Shaping HerCity

The question behind the question

Cities are not neutral. UN Women's report Safe Cities and Safe Public Spaces (2017) and a 2015 Dublin study funded by Dublin City Council's Safe City Programme converge on the same finding: fear of gender-based violence restricts women's mobility, their participation in public life, and their access to leisure, with sharper effects for migrant women, lesbians, and gender-diverse people.

The cost is rarely visible as a single dramatic event. It accumulates as a thousand small adaptations. The women in this study described changing routes and schedules to avoid streets perceived as risky after dark; choosing whether to take a bus or wait for one depending on who else was at the stop; paying for taxis to get home safely, a recurring expense their male peers simply do not have, effectively a tax on accessing the same nightlife, jobs, and social life; reconsidering clothing choices before going out; and sustaining a constant background vigilance, scanning, calculating, staying alert to pre-empt harassment before it happens. Each adaptation is small. Together they amount to a city only partially inhabited: a structural limit on where women go, when, with whom, and at what cost, financial, social, and cognitive.

That framing shaped the research question from the start. I deliberately did not ask "how can we protect women?", a question that produces panic buttons and tracking apps. I asked instead:

How can urban interaction design be leveraged to create or transform urban spaces, fostering inclusive and enjoyable experiences for diverse women in Dublin today?

The unit of analysis was not danger. It was the full urban experience: freedom of movement, belonging, trust, and joy.

Listening to the city

To capture something as layered as urban experience, no single method would do. I designed a multi-sited ethnographic process where each method captured a different layer of the phenomenon, then triangulated across them:

• Six semi-structured interviews explored where women feel constrained or free, and why.
• A 15-day diary study with six women (locals and migrants, diverse sexual orientations) captured experience in real time: voice notes, photos, annotated routes, rather than in retrospect.
• An exploratory survey (15 women, aged 21–40, 35% Irish / 65% migrant) quantified patterns.
• Two rounds of collective mapping in parks and at Trinity College put a physical map of Dublin between participants and invited them to mark, annotate, and discuss.
• Participant and non-participant observation in venues and in women's WhatsApp and Facebook groups revealed how safety information already circulates informally.
• Findings were cross-validated against Google Maps reviews, checking whether places participants flagged matched what other women reported publicly.

The findings drew a precise map of invisible constraints:

"I tend to do things close to home if I am on my own… not to say I won't travel far by myself, but I think just as a habit of safety I prefer areas that are close to where I live that have some activity and are well lit."Research participant

Migrant and LGBTQ+ women carried additional layers: racism, xenophobia, harassment tied to visibility.

Synthesising the research: personas

To carry the findings into design without flattening them, the research was synthesised into three personas. These are not generic UX deliverables: each one is an evidence-based composite built directly from the interviews, diaries, survey responses, and mapping sessions, and each encodes a distinct pattern in the data. Aylin carries the experiences of young migrant women navigating xenophobia and night transport; Maria, those of LGBTQ+ migrant women seeking belonging and inclusive social spaces; Sarah, those of local women negotiating harassment in nightlife they refuse to give up. Their quotes are drawn from participants' own words.

The survey makes this concrete: 85% of respondents reported experiencing some form of harassment, aggression, or unwanted behaviour in Dublin, and among those who experienced it, all said they did not formally report it. They turned instead to trusted friends. The survey did not measure why they chose not to report, but the pattern is consistent with broader research on the underreporting of gender-based harassment and on the role of informal support networks. When formal channels go unused, the care network does the institutional work. The question the project carried forward was what design could do for an infrastructure that already works but cannot be seen.

The reframe

That finding forced a conceptual decision that shaped everything downstream.

The dominant paradigm in "women's safety" technology is surveillance: track my location, alert my contacts, map the danger zones. It treats women as subjects to be protected and the city as a threat to be monitored. Danger-zone mapping carries a further problem: it can stigmatize particular neighbourhoods or venues, marking them and the people who live and work there, without addressing the underlying conditions that make situations unsafe in the first place.

The evidence pointed elsewhere. If women's actual safety practice is built on trust networks, mutual support, and knowledge of good places:

Drawing on Bonu's (2018) distinction, the project explicitly opposed the neoliberal notion of security, which preserves the status quo through monitoring and punishment, and adopted safety as radical care: an active, collective practice of agency and communal support.

The conceptual decision that shaped every design choice: rejecting the surveillance paradigm in favour of safety as collective care (after Bonu, 2018).

Three concrete design commitments followed:

1. No blacklists. Unsafe venues are not publicly shamed on a fear map; they are flagged privately to authorities and organisations, and offered practical tools, resources, and training opportunities to change. The aim is not to identify problematic venues but to support behavioural and institutional change. Public visibility rewards good places.
2. No panic buttons as core interaction. The primary gestures are recommending, reviewing, accompanying: acts of care, not alarm.
3. Positive incentives over punishment. Venues are invited to earn recognition, not threatened with exposure.

Designing the ecosystem

Rather than a single product, the response was three interconnected interventions, each making a different piece of the existing care infrastructure visible and actionable. They were conceived through a participatory, research-led process with co-creative moments at key decision points: participants reviewed and redirected design proposals during the second collective mapping round, and field testing reshaped the artefacts themselves.

Institutional layerSafe Space Award

A physical recognition displayed by venues that women collectively identify as safe and inclusive, inspired by sustainability certification models. It converts women's distributed knowledge into a visible incentive: venues compete to earn and keep the recognition, shifting business behaviour citywide without a single punitive mechanism.

Situational layer'Stop It!' interaction card

A football-style red card placed on tables in bars. A woman experiencing harassment can raise it: a non-verbal, non-confrontational signal that makes the situation visible to staff and bystanders without requiring her to escalate verbally against a potentially aggressive response. It is a game mechanic deployed as social protocol, and the in-situ counterpart of the venue intervention: where the award recognises good behaviour, the card gives venues and customers a concrete tool for responding to bad behaviour in the moment.

Community layerShaping HerCity platform

Web and mobile app where women map and review trusted places, find companions for the journey home, by taxi, walking, or public transport, because the cost of safety is also economic, access support resources, and feed the award system with their reviews. The platform is the digital home of the care infrastructure the research found: it gives the WhatsApp group, the trusted friend, and the known-safe café a shared, persistent, searchable form.

Taking it to the street

Most concept projects stop at the prototype. This one went outside, and the street answered back.

The Safe Space Award was deployed in six real locations: two cafés, two bars, and two public spaces, the gates of St. Patrick's Park and Trinity College, over a month and a half. I observed reactions in situ and ran informal interviews with owners, staff, and customers. Staff expressed pride in the recognition; passers-by photographed the signs; the artefact behaved, in miniature, exactly as the incentive logic predicted.

A QR code on each sign linked to a short concept-evaluation survey (Google Forms), asking women whether peer recommendation would change their trust in a venue and gathering reactions to the three proposals. During July, 145 people scanned the code and 52 responded; 30 women (aged 19–49, 60% Irish, 40% from thirteen other nationalities) completed the full survey.

"Very helpful. I frequently feel unsafe in social situations and knowing that a location has been deemed safe by other women would significantly increase the likelihood of me patronizing that place."Concept-evaluation survey respondent

The 'Stop It!' card produced the project's most instructive finding, twice. First, in testing with volunteers in bars, the card spontaneously evolved: used as a coaster, it sits visible on the table from the moment women sit down, sparking curiosity and conversation. The artefact found its better form in the field, not in the studio.

"I think the use of cards as coasters is quite interesting, different and a much friendlier tool to address harassment. Campaigns should be made to introduce their use in bars, I think they could be very useful!"Field-testing participant

Second, and more important: the venues where women had reported negative experiences never responded to contact attempts. The places that most needed the intervention were precisely the ones design alone could not reach. This is not a footnote; it is a research result. Voluntary adoption selects for already-good actors. Reaching the rest requires institutional leverage: local government sponsorship, licensing incentives, NGO partnerships. The project mapped, empirically, the boundary where design ends and governance must begin.

The platform, built on evidence

The digital layer received the same evidence discipline as everything else: every feature traces back to a research finding, and every interface decision was tested before it was kept.

From research findings to product features

The feature set is the research translated into interactions. Women confide in trusted friends, not authorities, so the platform is built on peer reviews and women-verified places rather than official ratings. Taxis home are a recurring safety expense, so Share a Ride and Travel Together connect women heading the same way, by taxi, on foot, or by public transport, addressing safety and its economic cost at once. Safety information already circulates in WhatsApp and Facebook groups, so the platform gives that circulation a persistent, searchable map. And because support needs differ, the app aggregates local resources: women's groups, services, and emergency contacts.

"Knowing where to get to and be safe is critical. I particularly like the ride feature. It would be game changing for me."Concept-evaluation survey respondent

Storyboarding the system before building it

Before any screen was designed, storyboards tested the logic of the whole ecosystem: how a woman encounters the 'Stop It!' card and uses it to address harassment in a bar, and how the Safe Space Award leads her to discover the app, download it, and use the ride-sharing feature to get home safely. The storyboards made the connection between physical artefacts and digital platform explicit, and exposed the moments where the system had to work as one.

How to address bad experiences

How to empower and enhance good experiences

Information architecture and low-fidelity flows

Information architecture came from an open card sort (10 participants): the dendrogram showed 100% of participants grouping the safe-space map with reviews, which became the home screen's organising principle. Paper sketches and low-fidelity wireframe flows defined the navigation skeleton before visual design began.

A/B testing the menu: making size signal priority

Moderated usability testing (7 participants, ages 21–45, remote and in-person, using observation, contextual inquiry, and scenario-based tasks) showed the main menu's concept landing: its unconventional layout intrigued participants enough to explore, and every one of them reached for the map pins to see nearby safe spaces. The problem was hierarchy. Participants were confused by the button sizes: they expected the most important features to be the largest and secondary ones smaller, and the menu didn't honour that expectation, leaving the map and ride features visually level with everything else.

Rather than redesign on intuition, I ran an A/B test of two main-screen layouts. Interface A was a stacked vertical list: one action per row, revealed by scrolling, the familiar pattern users meet in most apps. Interface B exposed every action at once in a grid and used tile size to encode priority, giving the most important features the largest targets. One question: should the main screen lean on a familiar scroll, or show its options at once and rank them by visual weight?

Interface B won on every dimension: 7/7 found it more visually appealing, 7/7 reached the Safe Space Map more easily, 6/7 located features more easily, 6/7 reached the ride-share feature more easily, and 6/7 preferred it overall. The single dissenter preferred A purely for the familiarity of scrolling: a real signal, not noise, and the reason the final design kept scrolling rather than discarding it.

Two principles explain the result. Visual hierarchy: when button size maps to importance, the layout tells users where to look before they read a word, resolving exactly the expectation the usability test had exposed. Recognition over recall: showing all primary actions at once lets users recognise what's available instead of remembering to scroll for it, which matters most in the hurried moments this product is used in. Participants drew the line themselves: scrolling is fine for secondary information, but the main screen should prioritise.

That became the rule for the final design: primary actions exposed and size-ranked on the main screen, secondary options kept in a scrollable menu, not a blanket swap of grid for list. The same session produced two further changes carried into the product: a UX-writing fix, where the ambiguous “Show me / My location” control became “Show my location”, and the driver-gender and vehicle-type controls participants asked for, which became filters in the ride-share flow.

A visual identity designed as an argument

The brand system is part of the product logic, not an isolated branding exercise. The primary colour is a deep violet (#47387A), historically the colour of feminist movements, paired with the rainbow of the LGBTIQ+ collective in the HerCity logo, signalling from the first glance that "women" here explicitly includes bi, lesbian, queer, trans, and non-binary women. The mark doubles as the system's connective tissue: the same rainbow circle appears on the Safe Space sign, the 'Stop It!' card, the app's location pin, and the website, so that a woman who has seen one artefact recognises the others on sight. In a product whose core mechanism is trust, that instant recognition is a functional requirement. Typography (Open Sans / Montserrat) and a warm secondary palette keep the register closer to community and joy than to alarm: a deliberate inversion of the red-and-black visual language of conventional safety products.

High-fidelity flows and key screens

Every screen here serves one goal: helping a woman move through the city with more confidence and less exposure to risk. The flows are shaped by how women already manage safety: trusting peers over institutions, trading tips in group chats, and treating the taxi home as a recurring cost of staying safe, so the product turns those informal practices into shared infrastructure rather than inventing new behaviour.

Search and Review work as one loop: women find places other women have verified as safe and inclusive, by category, name, or address, and leave reviews that feed the same map: a self-reinforcing, women-verified database instead of official ratings. Share a Ride and Share Your Journey answer the night-time gap directly: one connects women heading the same way so they can travel together and split the fare; the other shares a live location with chosen contacts and confirms safe arrival on its own. Together the screens make the case’s central reframe, safety as collective care, into something a person can act on in the moment.

What I learned, where it goes

On evidence and scale. The street test was deliberately small: six locations, a few dozen survey responses, a month and a half. The signals are early, not conclusive, and I treat them as such. What the scale did allow was something most concept work never gets: contact with institutional reality.

On the limits of design. The strongest lesson is systemic. Behaviour change in venues responds to incentives, and incentives need backing. A recognition scheme works for venues that want to be good; it needs public-sector weight to reach the ones that don't. Any serious implementation runs through local government, licensing bodies, and women's organisations, a finding that turned the project's future roadmap into a governance question as much as a design one.

On infrastructures of care. Reframing safety away from surveillance was not an aesthetic choice; it was what the evidence demanded. The project's deepest contribution is the finding it is built on: women's informal care infrastructures, the trust networks, the circulated knowledge, the known-safe places, already do the work that institutions fail to do. Designing for them means making them visible and giving them leverage, not replacing them. That thread, from informal care infrastructures to designed infrastructures of encounter, is the line of work this project opened, and it is what external reviewers responded to.

Next: adapting the model to an Argentine city, testing whether an infrastructure of care designed with Dublin's women can translate across urban cultures, and pursuing the institutional partnerships the Dublin test proved necessary.