CLIENT PROJECT

AVA 

PROJECT DETAILS

Duration: 2.5 week sprint

Team size: 4

My role: UX designer

Date: February 2020

SOFTWARE USED

Sketch

Affinity designer

Adobe Photoshop

HOW THEY FELT

  • People felt very overwhelmed and unsure

  • A time of stress with little structure and support

HOW WE HELPED

  • Supportive, knowledgable language that isn't patronising.

  • Always educating users as they go - so they feel in control and in the know

  • Helping them uncover and learn about their strengths

MY PERSONAL INPUT

  • Part facilitation of design studio

  • Organising project files

  • Drawing out wireframes designed and ideated by the team

  • Part of leading the project process

  • High-fidelity designs

  • Photography to document our process

PROJECT OVERVIEW

Discover

 

User research
GAINING INSIGHT INTO
THE PROBLEM

Competitive analysis
LOOKING AT WHAT COMPETITORS DO WITH
CV GENERATING TOOLS AND CHATBOTS

Affinity mapping
COLLATING OUR FINDINGS INTO KEY TRENDS WE CAN
TAKE FORWARD

Define

 

The Persona
CREATING A TYPICAL PERSONA BASED ON THE INFO GATHERED AT RESEARCH STAGES

Problem statement
DEFINING THE PROBLEM THAT OUR PERSONA SARA
IS FACING, SO THAT WE CAN USE THIS TO DESIGN SOLUTIONS FOR HER

Develop

 

Design studio
CREATING QUICK, FAST IDEAS ABOUT SOLUTIONS TO SARA'S PROBLEM

Paper prototyping
A QUICK AND SIMPLE WAY OF GETTING IDEAS IN FRONT OF USERS TO TEST HOW THEY WORK THROUGH THE USER JOURNEY

Mid-fidelity testing
CREATING A BASIC CLICKABLE PROTOTYPE FOR USERS TO TEST IN A MORE REALISTIC SCENARIO, WHILST ENSURING WE CAN STILL WORK QUICKLY

Deliver

 

Branding
CREATING A QUICK STYLE GUIDE TO IMPLEMENT ON THE DESIGNS GOING FORWARD

High-fidelity mockup
USING OUR BRAND TO FINALISE THE LOOK AND FEEL OF THE AVA WEBSITE

8 MINUTE READ

Brief

Discover

The challenge for this sprint was to design the process for a CV chatbot called AVA. AVA was an idea that came from the client’s existing company, Sidekicks, where recruiters work with first and second job seekers, to build them a CV that is fit for employers and also passes ATS checks.

GAINING INSIGHT INTO THE PROBLEM

User research

We began the project with a kick off meeting with the client. We introduced ourselves, asked questions and outlined the way the project would run. 

 

We quickly learned that the client was very passionate about helping people. Hardly any of her candidates knew about ATS (Automated Tracking System); a system that scans people’s CVs, looking for keywords and phrases, and rejects them if they don’t meet certain criteria or aren't in word format. She met lots of talented, capable candidates ready for a role, who weren’t getting replies from applications. Whilst guiding people to format their CVs correctly, she realised a bot might be a more efficient, accessible way to help. She provided us with a database of questions that a chatbot needed to ask, in order to get exactly the right information out of people.

 

We wanted to reach out to a large target user group in order to find out about how they were feeling. We did this in two ways:

  • We created a survey that featured questions based around our key points and assumptions from the briefing meeting, in order to validate the features and theories that were already in the client’s mind. We also gained extra insight into the users experience of applying for jobs, creating CVs and job searching in general.

  • We asked a similar set of questions in face to face user interviews too, so that we could get a more detailed understanding of how people were feeling.

LOOKING AT WHAT COMPETITORS DO WITH CV GENERATING TOOLS AND CHATBOTS

Competitive analysis

It was important that we look into competitors and what they were doing. We looked at both companies who provided CV generator tools (some suggested by client, some by users, and others from our research) and companies who had effective chatbots.

 

There were some key takeaways from this competitor research that were important in combination with our user research, which I will cover in the section below.

COLLATING OUR FINDINGS INTO KEY TRENDS WE CAN
TAKE FORWARD

Affinity mapping

We collated the key findings from all of our research in order to identify trends that we could look into going forward.

Initially we had a large number of trends that appeared in the affinity map. We discussed these and noted the key findings, but then decided to take a few sections and continue to arrange them further. These were:

  • Tone of voice - we had a clear picture of what people expected from recruiters (friendly, professional and knowledgeable) and who they seeked advice from when they were applying for jobs (older siblings and family members). We took this forward in order to inform how AVA might communicate with users.

  • Chatbots - People expected chatbots to be “Friendly, personalised, quick.” Intercom, Do not pay and Plum were all examples of companies that inspired us - great onboarding, helpful hints and suggestions of what to type next and clear information on where you are throughout the process.

  • CV Content - “I know what I’ve done but don’t know the right thing to say.” Users didn’t know what employers were actually looking for. The client met people who captained sports teams for example, who didn’t realise this showed leadership skills. Babylon Health had a clever way of asking people the right questions in order to get the information they need from them. We felt that we could learn from this to create AVA.

  • Account creation\login - The client didn't want the user to have to log in; showing she was trustworthy, and would not sell their Data. Our research highlighted a lot of confusion about this. “I would expect to have a login if I’m paying for
    the CV.”

    We felt it was important to test this further, so we created a quick test that we carried out with 9 users. We asked them to read a quick summary of what AVA does, and the information they would need to input. We added two buttons; 'create an account and continue', and 'continue without creating an account'. Without hesitation, 8 out of 9 users went straight to the create an account button. When asked why, they explained that if they were inputting data about themselves, they would want to know that it was in an account that they could access, prove they’d made payments to, and potentially delete in future.

Define

CREATING A TYPICAL PERSONA BASED ON THE INFO GATHERED AT RESEARCH STAGES

The Persona

Through findings from user interviews, we created a typical user that we could keep in mind moving forwards. Sara is fresh out of university and looking for a new job as a PA (a typical situation suggested by the client).

DEFINING THE PROBLEM THAT OUR PERSONA SARA IS FACING, SO THAT WE CAN USE THIS TO DESIGN SOLUTIONS FOR HER

Problem statement

With Sara in mind, we then began to think about her current situation. Using a storyboard, we could visualise how our solution might help Sara in her journey. This led us to then produce a problem statement, which sums up what Sara is struggling with, and provides us with a direction when designing our solution.

 

Problem: 

Sara needs a way to build a professional CV that will pass an ATS check, because she struggles with understanding what to put on her current CV, and she doesn’t receive responses  when applying for roles.

 

This problem focuses on some of the key points gathered from the user research:

  • Users are unaware of ATS

  • Users don’t understand what to put on their CV

Develop

CREATING QUICK, FAST IDEAS ABOUT SOLUTIONS TO
SARA'S PROBLEM

Design studio

Moving into the development stage, we met again with the client. We ran a design studio in which we took turns in facilitating. We wanted to bring everyone together, and get all our ideas down so that we could then discuss what we would take forward. After some entertaining warm up sketching sessions, we began to ideate around ways we could solve Sara’s problem, both using a chatbot but also trying to be as free as possible with ideas. We used dot voting to narrow down to some key features that we thought were particularly helpful in solving Sara’s problem.

Back at the studio, we took the highest voted features, and arranged them into a feature prioritisation map, in order to work out which features we’d need to include in our MVP (the simplest version of the product that solves Sara’s problem.)

A QUICK AND SIMPLE WAY OF GETTING IDEAS IN FRONT OF USERS TO TEST HOW THEY WORK THROUGH THE USER JOURNEY

Paper prototyping

We began to develop some initial paper prototype wireframes of Sara’s potential user journey through the website, utilising the prioritised features from the design studio. 

We tested this wireframe journey with 9 users, asking them to:

 

  • Read the homepage and create an account

  • Chat with AVA

  • Pay and download your CV

 

We had some really interesting feedback at this early stage, and we implemented some changes during the process.

 

Examples of the kind of changes we made are as follows:

After the chat with AVA, some users wanted to review what they'd written before they moved on. We added a button to take them to a transcript of the conversation.

The language used confused users at this point - they didn't know what this step was. They were also confused by the cost showing up too early, so we removed it.

CREATING A BASIC CLICKABLE PROTOTYPE FOR USERS TO TEST IN A MORE REALISTIC SCENARIO, WHILST ENSURING WE CAN STILL WORK QUICKLY

Mid-fidelity testing

After making iterations to the paper prototypes, we then moved into Mid-fidelity (a greyscale clickable prototype) in order to test users on screen. We tested with 11 users at this stage, and again made changes based on their feedback and reactions.

An example of changes is as follows:

Users skipped right over the crucial information on the homepage, and then became confused later on because of this. We significantly reduced the amount of copy, and this made the account CTA more visible too.

Deliver

CREATING A QUICK STYLE GUIDE TO IMPLEMENT ON THE DESIGNS GOING FORWARD

Branding

Alongside existing styles, we developed the brand slightly, reflecting our research on tone of voice and recruiter expectations. We created a colour palette based around the existing AVA blue, and Lato typeface, and we introduced curved edges and a new illustration style that helped to reflect the more friendly, human side that users expected from a figure like AVA. 

USING OUR BRAND TO FINALISE THE LOOK AND FEEL OF THE AVA WEBSITE

High-fidelity mockup

We implemented our brand onto a final clickable prototype. We tested this on 10 users, and had some really encouraging feedback: “I found the whole experience very easy & calming” and “Very professional and clear layout”. We created a video walkthrough, showing Sara’s journey through the website, beginning with reviewing the homepage and creating an account, through some examples of the chat with AVA and how the communication might look, and finally to Sara’s choice of template and payment & download. 

A selection of high-fidelity wireframes in Sketch

We ended the project with a presentation of our journey through the two week sprint to the client. They were thrilled with the outcome, and very excited about moving forward with AVA. We are very proud of our work as a team, and we can’t wait to see what the client does with our solution! 

High-fidelity walkthrough of Sara's journey.

Insights

I thoroughly enjoyed this project. I learned a huge amount, both in terms of soft skills, and the UX process itself:

  • Process - We established great ways of working as a team from the outset, and this meant that we could focus all of our energy on producing the best work possible.

  • Keeping the client in the loop - We outlined the project with the client early on - this helped us keep on track, and helped them feel involved through the process. They were a valuable part of the journey, and invested in all stages of the project.

  • Adaptability - we were provided throughout the course with a toolkit of UX methodologies. It was up to us in this project to choose which tools to use, and when. This was a great challenge, and helped us stay in control.

  • Design studio - this was a really positive and productive experience. We learned the value of warming up and doing fun tasks initially, so that when it became more serious, everyone was feeling comfortable and confident. I really enjoyed seeing the variety of ideas during the studio, and it reinforced to me the power of teamwork.

  • Practise! - We structured our project efficiently, which gave us time and space to practise our presentation. This was new for me, as I usually worry about over-rehearsing, but I was proved wrong; rehearsing helped to solidify the information in my head so that I could bring more energy to the presentation, and focus on how the audience was feeling/reacting as I spoke.

What next?

Going forward, we made a few suggestions about what the client could look at next:

  • We did some further prototyping and testing around a new scenario, in which the user could go back into their account and change their CV template.

  • We looked into the possible integration of Sidekicks (the client’s other company) and how we might be able to introduce their job listings as a final stage of the process, after the user has downloaded their CV.

  • We also considered an optional personality test at the beginning of the process, so that unsure users could learn a bit more about themselves and the kind of jobs that would suit them before they began to chat with AVA.

  • LinkedIn