Skip to main content
All work

Case 02Software engineering · One Services Estimator

41% less time spent estimating

Turning an under-used estimation portal into a guided workspace estimators actually choose to use.

Context
Led in-house at TCS
Role
Lead UX design
Platform
Web application
Timeline
6 months

Illustration of the interface pattern, recreated in code. Not the shipped product, which belongs to the organisation it was built for.

Outcome

Outcome

  • 41%reduction in estimation time
  • 26%improvement in estimate accuracy
  • 61%increase in user satisfaction

Overview

What we were working with.

In enterprise software services, fast and accurate estimates win bids and protect margins. The One Services Estimator served estimators, pursuit leads, delivery architects, reviewers and developers, each with their own workflow and dependencies.

The problem

Only 20% of estimators used the official tool. Most fell back to manual Excel templates, which meant inconsistent estimates, slow reviews and little shared visibility.

Who we designed for

Kurt, estimator

“I need to confidently size a feature without constantly second-guessing myself.”

Goals

  • Produce accurate, defensible estimates quickly.
  • Rely less on fragmented input from subject experts, with a clear guided process.
  • Trust estimates through better visibility, consistency and supporting data.

Pain points

  • No indicators of how confident an estimate is.
  • Re-estimation with experts across time zones is slow and chaotic.
  • Translating customer requirements into effort is hard.

ScenarioEstimators turn customer requirements into project estimates while coordinating many stakeholders and fragmented inputs. The real challenge was doing that with confidence when requirements were unclear and re-estimation crossed time zones.

Process

How we got there.

  1. 01

    Research

    • Ran a usability inspection against Nielsen's ten heuristics.
    • Observed 12 estimators completing real requests, documenting actual against intended workflows.
    • Reviewed competing estimation platforms.
  2. 02

    Personas and journeys

    • Refined three core personas: estimator, reviewer and pursuit lead.
    • Mapped each end-to-end journey to find pain points and opportunities.
    • Found that 60% of estimation time went on coordination and rework, not estimating.
  3. 03

    Design

    • Rebuilt the information architecture around user tasks instead of system features.
    • Moved from skeleton mock-ups to interactive Figma prototypes.
    • Designed high-fidelity screens to brand standards and WCAG 2.1 AA.
  4. 04

    Test and hand off

    • Ran three rounds of RITE testing with eight users each.
    • Ran formal usability testing with 15 participants across all three roles.
    • Handed off component specs, type, colour tokens, interaction states and breakpoints.

Design decisions

The calls that shaped it.

  • Workspace, not portal

    We reshaped the entry experience so people could start a new estimate, pick a template or return to recent work in one step, replacing the pull back to Excel.

  • Quick and detailed modes

    A dual-mode approach balanced speed against rigour through progressive disclosure. The pattern has since been adopted in other enterprise tools.

  • Collaboration in context

    Instead of a full messaging system, lightweight comments, mentions and status indicators delivered most of the value for a fraction of the complexity.

What we learned

What we carry into every project.

  1. Research earns its time

    Spending 30% of the timeline on discovery (heuristic evaluation, contextual inquiry and customer feedback) meant we solved real problems, not assumed ones.

  2. Progressive disclosure

    Letting people choose quick or detailed estimation served two competing needs without compromising either.

  3. Small features, big value

    Contextual collaboration gave roughly 80% of the benefit of a messaging system for 20% of the effort.

Next caseReimagining special occasions through ambience control

Let's build it.

Tell us what you are making. We reply to every enquiry within one working day, and you will hear from the people who would actually do the work.

London, United Kingdom · Working worldwide