In 2023, Washington State Employees Credit Union had a new brand and an old problem. The artwork was ready to change. The machinery behind its payment cards was less accommodating. Design restrictions, stalled programming and late deliveries turned a rebrand into an operations headache. Some cards reached branches and contact centers up to two weeks late, according to Arroweye’s account of the project.
- Arroweye makes, personalizes and ships payment cards on demand.
- Customers avoid buying and managing preprinted card inventory.
- The buyers are banks, credit unions, fintechs and program managers.
- CPI Card Group bought the business in May 2025.
WSECU found Arroweye through its processing partner, MAP. It moved its EMV program, then its contactless and dual-interface programs. The published case study reports eight designs and more than $100,000 saved in seven months. What changed was the point at which a design became a physical commitment.
01 / The expensive art of guessing
A conventional card program can begin with a prediction. How many cards will customers want? Which design will they choose? An issuer orders branded stock, holds it, then has individual cardholder information added as needed. The forecast gets a physical form. When a logo changes or demand disappoints, that form becomes awkwardly tangible.
Arroweye moves manufacturing closer to the order. Its digital process brings production, personalization and fulfillment together, with the card, carrier and inserts created for the requested program and recipient. The customer pays for cards as needed. There is less reason to purchase a warehouse’s worth of confidence in a forecast.
That distinction deserves precision. “Zero inventory” describes the customer’s preprinted stock. A factory still needs materials, chips, equipment and capacity. Arroweye removes a particular obligation from the issuer; it does not abolish the supply chain.
02 / A factory with a personal touch
The business reaches back to a less industrial irritation. An early company biography says co-founder Doron Friedman received chocolates without a personalized note. He saw an opportunity to make mailed gifts feel personal. Fellow co-founder Brian Huse helped develop the production systems. The company’s current history dates its digital on-demand model to 2000.
The capabilities accumulated. Visa and Mastercard certification followed in 2008, Discover in 2010, American Express in 2012. Customer Connect arrived in 2013, EMV On-Demand in 2014 and contactless dual-interface production in 2019. A company built around customization had to learn the exacting habits of payment-card manufacturing.
Today the output includes credit, debit and prepaid cards, with chip and contactless options. Material choices, finishes, orientation and the surrounding paper carrier give issuers room to express a brand. Arroweye also emphasizes printing personalized information beneath the laminate, protecting details within the card’s construction. The little object in a wallet carries a surprisingly elaborate production brief.
03 / Software still needs an envelope
Arroweye sits between the organizations designing financial products and the people receiving their cards. It sells production services to institutions and program operators. A consumer usually encounters the result without ever encountering Arroweye’s name. The factory is doing its job when another company’s brand gets the attention.
That explains its platform partnerships. In 2022, Synctera selected Arroweye for customized cards supporting its t-minus10 developer experience. Developers could build banking products in software, but those products still needed a physical expression. Synctera co-founder Peter Hazlehurst put the requirement plainly:
“Even in a digital-first era, FinTechs need physical cards to appeal to a diverse customer base.”
Peter Hazlehurst / Synctera / 2022
Different customers stretch the same capability. Affiniti Finance uses Arroweye for association-branded credit cards across nine industries. Each association can need its own design and carrier. In October 2024, Velera selected Arroweye for production, personalization and fulfillment of dual-interface debit and credit cards. Velera’s network reaches more than 4,000 financial institutions; those are potential distribution relationships, not a direct Arroweye customer count.
There is a more urgent version, too. Arroweye’s Qolo case study describes a state-government disbursement project in late 2022. It reports nearly 740,000 cards fulfilled within eight weeks. Here, manufacturing flexibility served a deadline attached to public assistance, rather than a marketing calendar.
04 / The $45.55 million vote
On May 6, 2025, CPI Card Group acquired Arroweye. The next day’s announcement put the price at $45.55 million, before customary adjustments. CPI was buying on-demand capabilities that complemented its existing payment-card business. For Arroweye, the combination offered access to a larger company’s customer relationships and resources.
CPI Card Group’s purchase of Arroweye
The subsequent numbers show an operating business inside a larger one. CPI’s annual filing included $42.8 million of Arroweye revenue after the acquisition through December 2025. Its first-quarter 2026 filing reported another $16.1 million contribution. These are different reporting periods, not evidence of a sudden growth rate.
Integration has a bill. CPI reported $2.8 million of second-quarter 2026 integration costs primarily related to Arroweye. In August, it also said the acquired business was performing ahead of its original investment case. Both facts belong in the picture.
05 / Copy the sequence
For an issuer considering Arroweye, the useful comparison is total program cost: production, setup, stored stock, obsolete designs, change requests and delivery. WSECU’s savings illustrate one outcome. They are not a price list. A stable program with predictable demand may value a different mix of cost and flexibility.
Other suppliers compete here. IDEMIA provides card issuance and personalization; Giesecke+Devrient markets print-on-demand services. A sensible procurement exercise asks each provider exactly what gets made after an order, what must be committed beforehand, and which changes trigger a delay. A virtual-only product, meanwhile, has little need for a physical-card factory.
An issuer can turn that comparison into a small trial. Choose a live design, define the data handoff, agree on proof approval and measure elapsed time through delivery. Then request a change. The revealing number is how much work, stock and waiting that alteration creates. This is a suggested buying test, rather than a result Arroweye has promised for every program.
Timing also has boundaries. Fast manufacturing cannot settle an issuer’s unfinished artwork, resolve its internal approvals or make a postal journey disappear. A team that needs customers to walk out of a branch holding a card should evaluate instant issuance separately. That is a different delivery requirement from centralized manufacturing and mailing, even when both services appear in the same parent company’s portfolio.
The broader lesson is an operational inference: postpone an irreversible choice until you have better information. You can copy that question without copying Arroweye’s machinery. Before ordering the next batch, ask which assumptions you are about to laminate.