Built in-house, it captured each request correctly and handed it to production one at a time. It had no notion of a production order, so a customer's many orders in a day were many separate things, each tracked, or not, by whoever touched it.
A thousand orders that were all different
A maker of personalised ID badges and lanyards, where no two items are alike, added the production-order layer its own ordering system never had: orders from the same customer are grouped and tracked from art to delivery, each designer has a dashboard, and a structured collection process cut delinquency by more than 90%, by the company's account.
A maker of personalised ID badges and lanyards, where no two items are alike.
Exxpress makes PVC badges, lanyards, proximity cards, holders and accessories, with no minimum quantity. Each badge is unique, personalised one at a time rather than in batches, and a single customer may order many in a day for different people and different sites.
Personalised identification materials
2014Year the company started, in a bedroom with a portable phone
16,000+Companies served, from the company site
No minimumQuantity on a badge order
Own deliveryIn its metropolitan area, with shipping beyond it
>90%Reduction in delinquency, in large part due to the structured collection process
12–15 → 24–28Layout and file-closing emails per day the creation routine sent before, and sends now
16,000+Companies served
2014Year the company started, in a bedroom with a portable phone
“We didn't have the concept of a production order. We built our own online badge-request tool, and when we built it we only worked with the concept of an order: an order comes in, we make it. That made sense at the scale we ran in 2018. But when the same customer places many orders in a day, with different art, for different destinations, and more the next day, our system had no logic to group those orders and coordinate their progress inside the company. We saw a very large gap, and it was producing financial losses from errors, from things going wrong in the middle of the process.”
ExxpressCompany leader, in the recorded interview
BeforeAfter
- Orders handled one at a time, with no groupingA customer's orders grouped into a production order, tracked across the company
- Progress known by askingStatus per item on the production order; the ordering system's events feed it
- Layout requests as a full formA sales app for layout requests showing only the relevant fields for that role
- Designers finding out what was nextA dashboard per designer with the day's work and the status of each item, and a side chat
- Art sent at 12 to 15 emails a day per routine24 to 28 a day, by the team's count, with the same routine
- Overdue receivables handled case by case with no viewA collection process with open amounts, treatment per case and metrics
- One learning curve for everyoneEach role sees an app shaped to its own work
What was not replaced. The in-house ordering system stayed as the customer front door and as the source of events. Email stayed as the art-approval channel. The creation and production teams are the same; they now see the work grouped and assigned.
The bottleneck was operational, not commercial
Exxpress makes personal identification materials for companies: PVC badges, personalised lanyards, proximity cards, holders and accessories, with no minimum quantity and its own delivery in its metropolitan area. It started in November 2014 and reports more than 16,000 companies served. Its distinguishing feature, in its own words, is that it makes nothing for the shelf: each badge is unique, personalised one at a time rather than in batches, and a single customer may order many in a day for different people and different sites.
A made-to-order operation lives or dies on tracking. A batch producer can count units; a badge maker has to know, for each item, whose it is, which art it carries, which order it belongs to, where it is in the process and where it ships. When one customer places thirty orders in a day and more the next, "an order" stops being a useful unit and "everything this customer has in flight" becomes the thing that has to be managed.
That is why the bottleneck was operational and not commercial. Orders were coming in through the company's own system, which worked well at the scale of 2018 and was built around the idea that an order arrives and gets made. What was missing was the layer above it: grouping orders into production, coordinating their progress across the company, and knowing where things went wrong before the loss landed.
A good in-house system for the front of the process and nothing for the middle and the end
“We didn't have the concept of a production order. We built our own online badge-request tool, and when we built it we only worked with the concept of an order: an order comes in, we make it. That made sense at the scale we ran in 2018. But when the same customer places many orders in a day, with different art, for different destinations, and more the next day, our system had no logic to group those orders and coordinate their progress inside the company. We saw a very large gap, and it was producing financial losses from errors, from things going wrong in the middle of the process.”
With no systematic tracking of progress, things went wrong mid-process and the company found out when the cost arrived. The company's own words are financial losses from errors.
The company could have extended its own system. Doing so was not fast enough for the time it had, which is why it looked for another way to build the missing layer.
Designers closed files and sent art for approval at a pace set by finding out what was next: 12 to 15 layout emails a day per routine, with interruptions to ask where things stood.
Overdue invoices were a large problem, with no process defining the treatment of each case and no metrics on what was open.
The pattern: a good in-house system for the front of the process and nothing for the middle and the end. Nothing was broken; the company was growing past the scale its tool was built for. Every day of growth added orders that were captured well and coordinated by hand.
Item by item, what did each job before and what does it today
| What did this before | What does it today |
|---|---|
| Orders handled one at a time, with no grouping | A customer's orders grouped into a production order, tracked across the company |
| Progress known by asking | Status per item on the production order; the ordering system's events feed it |
| Errors discovered as financial losses | Process control with dashboards, so problems are visible mid-process |
| Layout requests as a full form | A sales app for layout requests showing only the relevant fields for that role |
| Designers finding out what was next | A dashboard per designer with the day's work and the status of each item, and a side chat |
| Art sent at 12 to 15 emails a day per routine | 24 to 28 a day, by the team's count, with the same routine |
| Overdue receivables handled case by case with no view | A collection process with open amounts, treatment per case and metrics |
| One learning curve for everyone | Each role sees an app shaped to its own work, so learning it matches the job |
What was not replaced. The in-house ordering system stayed as the customer front door and as the source of events. Email stayed as the art-approval channel. The creation and production teams are the same; they now see the work grouped and assigned.
The chain, end to end, for one customer's day of orders
- 1The customer places orders in the company's online badge system, each with its own art and destinationExternal
The front door stayed.
- 2The orders arrive as events and are grouped into a production order for that customerAutomatic
The layer the ordering system never had.
- 3Sales requests any layout work through its app, filling only the fields that role needsPerson
Not a full form of fields the requester did not need.
- 4The designer sees the item on their dashboard, closes the file and sends the art by email for approvalPerson
The dashboard replaces the question of what is next.
- 5The customer approves the artExternal
Email stays the channel.
- 6Production personalises each item; status moves per item on the production orderPerson
An operator has to personalise the item.
- 7Items ship to their destinationsExternal
Different art, different sites, still one production order.
- 8Invoices are issued; unpaid ones enter the collection process, where each case's treatment is recorded and open amounts are trackedPerson, with automatic tracking
Someone has to make the collection call; the tracking around it is automatic.
The hard links are 4, 6 and 8. A designer has to close the file, an operator has to personalise the item, and someone has to make the collection call. The design does not remove those; it removes the finding-out around them. The designer dashboard replaces the question "what's next"; the production order replaces the question "where is this customer's stuff"; the collection process replaces the question "who has been chased and who hasn't". The team's own summary is that the value is not only automation but control of the process and the transparency the dashboards give.
“When I started using the system I only did file closing. Back then the routine of closing a file and sending the art to a customer was around 12 to 15 a day. Today the creation team has doubled that, around 24 to 28, because each designer has their own dashboard with everything they need, what to do today and where it stands. Information moves through the side chat and nobody has to stop anyone.”
What changed, and what the source does not measure
| Indicator | Result | Where it comes from |
|---|---|---|
| Delinquency | Reduced by more than 90%, attributed in large part to the structured collection process and its metrics | Recorded interview; self-reported, no baseline, period or amount |
| Creation throughput | From 12 to 15 to 24 to 28 layout emails a day per routine | Recorded interview; the team's own count, undated |
| Production coordination | From orders tracked one by one to production orders tracked across the company | Recorded interview |
| Error losses | Described as the gap the build was meant to close; no figure before or after | Recorded interview |
| Learning curve | Each role learns an app shaped to its job rather than the whole system | Recorded interview |
The delinquency figure is the company's, with no numbers behind it. More than 90% is stated as a reduction with no starting rate, no period and no amount, and the company itself attributes it "in large part" to the collection process, which leaves room for other causes. This write-up reports it as stated and does not convert it into money.
The throughput figure is a routine count, not a measurement. The team gives a before and an after for one routine, from memory and undated. It is consistent and specific, and it is not a timed study.
"Doubled" is the team's word for 12 to 15 becoming 24 to 28. The arithmetic holds at the low end and slightly exceeds it at the high end; the write-up keeps the ranges rather than the summary.
No company-scale figure from the interview exists. Orders per day, badges per month, revenue and team size are not in the source. The 16,000 companies figure is the company site, undated.
What this case does not measure
Worth naming, because it is usually what gets inflated.
Both quotes come from a recorded interview whose transcript names neither speaker: one speaks for the company direction, the other for the creation team. No process was independently observed.
Delinquency rate before, loss from errors before, orders per day: none is given.
It is proprietary, event-driven and built around orders; nothing else about it is claimed.
Several passages were reconstructed from context. The "scale of 2018" reference dates the ordering system, not the interview.
He describes researching a way to build the missing layer and doing so on the platform. Jestor now builds and maintains the system on the customer's behalf; those passages are not used as evidence. His remark about vertical ERPs never fitting any operation is a general observation and names no product.
It appears on the company site as a running total.
Late payment is the normal state of B2B trade, and write-off is the tail of it
| Indicator | Number | Source |
|---|---|---|
| Share of B2B credit sales paid late, one Northern European market, 2025 | 41% | Atradius, Payment Practices Barometer, Western Europe, 2025 |
| Share of B2B invoices written off as uncollectable, same market | 7% | Same survey |
| Share of B2B credit sales paid late, one Gulf market, 2025 | 58%, with bad debt at 8% of overdue invoices | Atradius, Payment Practices Barometer, 2025 |
In well-run markets, about four in ten credit invoices are paid late and several percent are never paid. A supplier of low-ticket, high-frequency items to thousands of business customers is exposed exactly where the survey numbers sit: many small invoices, each too small to chase by hand and too many to ignore.
The difference between an overdue invoice and a written-off one is usually whether anyone followed up, and whether the follow-up was recorded so the next person could continue it. The company's own explanation of its result is that the treatment of each case became visible and measurable; that is the mechanism the survey figures point to.
The Atradius numbers describe surveyed firms in two markets. They give the scale of the problem; they do not give this company's rate before or after. Its own open-amount metric, as it now measures, is the number to plan on.
The system is built to fit, and responsibility for it stays with us
Exxpress had already built its own software once, and built it well; the ordering system still runs. What it learned is that the layer above it, production orders, dashboards, collection, needed to exist faster than an in-house build could deliver. The question a bespoke system has to answer is what happens in year two: a production order nobody can extend when a new product line arrives, or a collection process nobody can adjust when terms change, becomes the next gap in three years, with a login.
One builder owns each request from start to finish, and one is always in execution.
A new role app, a new dashboard or a new automation comes in through the same channel, without becoming a new project.
If what was built is not right, it is rebuilt. Unlimited revisions within the subscription.
Seats are never the billing unit, which matters when sales, every designer, production and finance all touch the same production order.
People learn to use their app the way they learn any app, by opening it. Building, configuring and maintaining stays on our side.
Full export at any time, by CSV and API. SOC 2 compliant, no exit fee. Pause in one click and the systems keep running.
Methodology and sources
The before-and-after descriptions and both quotes come from a recorded interview with a company leader and a member of the creation team, used through its machine-generated transcript, which names neither. Quotes were cleaned of transcription noise without changing their content. The account is the company's own and was not audited. Passages describing the company leader building the system himself were not used as evidence.
Founding date, product lines, no-minimum policy, delivery area and companies served from the company's own site and its founder's account there.
Atradius, Payment Practices Barometer, 2025 editions for Western Europe (Nordics report) and the United Arab Emirates.
No delinquency rate, amount recovered, loss figure or order volume is claimed, because none was given. More than 90% is the company's own reduction figure, with no baseline or period. The 12 to 15 and 24 to 28 counts are the team's memory of one routine, undated. The speakers' names are not invented.