A global search for Welcome to the Jungle

Anne Moreno
12 min readNov 20, 2020

Welcome to the Jungle is a french media and tech company launched in 2015. The company is present in Europe : Paris, Barcelona, Prague, Bratislava. At the start, its aim was to create the first media on employment and today it is the reference in that field.

For them, job-hunting is a jungle!

Their goals are :

  • to build the best candidate experience possible
  • to facilitate quality matches between candidate and recruiters, based on culture and aspirations

Basically, their main goal is to guide and support users at every stage of their professional life.

How ?
They offer an innovative job board (visual & modern) with more than 3 000 detailed company profiles and a wide variety of high-quality content (articles, videos, series, podcasts, magazines or events) to help young professionals get a job and evolve.

Welcome to the Jungle proposes also other products to professionals (WTTJ Originals, WTTJ Home, WTTJ Event, WTTJ kit) that won’t be detailed in this case study.

This project was carried out in two weeks remotely with the support of two members of Welcome to the Jungle, Head of design & Product manager.

The brief

On the website, there are three main pillars : media, explore companies and find a job. Currently, users can only search in each section (it’s called a scoped search). It is useful because it has the potential to help users find what they are looking for faster.¹ But sometimes, it can lead to frustration if users have to switch to one section to another to find the answer they’re looking for.

In this context, Welcome to the Jungle sent me the following brief :

« To enable users to search through all the types of resources available on our site, regardless of where they are located on the site. »

The idea is to propose a global search.

The kick-off

Set up of the meeting

I planned a 1 hour kick-off meeting in remote with the team of the project to :

  • understand the organization within Welcome to the Jungle
  • discover the scope, opportunities & constraints of the project
  • learn more about users’s feedback
  • discuss about the progress of their work on this topic

For that, I prepared a mural template of our kick-off meeting with a set of questions, my UX strategy planning (in progress) and some exercises (to do with the client) as a user journey map, for example. At the beginning of the session, I shared the access to the whole team.

Along the session, I fulfilled the whiteboard with the information given by the client and allowed them to participate and bring their own ideas.

kick off meeting on Mural

What I have learned

From this kick-off meeting, I got responses about :

Users of Welcome to the Jungle
From student, young workers to senior workers, Welcome to the Jungle has a large panel of users. This is the reason why I didn’t create a persona for this project because it didn’t provide something valuable for the design process and the client.

User flows
There are two main user flows on the website :

  • users who follow Welcome to the Jungle on social media or/and newsletter and read articles from there and after that, they quit the website.
  • users who are job seekers and go mainly on the « find a job » section.

Secondary research

I understood that the brief did not come from a specific need or frustration of users but it was more part of the idea of improving the user experience of their website.

For this challenge, I needed to better understand users in order to implement this new feature in a relevant & valuable way for them.

Where to begin ?
Firstly, I created a survey to collect quantitative data about users and their behavior. It allowed me to confirm/invalidate my assumptions and those of the client.

Survey

To be sure that I ask the right questions, I needed to know what data I need to collect and from who. For that, I made a lean survey canvas.

Lean survey canvas (Mural template)

The data I need to collect
The survey should be short and clear (around 10/15 questions). I asked general questions first (who they are, what device they use) and finished with more specific questions :

  • what are they looking for
  • if they search any information related to their job search (as tips, articles, job description)
  • what kind of related content precisely

The testers
I sent it to Welcome to the Jungle users to help me understand them better and confirm/invalidate the assumptions of the clients.

survey : 127 responses

The survey confirmed all the assumptions of the clients but also highlighted the fact that users search also for content related to job search on different platforms and the fact that they search mainly on their desktop.

Interviews (qualitative data)

In parallel of this survey, I conducted also 5 user interviews with users of Welcome to the Jungle and non-users to understand their behavior, needs and frustrations regarding searches for job offers & topics about work.

Two types of questions
For Welcome to the Jungle users, I asked them to show me their path on the website, what they do exactly and why. For non-users, I asked them what kind of job board they use and why.

This schema illustrates the two main user flows on the website :

Two user flows for two different state

I gathered some powerful insights regarding to the research of media content:

Conclusion
This research shows us that there is no transversality on the website and users don’t search actively for other content than job offers on Welcome to the Jungle.

In parallel of these interviews, I took a look at the existing website.

Audit of the website

Sitemap (as is)

Sitemap as is

As you can see, the website has a lot of subsections, especially in the “media” section where they propose a lot of relevant content for each stage of the user’s life. This varies depending on the language though.

I discovered two main pain points, highlighted also by users :

  • some content are difficult to find (the collections for example)
  • some pages are very long

« Sometimes I forget what I was looking for, after scrolling for thousand years ».

Heuristic evaluation

Then, I assessed the usability of the website following the ten principles.
To do that, I used this useful tool : a Trello board that enables you to do an heuristic evaluation easily following the Nielsen’s heuristics.

To sum up, the search bar of the website don’t :

  • prevent user errors (slip type mistakes) (Heuristic 5)
  • provide explanation when you don’t get any results (Heuristic 9)
  • offer search suggestion or option to reformulate your search (Heuristic 5 & 6)
  • offer recent searches (drop-down menu)

Conclusion UX part

The audit of the website and these interviews confirmed many insights from the client. To sum up, Welcome to the Jungle has a rich and wide variety of content but users need to be pro active to find it & read it. But, except for jobs, they don’t search actively. They just came across interesting content by chance.

This statement has to be taken into account for the next step of the design process.

Problem statement

Based on my research, the problem statement is the following :

How might we offer for job seekers relevant and personalized content that would help them find a job and evolve without them having to look for it ?

Ideation

I conducted a one hour ideation session with four different members of the squad to really get a variety of response : the Product manager, the Head of design, one Front-end dev and one Back-end dev.

With the problem statement in mind, we did one crazy 8’s about the implementation of the global search.

For the crazy 8’s, the idea is to generate lots of ideas in a short period. We draw 8 ideas per minute, i.e 40 sketches in total. After that, we shared our sketches and explained it to the others participants. A dot voting was held and each participant selected 3 best ideas.

Ideation session on mural

For the new feature, the main ideas that emerged are the following :

  • to offer personalized content (depending on your status : job seeker, internship)
  • to propose all types of content (from each section) & the possibility to switch between the different types (with a tab)
  • to help users refine their search (with tags, related searches)

Mid-fi prototype

With these ideas in mind, I made a first version of the solution for a desktop version : a hand-drawn low-fi wireframe and then a mid-fi prototype in Sketch.

The idea was to propose a new search bar on the upper-right corner of the homepage. A tooltip will notify users the presence of the new feature.

Homepage with the new feature

When you click on the search bar, it expands. You enter your search and when you press enter, it directs you to the results page where you got all the results from the whole website sorted by category. You can switch from one section to another to get different types of results (articles, event, jobs).

Global search : results page

Usability test

I conducted 5 usability testing with Welcome to the Jungle users.

Why users of Welcome to the Jungle ?
They are already familiar with the interface, so the usability test won’t highlight frustration or pain points out of scope.

I asked them two things :

  • to search information about «UX design »
  • to apply for a job in UX design in Paris

Results

I got powerful insights from users and collected them in an affinity diagram. I sorted the notes by screen and here are the most relevant pain points :

On the homepage :

  • the representation of the new feature : the magnifying-glass icon is difficult to understand & to find (it’s an icon without a box or a label)
  • its location : it’s not strategic and valuable to put it near the parameters.

On the results page :

Filters
Users were a little bit confused. They didn’t understand these filters below the search bar especially :

  • their functionality : “are they filters or historic search ? how were they selected ?”
  • their state : if they were applied or no
  • choice and customization by users : “how can we add other filters ?”

Tabs
The tab allows users to see the results sorted by type but users pointed some elements :

  • lack of visibility (they didn’t see it at first and then get confused because it recalls the existing pillars of the website)
  • to replace the first section (by default) : « job » by « all category »

AHA Moment (The discovery)

At the end of the usability testing, I made a debrief with users and asked them some questions about the new feature and their experience.

Overall, users like this new possibility offered on the homepage :

« The global search opens up perspectives that I would not have imagined at the beginning of my search ».

But the usability testing highlights one big problem : users don’t see the point of using this new feature when they want to find a job. This is problematic when you know that the majority of users are job seekers. This new user flow takes more time and you can’t precise your search as you can do with the advanced search in the job section. And actually, when they search for jobs opportunity, they are task-focused and they want to get what they want as fast as possible.

These elements made me realize that I had to take a step back from the original brief.

Two opposite behaviors

Indeed, when you are a job seeker, you have a very precise search in mind. You apply some filters and know what you want to search. On Welcome to the Jungle’s website, you go directly on the « find a job » section and you get what you want in a quick and efficient way. On the other side, when you are in a passive search, you want to have a broad search to open up your perspectives.

How can we propose one search that will satisfy & respond to two completely different expectations ? This is contradictory!

That’s why I decided to separate the job section from the global search with connections between the two.

UI part

Design system

How make the wireframe come alive ?

Welcome to the Jungle created a new visual identity in June, 2020. They changed their logo and they displays now a new color palette with yellow as a lead brand color (instead of the green). The color palette and the two typefaces (Work Sans & Welcome) give a contemporary and dynamic mood.

I created my high-fi prototype based on this new design system and followed their guidelines.

Extract of the design system

To do that, I listed all the atoms, molecules & organisms that I would need to reproduce the different user flows. Then, I reproduced them as symbols & nested symbols in Sketch. And lastly, I just added some elements to the existing components (drop-down menu, checkboxes, cards for events).

I also tried to reproduce all the micro interactions that the user encounter on his path as hovering over cards or highlighting.

High-fi prototype

I made a first version of the high-fidelity prototype, iterating from the previous usability testings.

The main differences with the mid-fi prototype are :

  • the location & representation of the global search : the icon has been replaced by a button (label+icon) located near the other pillars at the upper right corner. Also, when you click on it, a full-width search bar appears and hides the other sections. It tells users that it is a complete different experience from the scoped searches.
  • A drop-down menu : it allows users to go directly on specific section from this global search bar.
  • « All category » : by default, on the results page, you arrive on the « all category » section with all the results (articles, events, jobs, companies).

I tested the V1 with 3 users of Welcome to the Jungle.

Homepage & results page

Results

From this usability testing, I gathered all the insights in an affinity diagram :

On the homepage :

The drop-down menu

  • the first sentence confused the users, they didn’t know what to do next, if they have to type again the search.
  • its relevance : this drop-down menu allows users to go directly on the other sections as Media, Find a job. But this is contradictory to propose this possibility (the search bar hides these sections so it is a non sense).

On the results page :

Tabs

  • lack of visibility : users didn’t see it at first
  • redundancy : between the labels of the tab and the titles below

« Discover all » button

  • lack of visibility
  • its meaning : users didn’t see that the button was linked to the title on the left (principle of proximity).

From these usability testings, I iterated with the insights collected and proposed a second version of the solution.

Solution

Check out the final version of the two main user flows :

  • « Global search » path
  • « Find a job » path

Next steps

For the next weeks, it would have been interesting to :

  • usability tests of the V2
  • meeting with the squad to discuss about the different versions
  • to think about the error state : for example, proposing another screen when there is 0 results founds with two buttons : one that proposes to reformulate the search and another button that links to global search

For the next quarter :

  • set up an Onboarding process (to offer even more relevant content at every step of the user’s professional life)
  • improve the filters by incrementing Boolean search

Key learnings

One thing : think outside the box!

At the beginning of the project, I tried so hard to put everything together and to identify a valuable solution for users with the exact brief in mind. But the solution was not fairly strong. So, I needed to take a step back and to propose something different based on all the materials collected with user interviews and usability testing.

Thanks for reading and thanks to the team of Welcome to the Jungle for this interesting challenge. If you have any questions or comments, don’t hesitate to reach me!

[1] : Katie Sherwin, Nielsen Norman Group.(January 18, 2015). Scoped Search: Dangerous, but Sometimes Useful https://www.nngroup.com/articles/scoped-search/

--

--