDI across all their retailer connections and cut the engineering effort dramatically.
Society6 cut EDI onboarding time by 75%. 8th Avenue Food & Provisions cut EDI costs and resource spend by 50%. NFI, one of the largest private transportation and logistics companies in North America, went from weeks-long carrier onboarding to under 5 days.
These aren't A/B test improvements. These are order-of-magnitude changes in the economics of a compliance requirement that had been accepted as fixed cost for decades.
In 2025, a typical mid-size B2B company runs: Salesforce or HubSpot for customer management. NetSuite or Dynamics for ERP. Shopify for commerce. Snowflake or Redshift for analytics. These systems connect via REST APIs, webhooks, OAuth, and real-time event streams. A developer can integrate two modern SaaS platforms in an afternoon. Documentation exists. Sandbox environments exist. SDKs exist.
Then there is the retailer connection. Which requires EDI. In the format finalized in 1979.
EDI is the only place in the modern B2B technology stack where a file format from the Berlin Airlift era is still the wire protocol. It exists alongside everything else as a specific, persistent carve-out — not because nobody noticed, but because the network effect protecting it is absolute. Every Walmart supplier must speak EDI. Every Target supplier must speak EDI. Every supplier who has built compliance into their operations has absorbed the cost and moved on.
The gap Orderful is attacking isn't just the translation layer between JSON and X12. It is the entire model by which companies manage trading partner relationships — the managed service model that locks in complexity, the per-transaction VAN fees that scale against you as you grow, the batch polling architecture that gives you 30-minute-old data, the weeks-long onboarding process that slows every new retail partnership.
The modern version of EDI infrastructure — Orderful's version — is a network. Not a managed service. Not a VAN mailbox. A network where connecting once gives you access to 10,000+ trading partners, where onboarding a new partner is measured in days not weeks, where documents travel in real time and not on a polling cycle.
In 2024, Orderful released Mosaic — an AI-native EDI product that eliminates the most expensive part of traditional EDI: the mapping layer.
Every EDI implementation requires a map: a document that translates your internal data fields into the specific X12 field positions and validation rules required by each trading partner. Maps are bespoke per retailer. Maps change when retailers update their specs. Maps break when your internal systems change. Maintaining maps is the primary reason EDI requires specialists and the primary reason onboarding is slow.
Mosaic automates map creation and maintenance using AI. The first AI-native EDI product that eliminates mapping — not as a demo concept but as a production deployment.
This is where the EDI modernization story is heading: AI absorbs the translation and compliance burden that EDI specialists have carried since the 1980s. The knowledge that lived in people's heads — which field position carries the ship-to ZIP code, which qualifier code means "retailer-assigned SKU," which segment sequence the specific retailer's validation engine requires — becomes a trained model instead of a staffing dependency.
The format that Edward Guilbert invented to track cargo over the Berlin Airlift isn't going away. But the era of humans maintaining it manually might be.
Orderful sits at the EDI layer of the supply chain. It handles the translation, routing, compliance, and partner connectivity of supply chain documents. What it doesn't do — by design, by scope — is bridge the EDI data into the rest of the business.
A company using Orderful processes EDI transactions in real time: purchase orders, advance ship notices, invoices, shipment confirmations. That data exists inside the EDI platform. But the account manager in Salesforce still doesn't see the purchase order when it fires. The revenue forecast in the CRM doesn't update when the invoice is accepted. The operations dashboard in the database doesn't show the ASN. The EDI layer and the business software layer sit parallel, not connected.
This is where platforms like Stacksync complete the picture: real-time sync from the operational data layer — where EDI events live — into the CRM and analytical layers where business decisions are made. When a Walmart purchase order arrives via Orderful, the account team sees it in HubSpot immediately. When a carrier updates shipment status, the operations database reflects it in real time. When a compliance penalty fires, the revenue team knows before the invoice arrives.
EDI modernization made the supply chain data available in real time. Making it visible across the whole business is the next layer of the same problem.
1. EDI was invented by one man to solve a Cold War crisis — and he transferred it to the private sector himself.
Edward A. Guilbert, U.S. Army Master Sergeant, developed the standardized cargo manifest that would become EDI during the Berlin Airlift in 1948. He didn't just invent a format — he then left the Army, joined DuPont in the 1960s, created the first commercial EDI application with a tank lines company, helped found the Transportation Data Coordinating Committee in 1968, and served as its president for 19 years. One person invented EDI, commercialized EDI, and spent two decades governing EDI standards. Every purchase order sent to Walmart today traces through that one sergeant's solution to a 1948 logistics crisis.
2. The standard took 31 years to stabilize — longer than the commercial internet has existed.
Guilbert developed the core concept in 1948. TDCC published the first formal standards in 1975. ANSI chartered X12 in 1979. The UN declared EDIFACT in 1988. The "standard" that people treat as monolithic is actually a 40-year accumulation of overlapping standards bodies, each adding their own layer. The format that the 2025 developer is parsing on a Tuesday morning is the fossilized sediment of four decades of committee decisions.
3. Walmart mandated EDI before the internet existed — and has never modified the requirement.
The World Wide Web was invented in 1989. Walmart's EDI mandate dates to 1988. Every supplier compliance guide, chargeback structure, and trading partner requirement that Walmart uses today descends from a decision made before a single webpage existed. The mandate that governs how a DTC brand built in 2022 submits its purchase order acknowledgments was designed before its founders were born.
4. The people keeping EDI alive have more incentive to maintain it than to replace it.
EDI specialists — professionals who understand ANSI X12 mapping, VAN routing, and trading partner certification — are a genuine profession with job boards, staffing agencies, and annual conferences. The managed service providers built their $730M revenue business on the irreplaceability of that expertise. A vendor compliance manager who has spent eight years managing SPS Commerce doesn't own a migration success case if modernization works. They own a migration failure case if it doesn't. The rational choice, individually, is to defend the complexity that justifies their expertise. No individual made a bad decision. The system selects for its own perpetuation.
5. Orderful's network includes Walmart as a participant — the same Walmart that created the mandate Orderful is designed to work around.
Walmart is the institution that imposed the EDI tax on suppliers in 1988 and still enforces it with 3% OTIF chargebacks today. Walmart is also a shipper on Orderful's network, using Orderful's infrastructure for its own logistics operations. The company that created the supply chain compliance regime that Orderful's customers are trying to escape is itself a customer of Orderful's network. This is how infrastructure transitions actually work: the incumbent adopts the new layer not to replace itself but to gain efficiency — while the new system gradually absorbs the complexity beneath.
| Fact | Data |
|---|---|
| EDI origin | 1948, Berlin Airlift, Edward A. Guilbert, U.S. Army |
| First commercial EDI | Early 1960s, DuPont + Chemical Leaman Tank Lines |
| TDCC formed | 1968 |
| ANSI X12 chartered | 1979 |
| UN/EDIFACT standard | 1988 |
| Walmart EDI mandate | 1988 (pre-internet) |
| Orderful founded | 2016, San Francisco |
| Founder/CEO | Erik Kiser |
| Seed investor | Initialized Capital |
| Series A | $10M, Andreessen Horowitz, 2019 |
| Series B | $19M, GLP Capital Partners + a16z, 2022 |
| Total raised | ~$43.9M |
| Network scale | 10,000+ trading partners, 50M+ transactions |
| Performance metrics | 90% chargeback reduction (Liquid Death); 80% faster setup; 50% cost reduction (8th Avenue); 75% faster onboarding (Society6) |
| SPS Commerce (incumbent) | $730M annual revenue, 120,000+ companies, NASDAQ: SPSC |
| Onboarding comparison | Orderful: under 9 days avg vs. weeks for SPS |
| Walmart OTIF penalty | 3% of affected PO value |
| EDI market size | $2B+ (2024), projected $5B+ by 2032 |
| Latest product | Mosaic — AI-native EDI, eliminates mapping layer |
| Key customers | Liquid Death, ShipBob, Society6, 8th Avenue Food, NFI, Oura Ring, Caraway |
| Walmart in network | Present as shipper — the mandate creator is also a user |
Research sources: Orderful.com (product, blog, customer case studies, API documentation), a16z announcement (TechCrunch coverage), GLP Capital Partners Series B, Orderful homepage metrics (10,000+ partners, 50M+ transactions), ANSI X12 committee history, TDCC history, Berlin Airlift historical record, Walmart vendor compliance documentation, SPS Commerce investor relations (SPSC NASDAQ), McKinsey COVID supply chain acceleration research.
