Skip to main content

Case Study

Preparing for Scale (Part I)

Mula is a B2B SaaS platform that helps companies streamline their merchandise operations. The mission was to prepare the company for European expansion.

Role

Product Design Lead

Duration

2022–2023

Teams

Design, CS, Sales, Supply Chain, Warehouse

Context

Mula had been growing steadily without funding through a period of high demand for branded merchandise. By mid-2022 a competitor had acquired a 25% stake with an option to complete the purchase, valuing the company at €40 million. The European expansion had a timeline. Thirty people and a platform that had grown faster than the processes supporting it.

Two fronts needed work before scale was viable: operations, and the customer experience. Operations first. Expansion without a stable operational foundation would have multiplied the failures customers were already seeing (late deliveries, missed deadlines, broken handoffs) across every new market we entered. This case is the story of that first phase.

No Single Source of Truth

Stakeholder interviews in my first weeks revealed that no two teams described the order lifecycle the same way. Inconsistent answers about the same process mean one thing: no single source of truth. I mapped the full order lifecycle across all five teams involved.

Service map of the Mula order lifecycle
Service map — order lifecycle across the five teams involved

What came back wasn't fixable at the surface. Critical processes were running off-platform. Spreadsheets and Slack threads were compensating for gaps. This led to broken handoffs, unclear ownership, and operational failures that customers felt directly: delayed ETAs, missed deadlines, and, eventually, churn. At the current volume, the workarounds were manageable. At the next, they would compound.

01

Off-platform processes

Critical approvals and order tracking ran through Slack, email, and spreadsheets. No visibility, no single source of truth.

02

Unclear ownership

More than one person was considered responsible for the same step. In practice, nobody owned it.

03

Wasteful manual processes

Production specifications created manually. Order status maintained by hand across five teams. Every change meant rework.

Ownership was clarified during the mapping itself. Walking the broken steps with each team surfaced where responsibilities were unclear, and a single owner was assigned per process.

Standardizing the Design Phase

Production specifications were the manual process that touched every order. Different suppliers handled different product types and required different specifications. Designers maintained the templates by hand in Photoshop, rebuilding spec sheets per supplier per order. It worked at the current volume. It wouldn't at the next.

The decision was to stop accommodating each supplier's preferences and design one scalable format, agreed with the producers we worked with.

The intervention was closing the platform's configuration gaps so the specifications sheet could generate from platform data instead of being assembled in Photoshop.

Same team. 43% more products per designer. Average design phase down from 9.57 to 6.35 days per order.

No One Could See the Full Order Lifecycle

Order monitoring had three structural gaps: no consolidated view of the lifecycle, no early warning when a milestone was running late, and no preserved baseline for performance.

01

No traceability

When something went wrong, CS and Sales reconstructed history from Design briefs, HubSpot, Slack threads, and spreadsheet comments. No single record existed.

02

No early warning

The platform showed green until the ETA had already passed. No team could tell at a glance whether their part of the order needed steering.

03

Invisible performance

The platform stored only the current ETA, never the original. Every update moved the goalposts. There was no way to distinguish client-caused delays from operational failures.

The result was a Customer Success team that could only react to problems, instead of anticipating them before they impacted the customer.

To compensate for these platform gaps, five teams maintained a shared spreadsheet across the organisation, updated daily by hand. The workaround surfaced what was missing, and added problems of its own.

04

Duplicated effort

Every update had to be made twice: once in the platform, once in the spreadsheet. Five teams, every order, every change.

05

Always out of date

Updates were time-consuming so teams batched them. By the time status was logged, it was already lagging behind reality.

The Solution

The objective was to retire the spreadsheet by closing the gaps that had forced it into existence: a consolidated view of the lifecycle, early warning of milestone delays, and a preserved baseline for performance.

A unified timeline view

The spreadsheet was replaced by a unified timeline view, embedded in the order detail page and across all its instances. The view extended a chat surface that already existed in the platform, adding system events alongside the human comments it carried. Every event, ETA change, and team update now lives in one place, in chronological order. CS and Sales can reconstruct the full history of an order without leaving the platform. Supply chain and warehouse teams can spot which of their instances need steering at a glance.

Earlier warning signals

With every milestone now visible in one place, the traffic light could move from order-level to milestone-level. Yellow flags any phase running behind in a way that threatens the final ETA, surfaced on the list view weeks before the deadline. A delay in Design propagates to downstream instances automatically. Every team sees it without opening the order.

Measurable on-time performance

The original ETA is now preserved alongside the current one. Deviations are visible. Delays are attributed as client-caused or operational. For the first time, the business could measure its real on-time performance, understand why orders were late, and act on it.

The Outcome

BeforeAfter
Order lifecycle lived in spreadsheets and SlackEnd-to-end process traceable on-platform
Customer Success blind to developing delaysDelays caught internally, before client impact
On-time performance measurement was brokenReal ETA performance visible for the first time

Closing

Scale isn't something you add to a company. It's something the systems either support or don't. What customers experienced as delays came from an order lifecycle spread across spreadsheets, Slack threads, and five teams each holding part of the picture. Making it visible made it fixable. That's most of what designing for scale actually is.

Part (II): The Self-Service Platform. With the foundation stable, the customer experience work became the priority: redesigning the platform to do what the team had been doing manually. A separate case is in development.