Silo · Produce Supply Chain Navigation Overhaul
Silo · Produce Supply ChainSilo Next project
Navigation overhaul at Silo

Rebuilding a navigation that couldn't scale.

Silo's original navigation was built around two roles: buying and selling. As the platform grew, that structure stopped working, forcing a completely different, isolated view every time someone switched roles. I drove a design-led initiative to raise awareness of the problem and overhaul the IA, without breaking the workflows our users depend on.

Role
Lead Designer · Design-led initiative
Team
Design, Product, Eng · Silo
My focus
Awareness → end-to-end UX
01 · Why update?

A navigation built only for two roles. The platform outgrew it.

As Silo moved to a more modular product strategy, each new product area needed its own place to live, and the original role-based navigation had none to give. Accounting tangled with capital, inventory pulled away from orders, and the structure had nowhere left to grow.

1 → 5product managers, each building
a different product area
For Silo

A more modular product strategy meant more offerings, and a real risk: every new feature had nowhere to land.

For users

Before it got worse, users needed a more flexible experience, one that put what they actually needed within easy reach.

Before: one account,
two isolated views
Purchase Orders Vendors Buyer ▾ PL
Inventory Sales Orders Customers Packing Logistics Seller ▾ PL
Key issues
01
No hierarchy
No sense of grouping for similar features, and the horizontal space ran out fast as more got added.
02
Isolated views
Switching from seller to buyer swapped the entire experience, not matching how the team works.
03
Hidden core features
Accounting, a revenue-generating product offering, was buried in user settings like an afterthought.
02 · Raising awareness

Everyone agreed there was a problem. Getting it on the roadmap was the hard part.

As a passion project, the hard part wasn't the design. It was building enough momentum to earn a spot on the roadmap, then getting users comfortable with the change once it shipped.

User feedback at every stage
Feedback shaped every iteration, from early concepts through final review.
Cross-functional brainstorms
Initial conversations in product grew into brainstorms with design and engineering leads.
Keeping everyone in the loop
Weekly syncs and a dedicated Slack channel kept users and leadership in the loop.
03 · Defining the approach

Card sorting our way to a structure that made sense.

We ran two card-sorting exercises with product, design, and engineering leads, each sorting based on what their own product focus area actually needed. That moved us from divergent ideas to a convergent structure.

Exercise R1Divergent card sorting
{{ r1Stage }}
Exercise R2Convergent sorting & defining
{{ r2Stage }}
R1 called out all important features and grouped them into parent/child relationships. R2 organized which features surface on the navbar vs. nest within sections.
What we uncovered
01
Color meant something
Users had learned to associate green and blue with order type, purchase vs. sales, not with the buyer/seller split itself. That association carried over into the new design.
02
Bookmark by job role
A few wanted to bookmark certain pages based on their role: accounting, sellers, warehouse managers.
03
Keystrokes, never the mouse
Users relied on keystrokes, so we defined the right shortcuts and keyboard navigation.
04
Don't remove capabilities
Users were fine with the changes, as long as nothing they already had went away.
04 · Into the details

Tackling the work
across every page.

With a stronger approach from all the iterations and feedback, it was time to apply the new navigation across the whole platform, and make sure it didn't break a single workflow.

Audited every workflow

We went page by page, Orders, Warehouse, Logistics, Finance, Reports, mapping what moved where.

Preserved every capability

Feature parity was the rule: nothing users relied on could quietly disappear in the new IA.

Localized as we went

With most of our users Spanish speaking or dual language, translations weren't an afterthought, they were updated across every main page component alongside everything else.

The Audit
Every page, checked for navigation,
feature parity, and translation.
Navigation mapped
Feature parity
Translated
{{ boardRows }}
Complete
In progress
05 · Change management

Our users are old-school. The rollout needed a lot of care.

Most of our users came from a pen-and-paper background, using digital software at all was already a stretch for them. Any change to a tool they'd finally gotten comfortable with was going to be hard. We had to give engineering time to maintain both navigations until everyone felt ready to let the old one go. For long-standing customers, customer success stayed in touch throughout, real conversations, not just in-app messaging, so they felt cared for, not just notified.

Phase 1 · 1 week
Try out the modern layout

In-app onboarding with a video walking through the benefits and updates.

40%
of core users switched to try the new navigation.
Phase 2 · 4 weeks
Switch all to modern layout

Move everyone not already on it, with an in-app update that we're sunsetting the old one.

60%
of core users are on the new navigation at this phase.
Phase 3 · Final
Say goodbye to the old layout

A final in-app update: the official switch is here, and users can no longer revert back.

100%
officially switched. No requests to switch back.
06 · The experience

Final designs.
The modern navigation.

A persistent, scalable structure with the things users asked for, surfaced where they need them, and flexible enough to grow with the product.

Key improvements
Favorites
Quick access to the most-used pages.
Search
Find the page they're looking for, faster.
Reports & Finance
Now fewer clicks away than before.
Spanish translations
With most of our users Spanish speaking or dual language, this wasn't a nice-to-have — translations were updated across every main page component.
Order type, color-coded
Purchase and sales orders kept the green and blue color association users had already learned — one piece of the old system worth carrying forward.
Responsive design
Nav components flex down to smaller breakpoints.
Keyboard navigation
More accessible, with a shortcut to open search.
S Silo | Orders | Online Orders PL Parker Lawson Dexter Farms Co.
Orders
Purchase Orders Sales Orders EDI Orders Online Orders Price Sheets
Orders
Purchase Orders Sales Orders EDI Orders Online Orders Price Sheets
$
All Order Requests Reset filters
IDStatusCustomerRequested DateOrder Total
Purchase orderSales order
48591Open NEWDexter Farms10/29/23$9,240
48590Open NEWBlurred Limes Marketplace10/29/23$16,120
48589OpenLettuce Celebrate LLC10/28/23$5,830
48588Confirmed w/ adjustmentsInner Peas Produce10/28/23$11,940
48587ConfirmedDexter Farms10/27/23$8,415
48586CancelledBlurred Limes Marketplace10/27/23$3,050
48585ConfirmedLettuce Celebrate LLC10/26/23$14,200
48584OpenInner Peas Produce10/26/23$6,375
Recreated UI, opens from either the breadcrumb or the Orders icon in the rail, same menu either way.
07 · Learnings & reflection

We shipped a universal
change, without a single
request to bring the
old nav back.

With the team

Organizing Finance was an ongoing battle, those product owners wanted a separate navigation entirely.

It didn't make sense to build a separate experience for engineering to maintain, so we folded Capital under Finance until that line had clearer direction.

With the work

A universal change meant a lot of gnarly bugs surfaced as we built.

Updating to ensure feature parity
Addressing different breakpoints
Localization for Spanish speakers
With customers

Their attitude was indifferent, but we knew they were resistant to change. The game plan:

Kept users informed via email updates
In-product announcements
Over-communication with customer success
What's next

By the time rollout hit 100%, the accounting and capital overlap that helped cause the original mess had become one of the first clean separations the new nav proved out.

Same with freight. What used to be a single "Book Freight" link became a full Logistics destination as the TMS build-out grew, same place in the nav, it just grew into it.

Coming out of this work, we recognized the need to tackle universal page-level improvements next. As the design team grew, I was able to hand this next stage off, folding it into our design system and product roadmap.

No one asked to bring the old nav back.

{{ fabIcon }} {{ fabLabel }}