Mission

We want creating your own resources to be easy.

Wordwall.net allows teachers to create interactive games and printed materials for their students. Teachers simply enter the content they want and we automate the rest.

History

It all began in a secondary school classroom in London (UK) in 2006.

For generations teachers have stuck laminated words on a wall to support literacy exercises. Whilst working as a teacher one of our founders created a program where you could simply type a list of words for the same purpose. He added a random picker with a spinning noise like a TV game show. This was the first version of Wordwall.

2006

The Wordwall idea was born

2008

Wordwall sold bundled with IWBs

2010

Created our own ARS handsets

2013

Choice of graphical themes

2016

Wordwall.net was launched

2020

1 million daily visitors

2021

100k paid subscribers

A handful of other templates followed and that first template became known as Flip tiles. It can still be found in the product today. But the next real innovation came when we realised that there was a deeper connection between the templates that was emerging. A set of keywords used to make Flip tiles could just have easily been typed into a Wordsearch. The content a teacher typed in once could be reused in multiple ways. With this idea, the Switch template feature was born and has been a core feature ever since.

A few months later our two founders, Josh and Ben, took this fledgling product and started a company. Wordwall evolved into a tool that supported the high-tech classroom. To use it you needed either an interactive whiteboard (IWB) or an audience response system (ARS). It grew to a range of dozens of templates and received a warm reception. But there was a problem. High tech solutions were fun to play with and created initial excitement, but too many teachers didn't have the funding to buy such things. Hardware was often hard to manage for large groups, requiring too much technical expertise to be trusted. For many years Wordwall was a niche product with only a small fan following.

The company changed direction and in 2016 Wordwall.net website was launched. The concept was simple. Enable all teachers to create and share resources wherever they are in the world. This new product was made possible by the new HTML5 web standards that were for the first time meaning interactive content could be realised on any device. Gone were the installed software and reliance on expensive hardware. Making the product easy to use so that people could discover it for themselves became our number one priority.

Technology

Let us tell you how it all works

If you are considering joining us to be a developer or other technical role, you'll want to know more about our stack.

We take a pragmatic approach. Combining code that we've been using for nearly 20 years, with innovations from the bleeding edge. We develop our own unique technologies when we need to and build on widely available solutions when we don't.

Web application
The site is run using ASP.NET with the MVC pattern. Pages are generated using Razor. The server code is in C# .NET. This allows smooth integration with the game-engine which is also in C#.

Game engine
At the core of the product is the game engine. This is our own technology that we've been building since the early days of the product. It is composed of two key elements. The game logic and the view.

The game logic describes each template. And for each template it lists a set of reusable units that describes how it should work. These units might share common jobs such as marking, drag-and-drop, toolbars and menus. The same unit might crop up in many different templates. This system makes developing new templates faster and ensures a consistency between each template so each feels familiar to the user.

The game logic describes each template in an abstract sense and dictates how they behave. But it's the view's job to visualise it for the user.

The view draws sprites over a layered parallax background. It utilises the Canvas and Audio APIs found in every browser. We use XML mark-up to describe sets of graphics, sounds and animation sequences that we call themes. The same game screen can be made to look very differently by switching the theme. It's all fully responsive and will make sensible layout choices at all screen sizes.

Both game logic and view are written in C#. But they use a x-compiling technique to create two identical versions. One that can run natively on our servers and during automated testing, and one that can be delivered to the browser as a JavaScript file.

Hosting
Our product is hosted by Microsoft Azure. We have a scalable web-service that deploys a varying number of instances across the day and can automatically respond to peaks in demand.

The site's resources and all our users' files are hosted in cloud-based blob storage. We use scalable cloud databases in both SQL and NoSQL formats. We use Redis to provide our cache.

We ensure all our static resources are served up through a global CDN network.

Analytics
Making decisions based on data is very important to us. We collect and examine thousands of feedback comments every week. These are all read and collated.

We are always running A/B split test experiments on every aspect of the site. Every significant change to the user interface is verified with an experiment. Our experimental system is deeply embedded into our product.

We also collect anonymous data on how every template and theme is used, creating internal league tables of performance. These tables directs our efforts into which need attention and tells us what to make more of.

Deployment and testing
Our continuous deployment system is central to the smooth running of the product. Every bug fix or new feature is human reviewed and then runs against a suite of thousands of tests. The automated process culminates in a set of servers which can then be switched over with a single click.

We have a large range of unit tests which are developed in parallel with the code they cover. On top of this we have web-driver tests that cover all the features of the product by test-driving a real browser. This ensures that breaking the live site becomes a rare event.

Testing our templates presents a challenge when you consider the enormous range of content people might enter and the pathways they might take in using them. To solve this we use automated bots that play them thousands of times with different content and making different choices. We have other tests that use automated rules to evaluate the appearance of resources checking for issues and then applying it against huge numbers of examples.

Artificial intelligence
When a human designs a teaching resource they make a series of intelligent decisions to maximise its usability. We have to find ways of codifying these decision making processes.

We employ AI strategies in a diverse range of tasks from deciding how to layout items on the screen to figuring out how to best switch one template into another. Our mission is to make life easy for the teacher and to do this we need to find ways of solving tasks that traditionally rely on human judgement. Like an automated chess player we use computational power to try out more combinations than a human could, and we leverage our large data set of resources already made to train our systems.

Strategy

Our product is shaped by our principles

Wordwall.net was born 8 years ago and before that we had a 10 year history building PC software. We've enjoyed many successes and suffered many failures. From that experience we've developed a set of strategies and principles for how we now develop our product.

Focus on improving the user experience

We are building a world-class product. So we must demand excellence in our work and continuously strive to improve the quality. But we understand there are many different kinds of quality. The quality we strive for is not that which matters most to us personally, to our peers, to industry experts or to our competitors; but that which matters to our users.

Learn fast and embrace failures

We prioritise learning because our greatest risk is building what our users do not want or need. We measure everything we can and adjust our approach when we fail. We never assume something is working unless we have evidence. We progress with the small steps that help us learn at each stage, whether they succeed or not. We run experiments both to discover new insights and to validate our work. When new facts arise, we quickly change our minds.

Always work on what matters

We always want to work on what is most likely to create a better product for our users. The impact of the most important changes can outweigh thousands of less important ones. So we don't work on low importance jobs, whilst higher importance work is left undone. We are comfortable with saying "no", because we know it frees us up for what matters more. We don't work heroically towards artificial deadlines and will happily abandon projects that overrun or no longer seem important.

Use research to discover problems, use creativity and expertise to invent solutions

We can't use our expertise to tell us what problems our users are having. We must not take feedback from our users as direct instructions on what to build. We must instead perform skilful research to understand our users' needs and behaviours. It is the job of this research to uncover important problems and the job of our designers and developers to invent creative solutions that solve those problems. A solution must not be judged on its own terms, but on how well it solves the problem and how easily it does so.

Understand your own biases and use process to mitigate them

We know our viewpoint is very different to our users. We'll often get attached to our own ideas and opinions. It is easy to cherry-pick evidence that isn't representative or see patterns that aren't real. An issue we just stumbled across can seem far more important than it really is. We defend ourselves from these biases with processes that demand we use evidence to develop empathy with large numbers of users, that prevent us acting rashly, and that ask us to consider many possible viewpoints and solutions.

Team

Together even though we're apart

We’ve got a diverse team of about 25 people, all working to make Wordwall amazing. There are customer service gurus, artists and designers, all kinds of programmers, project managers and more.

The company was founded in the UK, but we’re now a fully remote team with people in 12 different countries.

Miguel’s experience of working
at Wordwall

Judi’s experience of working
at Wordwall

Jobs

Careers at Wordwall

We are a small company with big ambitions. Our user base is growing rapidly with low churn and we’re already in profit. We’ve got a product that works and we’re building for the long-term. We want people who want to stick around and grow with us. Join us and help make learning more fun!

Find out more or apply now

About

Mission

We want creating your own resources to be easy.

71,902,407

35

43

History

It all began in a secondary school classroom in London (UK) in 2006.

Technology

Let us tell you how it all works

Strategy

Our product is shaped by our principles

Team

Together even though we're apart

Jobs

Careers at Wordwall