Shell · Employee Experience · Aug—Dec 2022

I redesigned onboarding while being onboarded

My first project at Shell was fixing the thing I was living through. Four months, eleven interviews, three personas, and a pitch for a budget that did not exist.

My role

My role

Service designer & user researcher

Service designer & user researcher

Timeline

Timeline

August — December 2022

August — December 2022

Context

Context

Shell e-mobility division

Shell e-mobility division

Outcome

Outcome

6 concepts funded and shipped

6 concepts funded and shipped

11

New joiners interviewed

3

Personas

37

Mapped journey steps

11 → 6

Concepts prioritised to shipped

The problem

Everything existed. Nothing was findable.

The e-mobility division was hiring fast. Every resource a new joiner might need existed somewhere — scattered across multiple SharePoint folders with inconsistent access permissions, duplicated in several versions, with no indication of which was current.

There was no checklist. No 30/60/90 day plan. No single source of truth. Knowledge was simultaneously everywhere and concentrated in a handful of individuals, and new joiners were left to work out who those individuals were.

Making the case

The argument had to come before the design

Onboarding had been handed over as an afterthought with no budget attached. Before I could design anything I needed to establish why it was worth funding at all, so I built a secondary evidence base on how onboarding quality affects retention, satisfaction and how quickly people become productive.

[00%]

[what this figure shows]

[Author, Year]

[00%]

[what this figure shows]

[Author, Year]

[00%]

[what this figure shows]

[Author, Year]

Figures above are from published secondary research used to build the business case, not findings from this project.

Positioning

My first project, and I was inside the research population

I joined Shell and was handed this project almost immediately, which meant researching an experience I was in the middle of having. That cut both ways. It gave me unusually direct access to the confusion I was documenting, and a fast education in the division’s processes and stakeholders. It also meant being deliberate about separating what I was finding from what I was personally feeling.

Research approach

Eleven conversations, three phases, one map

I interviewed eleven people who had recently joined the division, selected to span different entry routes into it. The journey was broken into three phases — pre-onboarding, company onboarding, and functional onboarding — because the problems in each turned out to be different in kind, not just in degree.

In-depth interviews

Eleven recent joiners across three distinct entry routes into the division.

Secondary research

Published evidence linking onboarding quality to retention, satisfaction and time to productivity.

Ideation workshops

Collaborative concept generation producing eleven candidate concepts.

Who we were designing for

Three people, arriving three different ways

Rather than segmenting by job title, the interviews surfaced three patterns defined by what someone already knew when they arrived. The cycle had to keep all three engaged without repeating content for one or skipping it for another.

The new joiner

New to Shell and new to e-mobility. Needs both onboardings at once, with no existing map of either.

The internal mover

Knows Shell, new to e-mobility. Company onboarding is redundant; functional onboarding is everything.

The contractor

Arrives for e-mobility only, with restricted access to Shell systems and processes permanent staff take for granted.

Key findings

Three pain points, all of them about orientation

01

No idea who to ask

No clear guidance on how to work or who to approach for help. A line manager was often not from e-mobility at all, while the managers, PMs and stakeholders someone actually worked with were. Managing two parallel onboardings with no map between them was its own problem.

02

Information without shape

Responsibilities and tooling in the new department were unclear, and people wanted to know before arriving rather than after. They were then flooded with material with no sequencing. Most could not explain which team or wing of the division they belonged to.

03

No sense of the whole

People were unaware of the division’s multiple product lines and projects, so they had no picture of where their own work sat. The team was expanding fast enough that navigating it was difficult even for people who had been there a while.

Interview insights

Key takeaways from eleven interviews

Synthesis

Thirty-seven steps, three personas, one map

I mapped a nineteen-step pre-onboarding journey and an eighteen-step onboarding journey, both across all three personas, marking at every step which personas were involved and which were not. Seeing where the three diverged was what made a single cyclic programme possible rather than three separate ones.

Pre-onboarding journey — 19 steps across three personas

Onboarding journey — 18 steps across three personas

Ideation

Eleven concepts across the whole cycle

Ideation workshops produced eleven concepts spanning every phase of the journey — some addressing what happens before someone’s first day, others reaching into their first quarter. They were documented as a concept book so stakeholders could see the full set before anything was cut.

[concept name]

[concept name]

[concept name]

Prioritisation

Eleven concepts, and money for none of them

With no budget to build all eleven, I prioritised against employee impact, business impact and implementation effort — deliberately selecting a mix of quick wins and slower structural changes rather than optimising for either alone. Six made the cut, sequenced into a roadmap.

Concept prioritisation — employee impact against business impact and effort

Target concept roadmap — sequencing the six funded concepts

Getting it funded

The project began with no billing code

Onboarding was handed over as an afterthought. No budget attached and no billing code to charge time against, which is a fairly precise measure of how seriously it was being taken.

So the deliverable became an argument. I took the prioritised concepts and the evidence base to senior stakeholders and made the case for why onboarding was worth funding at all. They backed it — including budget to produce physical welcome boxes for new hires, which had started as one of the less defensible ideas on the list.

The outcome

A cycle, not a document

The six funded concepts were assembled into a single repeatable onboarding programme rather than six separate deliverables — a cycle that could run for every new joiner and improve each time.

01

Pre-onboarding

02

First day

03

First month

04

First quarter

05

Measuring success

The toolkit specified what happens at each stage, including email templates managers could send so new hires received the right information before they arrived.

Six of eleven concepts were funded and built, spanning quick wins and longer-term structural changes.

Onboarding toolkit — stage-by-stage guidance

Onboarding toolkit — stage-by-stage guidance

~25%

Estimated cost saving

~20%

Estimated satisfaction improvement

Both figures are directional estimates agreed with stakeholders at the time rather than independently audited measurements.

Reflection

What surprised me was the duplication. The division was not short of material — it had produced the same guidance many times over, in parallel, with no consolidation and no way of telling which version was current. The problem was never a lack of content.

The hardest part was getting people to speak honestly. Almost nobody wanted to say out loud that they did not really understand the division they worked in, or that they were executing tasks without knowing what they were building toward. Getting to that admission, repeatedly and gently, was most of the research.

© 2026 by Joel Mathew. All Rights Reserved.

© 2026 by Joel Mathew. All Rights Reserved.

© 2026 by Joel Mathew.

All Rights Reserved.