SAAS STARTUP B-EVE - 2020
B2C Anti-aggression social app & IoT wearable
Context
I spent 1y working on an efficient solution to stop street aggressions. I started this journey by participating in and winning the James Dyson Award, before launching my SaaS start-up and focusing on finding a great community-based app using extensive testing, interviews & Agile-based methodologies to find a viable solution to this worldwide problem.
What is it?
b-eve is a suite of two products combining the power of both hardware IoT device & companion app social features:
​
-
​IoT Device: This wearable bracelet solves the key problems of triggering anti-aggressions solutions when aggression occurs. By saying a keyword or shaking the wrist, it records the aggressor's voice, calls for help, sends GPS coordinates and triggers a heavy 100dB alarm to scare the aggressor.
​
-
Social Companion app: 90% of street aggressions happen when women are alone. The app solves this problem so that aggressions don't happen in the 1st place, by allowing users to meet up and walk back home together.
How did I validated it?
To build these 2 products, I conducted intensive user research of victims, security experts and policemen. I visited 2 emergency call centres to gather insights on this highly complex topic and understand why all the current solutions fail to solve this problem.
20+
User testings
2.3K
Website visitors
2
Research surveys
2.3K
Website visitors
5
Tested app
30%
Expected Gross M.
2
Tested IoT concept
Dyson
Award Winner 2019
1M views
Video in 48h
b-eve mobile app & IoT Bracelet
PHASE 1 - 3 MTHS
UNDERSTAND.
In this Phase, I defined the vision and market fit that ruled the product throughout the entire process. To do so I analysed the market and ran intensive user interviews to gather victims and police insights.
1. What are we solving for which market?
"Defend and empower women against the worldwide issue of public aggressions"
When starting the project, I knew several friends who had been aggressed in public late at night despite owning anti-aggression solutions such as paper gas. I suspected there was a potential gap in the market worth exploring to try to solve this very complex problem. This led me on a 1 year journey.​
​
Defining a product target market
As showcased by the below numbers, street aggressions & public harassment are worldwide issues that affects massive markets, showcasing a real business potential if a good product fit can be found.
-
India
-
​1st most dangerous country in the world for sexual violence
-
80% of women have faced harassment in public spaces
-
-
Brasil
-
In 2022, 7.4M women suffered physical violence
-
In 2020, 1 in 4 women suffered some type of violence
-
-
USA (source Gallup survey)
-
In 2014, 45% of women did not feel safe walking alone at night​
-
-
UK
-
In 2016, 75% of women living in London UK were victims of harassment or violence in public​​
-
​
I decided to focus on the French market
It was easier for me to gather data from it (interviews, survey..). Therefore I looked for France focused figures (source INED, 2015):
-
58% of women between 20/24yo have been victims of at least 1 violent act in public spaces​
-
1M women are victims of sexual harassment and abuse per year in public spaces
-
Only 13% of sexual harassment and abuse victims in public spaces engage in legal procedure
-
There are 8M women between 15 and 35yo in France, which is the main street violence targetted age range
​
Defining the product focus
The word of 'aggressions' refers to multiple types of aggressions:
-
Domestics violence: These are the most common aggressions-type. They are very hard to find answers to because of their highly private and repetitive nature.
-
Public aggressions: Are more sudden and unpredictable, and targetted to mostly 16/30 yo women in most cases. Defence against them can be quite straightforward compared to the other types of aggression.
-
Workplace violence: Lone women workers interacting with men are victims of multiple and repetitive aggressions (home nurses...)
​
I decided to focus my product on the Public aggression spectrum. It was the one I was the most familiar with, and the one with potentially the most straightforward solutions. However there are multiple overlaps within 3 aggressions-types, and finding solutions for one does solve several problems for the other.
2. What is the competition landscape ?
The safety solutions market is a growing and varied one, with multiple overlapping indirect and adjacent competitors using similar sets of features for very different usage (elderly fall detection or factory workers safety bracelet..).
Find below a list of the most important groups of competition within this self-defence sector.
​
Intensive surveys & interviews helped us gather feedback on each one of these competitors (cf '4.Gather user needs & feedback' below)
'Street safety' mobile apps
Direct Competition
-
Each country have their own apps, either from private businesses, owned by public organization or non-profit privates ones
-
They usually allow to easily share your location on your trip back home, tag a violent event you witnessed, and call for help in case of emergency
-
They are mostly free
Paper spray & 'hard' defense
Direct Competition
-
These products are made to directly 'fight back' against the aggressor
-
These are illegal in some countries and have many shortcomings, such as getting used against the victims
-
They are highly successful & and affordable products
Wearable alert trigger
Direct competition
-
These wearables offer quick access to an 'SOS call' or emergency SMS
-
They are usually dummy button that triggers mobile app features on the user's phone
-
The online price varies from 10£ cheap 'running bracelet' to fancy 100£ jewellery-looking products
Workers accident alert device
Adjacent competition
-
These products are highly reliable technical products made available to professionals working in tough dangerous environments and factories
-
These products are usually not available to the main public
Loud alarm & Flash
Direct Competition
-
110dB alarms are made to scare the aggressors. They are cheap 5/15£ solutions available online
-
Same for the intense bright flashlights made to immobilize them for a short period of time
​
Smartphone OS features
Direct competition
-
Apple, Samsung & Google have multiple safety-related features within their Operating system
-
It proves the large market interest for these types of products & features, but also shows the weakness of any startup in these area.
Elderly alert devices
Adjacent competition
-
These products are bulky technical bracelets, made to be highly reliable
-
They can be either dummy SOS buttons or contain smart fall-detection systems
-
Their price varies from 30£ to 100£ pounds online
3. Who is our product for?
-
'1 in 5 women' Vs '1 in 71 men' are victims of violence at least once in their life > The gender target will therefore be focused on women
-
The 1st women age group suffering the most aggression is the 16/19yo range
It represents a pool of 2M women in France -
The 2nd age group suffering the most aggression is the 20/30yo age range
It represents a pool of 4M people in France
​
Find below the Persona created following the above data.
The combination of Persona 1 & Persona 2 provides a total market size of 6M women in France.
This top-down market approach will have to be filtered down based on average income and location (targeting mostly city users).
A bottom-up market analysis could be done to verify the market size assumptions with more accuracy.
​
Intensive surveys & interviews helped us gather feedback and daily-life needs for each of these Personas (cf '4.Gather user needs & feedback' below)
Persona 1
High need, medium income
Independant Women
-
20 to 35yo active women living in city
-
Use to commute on their own, go out & party
-
They can go home late and on their own
-
Second main target group of public aggressions worldwide
-
They enjoy tech & digital products, and like to feel empowered by technology
-
They have 35K£/y and + income
-
They represent a population of 4 million in France
Indirect Persona 1
Indirect need, high income
Family Carer
-
35 to 50yo active Parents
-
They want to ensure their 15/20yo daughter would stay safe while commuting alone
-
They enjoy tech products and trust it to take care of their children
-
They have high income and are willing to invest in something to keep their children safe
-
They can buy as many products as they have children meeting the Persona 2 requirements: They have a max purchase power of 2 million units
Persona 2
High need, low income
Independent Student
-
16 to 19yo active citizen/student women
-
Use to commute, go out in a city environment most of the time.
-
They can go home late and on their own.
-
The main target group of public aggression worldwide
-
They are mainly using free-based services and social media
-
They have low income & purchase power
-
They represent a population of 2 million in France
Indirect Persona 2
B2B market
Lone workers women in high-risk environments
-
25 to 50yo active women working in risky environment professions
4. Gather user needs & feedback
I ran intensive user research for several months, trying to understand how street aggression occurs.
-
Expert Interviews & visit of Emergency Call Responder centre: In-person visit of 2 emergency call centres at night - interview of 7 call responders & trained response police squad team
-
Open-invite forum talks with student women: Qualitative feedback & debate about multiple competitor solutions and how/where aggressions occur (how the aggressor behaved etc..)
-
Research Surveys: Hundreds of quantitative feedback about competitor product performance, street fears & insecurity
Based on these multiple feedbacks, I divided a 'typical' aggression into a 6-phase process. See below a dumb-down version of it:
Phase 1: Preparing to go out
-
95% of street aggressions occur when the victim is on its own. Therefore the 'pre-going out' is key in preventing the aggression from happening
-
This phase is a source of apprehension for a lot of women that know they will have to walk home on their own
Phase 3: Harrassment
-
Usual behaviours from female victims are to ignore the aggressor, walk away, and sometimes try to call someone
-
Harassment evidence is hard to gather
Phase 5: After-assault
-
Victim can end up in shock lying down on the ground for several minutes, incapable of calling for help or sharing their location
-
Victim can try to run home or to a safe place
-
Several victims can be 'shy' of what happened, and not call for any help afterwards
Phase 2: Walking alone
-
At this stage, most women want to let some people know that they might be in potential danger. They already use multiple techniques to do so, using dedicated apps or text messages
-
Walking alone at night is a stressful moment for most women
-
Women use tricks to limit the risks of aggression (Pretending to be on the phone, strong surrounding consciousness...)
Phase 4: Physical assault
-
The victim usually freezes, making it impossible for her to counter-react
-
The aggressor usually takes hold of the victim, making it impossible for her to reach out to any device
-
Women find it very hard to reach and trigger any of the existing solutions (alarm key-chain, tear gas, quick sos call on the phone..)
-
Goods carried by the victim can be stolen or ripped apart at that stage, making use of the phone or any other protection mainly useless
-
Counter-defense products like tear gas is usually turned back against the victim by the aggressor
-
Police force can take between 5 to 20min to get on site if an emergency call have been made on time
-
Police can only access the location of the phone emitter up to a 10km radius based on the closest antenna used to make the call
-
Gathering evidence of the aggression is hard
Phase 6: Long-term effect
-
Long-term trauma is a recurrent thing for victims
-
Finding groups and people to talk about can be hard
-
If the victim decides to declare the aggression to the police, providing evidence is very difficult
5. Finding the market fit
Once I had gathered the data on competitor solutions and user feedback, I built a Product Positioning map to find a market gap within the current proposition. Here's what I found
-
At the moment of the aggression itself, the accessibility of the solution to repeal the aggressor is key. No product on the market offers an efficient counter-attack product that would also be highly accessible while being strained
-
No Product combines the above with a proper pre & post-aggression service - allowing for GPS tracking, SOS warning and support to the court
-
No product targets the challenge that "95% of aggressions occur to lone women in the street". Allowing women to go home not alone will always be more efficient than any defence system.
​
Here's a dumb-down version of the feature matrix I built:
Phase deliverables
-
'Competitor product breakdown': An exhaustive list of all the direct & indirect competitors, with feature breakdown & user feedback about them (find both online and using my own interview/survey data)
-
'Market fit analyses': A clear analysis of where the product should sit on the market
-
'Aggression-process breakdown': An analysis of the 6 phases of aggression and user needs throughout them
Phase challenges
-
It was challenging to apply the Lean methodology to such a complex & social topic
-
User interviews were hard to run because of the nature of the questions I had to asks about the assaults
-
Visiting an emergency call centre at night time and listening to these calls of profound distress were very hard
PHASE 2 - 2 MTH
IDEATE.
Using all the data, user feedback, business requirements, strategic vision, and competitor analyses gathered previously, I kicked-off a 2 sprints of feature ideations, from blue-sky workshop to an organized & ranked feature set to explore and trial later on.
1. Find features & solutions for each user need
The first step was to use each user need detected in the 6 Phases of aggression and try to find solutions against them. The decision of going hardware, software or both combined was still open at that stage.
​
See below a dumb-down version of it:
3. Combine features into meaningful Groups
Once I had this library of features, the next step was to define how best to interface and provide them to the end user. I therefore separated 2 sets of features out of it
-
Cloud-based app service for prevention
-
Takes full advantage of the software's internet capabilities to build social & community-based features that would work at preventing aggressions from occurring
-
-
IoT device & companion local app for defence​
-
Takes full advantage of the phone capabilities to call for help & record/share GPS location
-
Takes full advantage of the hardware to be the 1st hand-free activated device. It could be triggered under complete immobilization by the aggressor, or even if the phone is stolen
-
​
The detailed feature set of each product (digital & hardware) would be refined and tested later throughout the iterative trial & error AGILE process.
1. Mobile app: Safety features
-
Share the user's live GPS location for close ones to watch over the user going home
-
Quickly send an SOS message with GPS location to the Police force and/or contacts
3. Mobile app: Set-up the IoT device features
-
Set-up the IoT device
​
​
2. Mobile app: Social features
-
​Compare ways home to find the safest
-
Find other people to commute and walk with
-
Get in touch with other victims of aggression
-
Find quick help in your surroundings when being aggressed
-
Find safe places
-
Advice & contact on what to do after an aggression
4. IoT device: Embedded features
-
Hand-free activation of the defence mechanisms
-
Loud 100dB alarm to scare the aggressor
-
Aggression audio recording
Phase deliverables
-
'Feature set matrix': Detailed breakdown of all the features ideas, which data they require, and on which interface (mobile or IoT) they would be most relevant
-
'Feature Groups definition & Prioritization'
-
'Business plan'
Phase challenges
-
Finding entirely new defence methods was very hard, if not impossible. The challenge was to provide them in a way that actually helps the end-user.
PHASE 3 - 1 WEEK
PLAN & ORGANISE.
I now had to define how to develop & test both products: the social app & the IoT device.
1. Prioritization & Planning
Design & validate the IoT solution + mobile app defense solution - 2 months
-
Validate the feature-set​
-
Run Concept validation interviews​
-
-
Validate the form factor & HMI
-
Design & Build multiple hardware mockups​
-
Run usability testing​
-
-
Refine the final MVP product feature set, design & interaction
-
Use 'Shadow button' methodology to validate the final purchase intent
-
-
Size the product manufacturing cost
-
Take a decision on the cost viability
​
Design & validate the Social app solution - 2 months
The mobile app design consisted in cycle through the 3 steps below, until finding a successful concept
-
Iterate solutions & build an MVP wireframes​​
-
Validate the concept
-
Run Fake landing page concept
-
Run Concierge testing
-
PHASE 4 - 4MTH
ITERATE.
This phase was key in building an actual solution for both the IoT device & the social app that came out of the ideation. It is through their validation that I discovered their flaws and had to make the choice between continuing with this entrepreneurial project or not.
1.1 IoT Product - 1st iteration & validation
Using the feature Groups established in the previous Ideation Phase, and after a 2-month design process, I built the below product proposal.
Click on the picture to get a full breakdown of how the product works & what it offers.
Here's the validation process I used to design it, using Lean methodologies.
​​
-
Validate the feature-set​
-
I ran 6 Concept Validation interviews​ with former victims of aggression & 19/22 yo women students, gathering very positive feedback over the bracelet concepts & feature set
-
-
Validate the form factor & HMI
-
I did a design exploration of over 30+ wearable form factor concepts
-
I built a hardware 3D printed mockup & I ran an ergonomy study with 8 participants, testing the form factor ergonomy and easiness of use. The average note was 8.2/10, which showed a clear appeal for the small bracelet concept from my target personas.
-
I built a fully working prototype (alarm, BLE phone connection, shaking & voice recognition) & ran a usability study of the trigger & features with 8 participants. The feedback was positive on the voice activation methods, but the 'arm shaking' activation didn't receive great feedback. However, it did validate my original hypothesis that those trigger methods would allow to call for help even under immobilization by the aggressor.
-
1.2 IoT Product - Award & large-scale validation
I won the 1st prize Dyson Award and 1st French-Region Innovation Award with the initial concept, gaining lots of traction & and positive feedback from multiple anti-aggression charities, victims & media:
​
-
It allowed me to meet the 1st Lady of France, the French Minister of Women's Rights, and multiple companies' CEOs trying to support its industrial creation
​
-
An interview video of myself made over 1M views in 48 hours, with hundreds of valuable comments over the product that allowed for lots of enhancement for the 2nd iteration below
​
-
I entered an incubator which allowed me to plan to attend the CES the following year, giving me 1y to build a prototype and update the original concept (the project got abandoned beforehand)
​
2.1 IoT Product - 2nd iteration & validation
​A more polyvalent feature set based on large-scale user feedback.
Following hundreds of online comments, private messages received & in-person encounters throughout the journey, I enhanced the original proposition to make it a better, more grounded and versatile product.
This concept adds multiple user experience enhancements to ensure the product could work in more situations, and avoid putting the victim under more danger. Here are the main enhancements:
​
-
Shake your wrist turns the bracelet on & off: User feedback mentioned the risks of shaking the previous bracelet unintentionally during the day, creating a risk for unwanted SOS alert
​
-
Stealth & Sound SOS alert modes: Depending on how the product gets triggered (voice or side-button press), the feature alert set will be different, including the 100dB alert or not. User feedback stated the need to avoid attracting attention or annoying the aggressor even more with a siren in some cases
​
-
Haptic feedback: It will let the user know when people are watching over them over the mobile app, and if the product is sending SOS alerts if the Stealth mode is triggered
​
-
Anti-torn off feature: If the product gets torn off when being 'turned on', it will initiate a countdown before automatically going into 'SOS mode' and alert the police
​
-
Anti-turn-off trigger pattern: Ability to create a custom pattern to turn off the product once an 'SOS mode' is activated - making it harder for aggressors to turn off the on-going alert
2.2 IoT Product - Mobile app set-up
Part of this 2nd iteration was to include a companion mobile app to set up & personalise the IoT device experience.
-
Pick who should receive the SOS message: Decide who should receive the alert message
-
Customise the 'On/Off' mode: Turning On the product (by shaking it) can automatically trigger a 'Follow me' GPS sharing with contacts
-
Customise the alert triggered in each mode
-
Access tracking History
-
Training mode to allow users to train themselves using the bracelet
-
Retrieve the audio recording of the aggression (could be encrypted content that only the Police could access)
2.3 IoT Product - Technical requirements & costing
Once my technical requirements & feature-set well documented and written, I contacted a dozen of French & European R&D IoT consultancy to get a quote & cost evaluation of this 2nd iteration.
​
-
Developing & manufacturing a prototype of the product would cost around 250K€
-
Mass manufacturing of the product would require 300K€ to build the tooling and reach 1M€ to launch a small-scale product of 1000 units
-
Multiple additional costs to take into consideration:
-
Certifications CE/IP 15K€
-
BLE technology 15K$ yearly fee
-
Worldwide Patent 50K$ fee (+ 10K€ for the expert patent writing)
-
Shipment & storage cost
-
-
Estimated planning:
-
6 months to build a 1st POC​
-
+ 3/6 months for 2nd version POC
-
+ 1y to get the first 1000 units manufactured & shipped
-
Total of 2/2.5y until 1st product shipped to customer
-
-
Estimated product manufacturing cost of 70€/unit for 1000 units produced
3. Social app - 1st concept
Find people to go home safely together
The Social app work was a full-Agile project of its own, going through a 2 months iteration process testing 5 different concepts. I ran 2 types of MVP validation testing:
​
-
Concierge testing: Using the online popularity of the project, I would contact women who would fit my target Persona and age range & run a 30min interview to gather their feedback on the concept and Figma prototype I would run in front of them
​
-
Fake Landing page: I created 2 fake landing pages to advertise on Facebook and measure the engagement of both concept
1st version: Danger heat map + journey finder + find people around you
This concept concentrated multiple ideas in one to allow for efficient user feedback and detection of what would stick out of it. Here are the main features:
-
See live map of aggressions surrounding you
-
Way-finder of best route home to avoid 'dangerous' places
-
Find, chat and commute home with other community members
​
User feedback was mainly positive for the 3rd feature, which I decided to keep exploring in the following version.
2nd version: Plan safe trips with community member
It's the equivalent of the 'Book Uber ride in advance' feature in the Uber app. It creates a future event in which several people will meet up at a specific location to commute together safely.
​
User feedback was positive about it (6 out of 9 said being interested in using such app). The main feedback was about the pain of having to plan the journey ahead. The next concept tries fixing that.
3rd version: Find people around you now to share uber cost
It creates match between people within an area who want to go to the same location at the same time, and potentially share a Uber altogether.
​
User feedback was very negative regarding the obvious safety issue of meeting potential random & malicious people through this system.
There are no easy ways to solve such problem, and that contributed to putting an end to this series of App concept research.
4. Social app - 2nd concept
Meet people to go on event together
I tried to explore helping people meet up and go out more, but do so in a safe environment. This new approach was more optimistic and opened a new market of 'event planning' type of features.
1st version: Map-based Event finder & planner
This all-in-one event planner proposal concept offers these features:
-
Map-based event finder
-
Find other people nearby going to the same public event, and plan the way in/out
-
Plan your own Private event & plan how your friends will go home ahead
​
User feedback was positive on the idea of meeting new people to get to events together, which the next concept would focus on.
2nd version: Tinder-style swipe to find events partners nearby
'Tinder like' concept of swiping left/right people interested in the same event.
​
User feedback wasn't positive enough to keep exploring this route. Only 2/9 participants wanted to try the app which shows a clear disinterest over the concept.
5. Iteration Phase detected blockers
Here is a recap of this iteration phase's discovered blockers, allowed thanks to the quick trial & error of the Lean process followed throughout these 4 months.
​
-
Defense IoT Product: Too expensive & margin too thin​
-
The overall wearable concept and feature-set received great feedback, and would have been worth developing further if the manufacturing cost were not so high.
-
Prototype manufacture cost too high (250K€)
-
The estimated 30% gross margin wasn't enough to sustain a startup growth​
-
-
Funding hardware without a prototype was hard
-
Crowdfunding was too expensive (2€ spent in marketing for 1€ back)
-
Business Angels wouldn't invest without a fully working POC
-
-
I struggled to gather valuable evidence of end-user purchase intent, versus simple product and 'brand' hype (fake landing page unclear results)
-
Some features had too many legal risks (audio evidence recording wouldn't always be legal etc..)​
-
​
-
Social mobile app: Couldn't find an attractive concept
-
Making people meet to go home together does solve the core problem (95% of aggressions occur over lone women), however, it does create too many new issues, which ends up being counter-productive
-
After 5 app concept iterations, I couldn't find a successful concept, and the user testing feedback weren't good enough to justify any more work
-
Phase deliverables
-
'Technical IoT requirements'
-
'IoT Product proposal breakdown'
-
'Mobile app proposal breakdown'
-
'b-eve Startup website'
Phase challenges
Applying the Lean trial & error methodology to both software app & hardware IoT device simultaneously was highly challenging but taught me a lot about project management & the core values of Agile methodology, allowing me to fail fast.
