Zipify OCU

Making a growing product easier to improve

Zipify OneClickUpsell (OCU) is a Shopify app for creating and optimizing upsell offers across the purchase journey.

When I joined, the product had accumulated usability issues and design debt, while design still worked mainly as a delivery function.

As Lead UX Designer, I focused first on the way product decisions were made. I brought design earlier into discovery and planning, added stronger validation and post-release feedback, and then led a gradual improvement of the product rather than a full redesign.

Role

Lead UX Designer

Domain

E-commerce / B2B SaaS

Team

5 designers on the product / 12 in the wider
design community

Focus

DesignOps, product strategy, research, and
team development

Early signals

  • SUS increased from 64 to 78
  • Dead clicks decreased by 38%
  • SEQ introduced for new features

A product that had outgrown its way of working

OCU had accumulated usability issues and design debt. The interface was becoming harder to extend, while the design team’s work was still concentrated mainly on UI delivery.

Designers were usually brought in after backlog and roadmap decisions had already been made. Tasks arrived with limited context, research was inconsistent, and the team had access mainly to high-level product statistics. There was no shared design process, no clear development path for designers, and no reliable way to learn from released changes.

As a result, the team could improve individual screens while the causes behind recurring problems stayed in place.

The UX problem started before the interface

Several visible usability issues needed attention. I first wanted to understand why the team kept producing them and where design could have more influence.

I ran a DesignOps workshop with the design team, followed by working sessions with product, engineering, support, and other stakeholders. Together, we mapped how decisions were made, where handoffs failed, what other teams expected from design, and which gaps were slowing the work down.

DesignOps Workshop 1
Where designers were losing context or influence.
DesignOps Workshop 2
The same problems viewed with product, engineering, support, and other stakeholders.
DesignOps Workshop 3
The operational changes we chose to address first.

The clearest finding concerned timing and ownership. Design entered too late to help frame the problem and often left too early to evaluate the result.

Moving design into the whole product cycle

The previous workflow placed designers near the middle of delivery. A task was assigned during planning, the designer created a concept, technical limitations were discussed, and the work moved into development. User validation and post-release evaluation were not established parts of the process.

I led the work to reshape this workflow. In the updated process, design joined discovery, backlog shaping, and roadmap conversations.

Before detailed UI work began, the team defined the user need, mapped the relevant journey, stated the research question, and aligned on technical constraints. Concepts were reviewed earlier, prototypes could be validated, implementation received design support, and released changes returned to analytics and a success check.

To support the new workflow, I also introduced regular one-to-ones and design retrospectives, created growth plans connected to product needs and OKRs, and began building a shared knowledge base. Collaboration with support and engineering became a regular source of input for design work.

The intended result was a continuous cycle in which design could help frame the problem, test the direction, and learn from the outcome.

Design Process Transformation

Product discovery, User research, PP definition, Ideas, Business\Users Request

  • UX/UI designer
  • Lead designer
  • Creative director
  • Business analyst
  • Product manager
  • Support manager
  • Tech lead
  • Marketing manager
  • Enterprise manager

Backlog creation, Lean + VP + Strategyzer feature validation

  • Product manager
  • UX/UI designer
  • Lead designer
  • Creative director
  • Business analyst
  • Support manager
  • Tech lead
  • Marketing manager
  • Enterprise manager

Roadmap Planning

  • Product manager
  • Lead designer
  • Creative director
  • Business analyst
  • Support manager
  • Tech lead
  • Marketing manager

Workshop on personas, usecases and journeys brainstorming. Concept ideation

  • Business analyst
  • UX/UI designer
  • Lead designer
  • Creative director
  • Product manager
  • Tech lead
  • Marketing manager
  • Enterprise manager

Task is assigned to designer on Planning

  • Lead designer
  • UX/UI designer
  • Product manager
  • Business analyst

Competitive research of UX/UI patterns and analysis

  • UX/UI designer

Task definition User needs + User journey for design + HMW + research question

  • UX/UI designer
  • Lead designer

Concept creating sketching and wireframing

  • UX/UI designer
  • Lead designer

Stories decomposition validation of the concept based on tech limitations.

  • Business analyst
  • UX/UI designer

Concept check ideation and finalizing requirements based on concept

  • Lead designer
  • UX/UI designer
  • Creative director
  • Business analyst
  • Product manager
  • Support manager
  • Tech lead
  • Marketing manager
  • Enterprise manager

Concept detalization + Updating DS

  • UX/UI designer

Check by design team

  • UX/UI designer
  • Lead designer

Prototype Creation

  • UX/UI designer

User validation

  • Lead designer
  • UX/UI designer
  • Business analyst
  • Product manager
  • Support manager

Corrections implementation that came from feedback's

  • UX/UI designer
  • Lead designer
  • Business analyst

Prep for dev + Updating DS

  • UX/UI designer

Design finalization check before Dev

  • Lead designer
  • UX/UI designer
  • Business analyst
  • Product manager
  • Tech lead

Team presentation on business refinement

  • Business analyst
  • UX/UI designer

Tech refinement

  • Business analyst
  • UX/UI designer

Design support

  • UX/UI designer

Design acceptance before release

  • UX/UI designer

Analytics / Success check

  • Product manager
  • Lead designer
  • Business analyst

Cycles of corrections and validations

No specific roles

The updated process brought design into discovery and kept it involved through validation, implementation, and post-release review.

Choosing evolution before redesign

Once the operating model was clearer, I assessed the product itself.

A heuristic review, available usage data, and recurring feedback from support, engineering, and product exposed two layers of work: immediate problems in key flows and broader debt across the interface.

A complete redesign was possible, but the cost was high. It would take longer and force existing users to relearn a product they already knew.

We started with the critical flows and kept the broader interface update for a later phase. This gave us a way to improve the product without resetting the whole experience at once.

Product design approach

Cognitive load
Evolution

Product-oriented approach. Systematic development makes transitions to the new types easy and comfortable for the user. Demands a moderate level of cognitive load during the transition.

Cognitive load
Revolution

This approach radically changes the product's look while demanding more time and effort in realization. Works for somewhat outdated products. It requires a high level of cognitive load during the transition.

We chose gradual change because the product already had established user habits. At that stage, a full redesign would have added more risk than value.

Applying the approach to the Offer Builder

The Offer Builder became a practical test of the new approach.

We began by bringing feature requests, known pain points, business expectations, and technical constraints into one discovery space. This helped the team see the problem as a connected system instead of a collection of requested screens.

Discovery process
Requests, pain points, constraints, and assumptions brought into one view before detailed design.

I then mapped the journey of creating a first funnel. This helped us see where users needed to make decisions, where additional context was required, and how individual steps affected the rest of the setup process.

User journey map
The first-funnel journey showed where users needed more context and where one choice affected later steps.

I also reworked the information architecture around the Offer Builder. The new structure showed how offers, products, funnel steps, and settings depended on one another.

Information architecture
The new structure made the relationships between offers, products, funnel steps, and settings easier to understand and discuss.

Together, these artefacts gave the team a shared model before detailed design began. They helped reveal dependencies, narrow the scope, and turn a broad request into a sequence that product, design, and engineering could discuss together.

What changed

Initial measurements after the broader programme of changes showed:

64 78
System Usability Scale

Increased from 64 to 78 across evaluated core workflows.

-38%
Dead Clicks

Decreased by 38% after resolving ambiguous click affordances.

SEQ
Single Ease Question

Introduced for new features closer to release for fast qualitative feedback.

These results reflect the broader set of product and process changes across the team. Their main value was showing that the direction was working and giving the team better signals for future decisions.

What did not stick

The team also created a set of shared design principles. The workshop helped surface different expectations, but the principles remained too broad to guide everyday product decisions. We had not connected them to concrete examples, real trade-offs, or clear ownership, so they were not consistently used afterwards.

Looking back, I would define fewer principles and test each one against a live product decision before treating it as shared guidance.

A principle becomes useful when it changes what the team chooses to do.

What I took from the project

Zipify changed how I think about design maturity. It depends on when design enters a decision, what evidence supports it, and whether the team comes back after release to learn.

It also changed how I think about leadership. A lead should not become the person who makes every difficult decision. The better result is a team that gets better at making those decisions together.