DELIVER MY CONTAINER
Product Strategy Interaction Design Decision Design Marketplace Design B2B

Repositioning a Freight Marketplace into a Workflow Platform

Problem

Users could compare options and find cost savings — but couldn’t confidently commit. The platform stopped at selection, leaving users to manage risk and execution off-platform.

Decision

Reposition the platform from a comparison tool to a workflow entry point — enabling users to commit with confidence, not just compare options.


Metric Spotlight
Conversion rate +38% Up from 21% — driven by removing friction at the confirmation step
Time on task 4.1 min Down from 9.3 min across all tested cohorts
Drop-off reduced −57% Step 2 abandonment eliminated after copy and structure changes
Users saved money — and still didn’t come back.

The platform was designed like e-commerce — freight is not e-commerce.

When I joined DMC, the platform had been designed around a simplified, price-led model of procurement. The product assumed procurement could be reduced to a fast, price-led interaction.
In reality, selecting a forwarder is a high-risk decision shaped by uncertainty, liability, and operational consequence.


The model the product was built on did not reflect reality
Assumed Model (product logic)
Actual Decision Context (real-world)
Decisions treated as price comparison
Decisions driven by risk and reliability
Selection seen as low-risk
Selection carries commercial consequence
Outcomes assumed reversible
Outcomes are irreversible
Minimal information considered sufficient
Requires complete, structured information
The product was built on the wrong decision model.

Faster pricing didn’t translate into repeat usage

The platform improved speed at the start of the journey, but failed at the point that mattered — selection and commitment. As a result, usage increased, but transactions and retention did not.


Business goals (what success depended on)

→ Increase quote-to-award conversion
→ Drive repeat usage across shipments
→ Build liquidity across forwarders

User behaviour (what actually happened)

→ Avoided committing without certainty
→ Required clarification before selection
→ Defaulted to known forwarders

Users tried the platform — then worked around it

Users didn’t fail to choose — the system failed to support the decision. At the point where confidence mattered, the product introduced friction instead of resolving it.


User Behaviour Loop

Fig 1 — Users weren’t choosing between prices — they were managing risk, effort, and uncertainty. The issue wasn’t access to options, but the anxiety of selecting an unknown forwarder.

Cycle repeats
Uncertainty remains
01 — List User posts a job → Expecting fast, reliable comparison
02 — Compare Quotes reviewed → Price visible, scope unclear
03 — Evaluate Risk assessed → Missing detail blocks confidence
Scope unclear
Confidence drops
04 — Clarify Details confirmed → Requires effort to resolve uncertainty
05 — Friction Confidence drops → Effort increases perceived risk
06 — Exit User defaults away → Platform abandoned before commitment
Effort increases

User Research Quote

We already have forwarders we trust. The platform didn’t make it easier — it just added more to manage.

Logistics Manager, SME importer

Cost only mattered once everything else felt safe

Importers were not optimising for price alone. They were evaluating whether a forwarder could be trusted to deliver without introducing risk, effort, or uncertainty.

Selection only happened when confidence was high across multiple factors. Any uncertainty stopped the decision entirely.

How selection actually works
The Trust Gate Price is ignored at this stage
  • Reliability
  • Clarity
  • Effort
  • Risk
  • Fit
Fwd B: $1050 Discarded — Failed Trust Check
Only “Safe” Quotes Remain
Fwd A: $1200 Fwd C: $1400
Price Comparison $1200 (Fwd A) vs. $1400 (Fwd C)
Selection Fwd A

We redefined what the platform is for

The platform was built for price comparison. We rebuilt it to support decision and execution.

Before - price comparison tool
01 Quotes
02 Compare Price
03 Accept
04 End
Platform behaviour

Price replaced judgement

The interface reduced forwarders to price points.

Key decision inputs - scope, process, and delivery - were not visible at the point of selection. Users could compare quickly, but not decide with confidence.


After — Selection became the start of the workflow
01 Quotes Received
02 Evaluate
03 Accept
04 Manage
05 Track
06 Delivery Confirmed
Step 1: Evaluation

Decisions supported before commitment

Users can evaluate scope, delivery, and process before accepting.

Confidence is built before selection - not after.

Step 2: Commitment

Selection transitions into execution

Accepting a quote activates the job — it does not end the interaction.

The platform becomes the system of record from this point forward.

Step 3: Execution

Execution handled in-platform

Tracking, documentation, and updates are managed within the product.

Users no longer rely on external coordination to complete the shipment.

We traded simplicity for decision confidence

We prioritised decision confidence over speed and simplicity — accepting increased complexity at the point of evaluation.

01

More effort to evaluate

Quotes were structured around real service differences rather than flattened for speed. Each one took more effort to review — but users understood what they were actually buying before committing.

02

Harder to scan and compare

Service variation was made visible rather than uniform. Quotes resisted quick side-by-side comparison — but users could evaluate substance, not just price.

03

A heavier evaluation flow

Key detail was moved upfront rather than hidden behind progressive disclosure. The process felt heavier — but removed the need for off-platform back-and-forth at the point of decision.

04

A broader, more complex product

Workflow support was built into the platform rather than kept transactional. The product became harder to scope — but prevented drop-off after selection and supported real operations.

We removed the illusion of simplicity to enable confident decisions.

The interface was redesigned to support evaluation, decision, and commitment

The interface was redesigned not just to compare prices, but to support evaluation, surface differences, and create a clear moment of commitment.

New quote view
Platform behaviour

Trust as the entry point

The redesigned interface leads with forwarder credibility and service clarity. Price becomes the final filter — not the first impression.

Three Layers of the Redesign
A

Evaluation

Compare services, not just prices

Structured quotes aligned to a defined service. Consistent inputs enabling like-for-like comparison. Clear scope of what is being offered.

B

Decision

Make differences visible before commitment

Inclusions and exclusions surfaced explicitly. Variation between quotes made legible. Reduced need for interpretation or follow-up.

C

Commitment

Selection becomes a commitment, not a guess

Defined service carried forward at selection. Clear expectations set before engagement. Reduced likelihood of post-selection surprises.

Product in Use

Comparison view — annotated primary view

Expanded quote breakdown — detail layer

Selection moment — interaction / state change

We traded growth speed for product correctness

This shift moved the platform away from a simple comparison tool toward a more structured, opinionated product. It aligned with real user behaviour — but introduced product and marketplace risk.

Strategic Trade-offs
  1. 01
    Simplicity → Accuracy
    Less lightweight, but more representative of real-world complexity
  2. 02
    Broad appeal → Targeted fit
    Reduced universality, increased relevance for SMB importers
  3. 03
    Fast transactions → Informed decisions
    More deliberate selection, with potential impact on short-term conversion
  4. 04
    Marketplace neutrality → Product opinion
    Shifted from displaying options to shaping how decisions are made
Marketplace Risk
Risk 01

Reduced emphasis on price could weaken perceived cost advantage

Risk 02

Increased structure could discourage less experienced users

Risk 03

Forwarders could resist transparency or standardisation

Constraints Not Fully Solved

Off-platform behaviour when relationships were established

Fee model incentivising disintermediation

The product became more correct — but less forgiving.

Selection became clearer — and required less intervention

Before — clarification-led selection
Price-first, 2024
List
Quote requested
  • Minimal service detail
  • No scope defined
Quote
Price received
  • Assumption-based
  • Inclusions unclear
Clarify
Off-platform
  • Follow-up required
  • Decision delayed
Decide
Hesitation point
  • Risk unresolved
  • Default to known
Leave
Platform exit
  • Good results: off-platform
  • Poor results: churn
Users exited or reverted to known forwarders — good experiences moved off-platform, poor ones led to churn.
After — comparison-led selection
Service-first, 2025
List
Structured service
  • Defined scope
  • Clear inclusions
Evaluate
Visible differences
  • Like-for-like comparison
  • Variation made legible
Select
Direct commitment
  • Reduced hesitation
  • Expectations set
Continue
In-platform
  • No validation loop
  • Engaged new forwarders
Users progressed from comparison to commitment without off-platform validation.

Selection shifted from clarification-led to comparison-led, with users making decisions based on visible service differences rather than defaulting to existing relationships. This reduced the need to move off-platform to validate decisions, enabling users to progress directly from comparison to commitment.

Before vs After behaviour loop

The problem wasn’t the interface — it was the model

The platform did not fail due to lack of features, but due to a mismatch between how it was designed and how freight procurement actually works. It treated procurement as a transactional, price-led interaction — when in reality it is a decision and operational process shaped by risk, clarity, and coordination.

The Shift
01 Price optimisation
02 Decision confidence
03 Workflow continuity

This required moving from simplifying the problem → to representing it correctly — and designing for how users actually operate, not how the process appears on the surface.

Effective products don’t reduce complexity — they make it legible.