How can we make UK recycling decisions instant and accurate, right at the bin?

A self-initiated concept app aligned to Biffa’s visual system. Scan an item, apply your council’s rules, and locate the nearest compliant drop-off. Postcode-specific guidance in seconds.

Jump to Outcomes

Project:

Concept mobile app that makes recycling decisions fast and trustworthy for UK residents.

Timeline:

2 weeks (concept + validation)

My Role:

UX research
Survey design
User interviews
Affinity mapping
Information architecture
Lo-fidelity and mid-fidelity wireframes
Usability testing
Front-end prototyping

Team:

4 people:
UX: Anna Salvadori, Jan Helmers

UI: Pati Dlugosz, Bharath Kannan

Tools:

Figma / FigJam
Google Forms / Drive
GitHub
Trello

My Contribution

This was a four-person team: two UX designers and two UI designers.

On the UX side, the survey was designed, distributed, and analysed by me. I conducted all five user interviews and built the affinity map from the raw data. The lo-fidelity and mid-fidelity wireframes were primarily my work, produced in Figma and tested across two usability rounds. Jan Helmers and I shared the journey mapping and user flow work equally. The competitor analysis was led by Jan.

The hi-fidelity visual design was produced by Pati Dlugosz and Bharath Kannan. My role at that stage was to review screens against the UX findings and flag inconsistencies before handoff.

The concept landing page, built in HTML and CSS and aligned to Biffa’s visual system, was designed and developed by me independently.

TL;DR outcomes

Usability completion rates improved across two testing rounds (n=5 per round):

Registration task: 50% to 100%
Pick-up service task: 60% to 100%
Observed reduction in hesitation on the Home screen between iterations

Testers described the final flow as “straightforward” and “quick to book”. Council-specific results and certainty labelling reduced “I’ll double-check” behaviour during sessions.

Note: sessions used 5 participants per round. Findings are directional, not statistically conclusive.

Overview

Design and validate a concept feature set for a Biffa-aligned recycling app that gives postcode-specific disposal guidance and nearby drop-off options. The primary input is a photo of the item; if not recognised, users enter a short description.

For Biffa, resident recycling behaviour is not simply a public service issue. Contamination rates and disposal volumes directly affect the operational efficiency of council contracts. A tool that reduces incorrect disposal and failed trips supports both resident trust and Biffa’s position as a reliable service partner to local authorities. The problem this project addressed is operational as much as it is experiential.

Problem & Constraints

The problem UK recycling guidance changes by council. Answers are buried in PDFs. Drop-off sites may be closed or charge fees the user was not expecting. Residents who want to recycle correctly have no fast, reliable tool at the point of decision.

Design constraints This is a concept project, not an official Biffa product. Nationwide data integration, rewards systems, and carbon scoring were out of scope for this sprint.

Discover

Market Research

The competitor analysis, led by Jan, compared enterprise providers (Veolia, SUEZ, Viridor) and consumer apps (Horizon, Recycle Coach) against Biffa’s public site. The analysis identified patterns users already expect: a map and list toggle, postcode selection, a camera input, and plain-English guidance on what goes where.

A positioning chart plotted the field on two axes: trust in official local rules (low to high) and speed at point of use (slow to fast). Enterprise providers scored high on trust but low on speed. Consumer apps were faster but less authoritative. The concept target was the top-right quadrant: fast and trusted, combining council-specific results with photo recognition.

User Research

Quantitative Data – Surveys

A Google Forms survey was distributed via social channels and email. 13 responses were collected. At this sample size, findings were treated as directional and used to shape interview questions rather than draw conclusions independently.

Qualitative Data – Interviews

Five semi-structured interviews were conducted with participants based in Leeds, London, and nearby areas. Sessions explored recent disposal tasks, how people chose a bin or site, and what went wrong. Key patterns that emerged:

Residents were not confident without clear, local instructions, particularly for mixed materials.
Recycling points felt inconvenient. Wasted trips due to closed hours or unexpected fees were a common frustration.
Many items had no reliable barcode. Photo recognition with a manual description fallback matched how people expected the input to work.
Users cared about recycling correctly but cited confusion and inconvenience as the main reasons for defaulting to general waste.

Define

Making sense of the data

Affinity mapping

Interview and survey data was organised into an affinity map to identify recurring themes. Four patterns shaped the design direction:

Clarity gap: residents lacked confidence without council-specific, item-level guidance.
Trip friction: distance, unknown hours, and unexpected fees caused wasted journeys.
Input reality: photo recognition with a manual fallback matched real-world behaviour.
Motivation versus friction: environmental intent was present, but convenience gaps overrode it.

Value Proposition Canvas

A value proposition canvas was used to map user jobs, pains, and gains against the proposed solution. This confirmed the core tension: confidence in recycling decisions rises when guidance is local and explicit, and drops sharply when the system returns no result or the trip fails.

User Persona

A primary persona was developed from the interview data.

John Andrews, 37, London. Senior Business Consultant. Busy schedule, recently moved address, unfamiliar with local rules. Needs correct disposal guidance within seconds, with realistic trip information before committing to a journey. Frustrated by conflicting guidance, closed sites, and ambiguous mixed-material items.

The persona kept decisions grounded in realistic constraints rather than ideal-user assumptions.

User journey map

The current-state journey was mapped across five phases: noticing waste, searching for rules, finding a site, attempting drop-off, and aftermath. The lowest emotional points were contradictory guidance, unknown site status, and ambiguity around bulky or unusual items.

The journey made clear why friction occurs at the point of decision: users need local authority clarity and operational realities before they commit to any action.

Problem statements

Three problem statements were drawn from the research:

At the bin, residents lack authoritative, council-specific guidance and hesitate, leading to delays or contamination.
Planning a drop-off is guesswork because hours, fees, distance, and accessibility are not surfaced before the trip.
Edge items stall progress. When the system does not know, users hit a dead end and defer to general waste.

How Might We…

How might we deliver a council-specific answer in seconds using photo recognition, with a clear manual fallback?

How might we surface today’s hours, fees, distance, and accessibility before a user commits to a drop-off?

How might we turn an unknown result into a safe, actionable next step rather than a dead end?

Develop

Problem meets solution

Ideation & Prioritisation (Impact/Effort)

An Impact versus Effort exercise was run across all brainstormed ideas. Each idea was pressure-tested against the value proposition canvas to confirm problem-solution fit. Three features emerged as the highest-impact, lowest-effort priorities:

Photo capture with manual description fallback
Council-specific results with explicit certainty labelling
Drop-off list showing today’s hours, fees, distance, and accessibility

Minimum Viable Product

The MVP was defined as the smallest feature set that allows a user to get a trusted, local answer and a realistic plan within seconds.

In scope: photo capture, manual description fallback, council-specific guidance with certainty chips, drop-off list with live operational details, clear fallback for unrecognised items, and a one-tap report function.

Out of scope for this sprint: pick-up scheduling, rewards, community features, education hub, and nationwide data coverage.

User flow

The primary flow is linear. From Home, the user taps Scan, takes a photo, and receives a council-specific result with a certainty label. If the photo is not recognised, a manual description input reaches the same result screen. From there, the user selects a drop-off option filtered by today’s hours, fees, distance, and accessibility, then navigates directly.

If the system cannot match the item, a safe fallback is shown for that council alongside a one-tap Report item option.

Rejected Decision

Why photo-first over search or barcode scanning

Two alternative input methods were considered and rejected early in the develop phase. A text search input was dismissed because it requires users to know the correct terminology for an item before they can get a result, which replicates the exact knowledge gap the app is trying to close. Barcode scanning was dismissed because a significant proportion of recyclable items either have no barcode or carry damaged, unreadable codes, a pattern confirmed directly in user interviews. Photo recognition with a plain-English manual description fallback was chosen because it matches how people actually encounter items at the point of disposal: visually, with limited information, under time pressure.

Deliver

Wireframes to prototypes

Lo-Fidelity Wireframes

Paper sketches and low-fidelity tap-throughs were tested with five participants across the core journey: sign-up, scan an item, view guidance, and plan a drop-off or pick-up.

Completion rates at this stage: registration 50%, how to recycle 40%, pick-up service 60%.

Key findings:

The Home screen felt unfocused. Users wanted to know immediately whether they were booking a collection or finding a drop-off.
The scan screen had no visible cancel option.
The result screen included too many negative labels. Users wanted the next action, not a list of what was not possible.
Several participants expected a pick-up option immediately after item recognition, not later in the flow.

Mid-Fidelity Prototype

Changes were translated into a clickable mid-fidelity prototype and re-tested with a second group of five participants. This round focused on labelling, hierarchy, and task visibility.

Changes made based on feedback:

Scan affordance on Home: added a text label next to the camera icon.
Result screen: enlarged the primary action card; removed redundant negative labels.
Payment flow: removed one screen; saved payment details now carry forward to checkout.
Recognition screen: added Scan again and Choose from photos options; editable item title via pencil icon.
Driver comment field added before payment.

Completion rates after iteration: registration 100%, how to recycle 60%, pick-up service 100%.

Hi-Fidelity Prototype

The hi-fidelity visual design was produced by Pati Dlugosz and Bharath Kannan. My contribution at this stage was design review, UX feedback, and two specific decisions that came directly from testing findings.

Guided first step

Testing showed that first-time users were disoriented on Home. There was no clear signal indicating where to start. Based on this finding, a guided walkthrough overlay was recommended: a lightweight tooltip pointing to the Scan button on first launch, with the copy “Tap here to recycle an item” and a Skip option. The tooltip fades after first use to keep the interface clean on return visits. The copy and interaction logic were specified before handoff to the UI team, who executed it within Biffa’s visual system.

Certainty labelling

A recurring pattern in testing was that users lost confidence when guidance felt generic or ambiguous. To address this, certainty labels were introduced on the result screen, indicating whether an item Always, Sometimes, or Never belongs in a given bin for that council. This reduced observed “I’ll double-check” behaviour in the final usability round.

Final results

All three tasks completed at 100% (n=5). Testers described the flow as “straightforward” and “quick to book”.

One participant missed the sign-up option on entry. Log in and sign up were placed side by side on the entry screen to resolve this.

Concept Landing Page

A lightweight landing page was designed and built in HTML and CSS to introduce the app within Biffa’s existing site structure. It follows Biffa’s visual standards and explains the core flow in plain English: take a photo, receive a council-specific result, locate the right bin or drop-off point with today’s hours and fees.

Live link: https://asalvadori.github.io/Biffa_App/

Reflection

The project confirmed that UK residents have a genuine recycling intent problem, not just a knowledge problem. The friction is operational: unclear local rules, unknown site hours, and no reliable tool at the moment of decision. Photo recognition with a local-first result addressed this directly.

Two things would be done differently with more time. The MVP scope expanded across three iterations to include pick-up scheduling, payments, driver tracking, and rewards. Some of that expansion was driven by testing insights, but not all of it was validated before being added. Tighter scope discipline in the develop phase would have produced cleaner iteration.

The usability sample of five participants per round is a recognised limitation. Findings directed design decisions but should not be treated as conclusive. A second round of testing with a larger, more geographically varied group would be the next priority before any pilot.

Next Steps

Further usability testing with a broader participant group. Refinement of the item recognition and booking flows. Accurate council-rule data integration for a pilot area.