Helpful

Helpful’s first firebreak: what we learned

During June 2018, we had our first Gov.uk inspired ‘firebreak’, where we paused non-essential project work for 2-3 weeks so the build team could focus on their own individual projects; including researching new ideas & technologies, organising neglected repositories, remodelling dated templates and writing helpful guides and documentation. Urgent requests from clients were prioritised over firebreak activities.

Why we did it

There had been a very busy period of build work that meant certain, non urgent, jobs were being left at the wayside so we could focus on project work. We knew we wanted to work on our project processes and had been discussing a starter ‘toolkit’ (blog about this coming soon!) but we didn’t have any free time to actually act on them. By enforcing a firebreak and sheltering the build team from non essential build work, we were able to free up time to have these discussions and put in a plan of action.

How we did it

We had to plan our firebreak in advance to ensure we weren’t working on or taking on any new projects during the 3 weeks. We let our clients know that we would only be available for urgent request during these weeks and scheduled any non-urgent work for the beginning of July.

To keep track of what we were working on, we created a Trello card where every member of the team listed the tasks they wanted to work on, in priority order, to check off as we progressed. To ensure there was transparency and no overlapping of work, we checked in every morning via Slack to explain what we we had worked on the day before and would be working on that day. The team was free to work on what they wanted during these 3 weeks as long as it was clear what we were working on and how it would be of benefit.

Team away day

During the first week, the build team went to Brighton to work on our build processes and starter toolkit ideas, facilitated by a friend of Helpful, Stephen Hale (@hmshale). The idea was to remove ourselves from the office to ensure we weren’t distracted and to have a fun evening in sunny Brighton to let off some steam. 

The afternoon of day 1 was spent brainstorming and sketching the most common website components used in our builds and discussing best practice for each to ensure we were left with a list of key components and a set of rules for each, to speed up our build process and ensure we are thinking about best practice and encouraging our clients to do the same.

After fish and chips on the beach for dinner, the morning of day 2 was all about our project processes and how we could improve them. Mapping out on a table each step of the process and any pain points, we were left with a comprehensive project guide which we will be turning in to a visual map for our clients.

Show and tell

At the end of the 3 weeks, we held a show and tell session to show each other and the rest of the business what we had been working on. Each member of the build team showed examples of what they had been doing and what benefit it has to the business.

Work completed in the firebreak included:

  • new functionality such as an automated site mapping tool and unbreakable responsive navigation
  • implementing version controlled deployment of WordPress sites
  • client Google Analytics dashboard templates
  • spring cleaning of Zendesk, Trello and Dropbox
  • digital contract signing
  • the move to a whole new time tracking system

What we’d do differently

This was our first firebreak and something we will definitely be repeating in the future. Unsurprisingly, the main issue was around project work creeping in and taking time away from our ambitious list of firebreak tasks. This will probably always be the case as clients come first and unexpected urgent work will always crop up but if we are having these firebreaks more frequently, our list of tasks will begin to shrink.

We chose late June as it is usually a quiet time for clients with holidays and there was an obvious gap in our project planning however this also coincided with Helpful team member holidays which became tricky to manoeuvre around so we had to extend the firebreak from 2 to 3 weeks to accommodate. Some more forward planning, potentially even starting to plan our next firebreak now, would make this easier.

Although we have our starter toolkit components and improved project processes, we still need to build these in to tangible products to get real use out of them. Project work is kicking off again and we are struggling to find the time to complete these. We are trying to set aside time to complete these now but we should have ensured that any firebreak work was finished in the allocated time which we will need to plan better for next time.

Testing… testing…

On the Digital Action Plan we like to challenge participants who are already confident producing content across a wide range of channels. One of the ways we do this is by asking them to test and improve their content.

Here are a selection of examples which demonstrate the value of regular testing and improvement.

Netflix: emotional images generate most engagement

Variations in the images used to promote Orange is the New Black on Netflix. Note how over three seasons the focus has shifted from an ensemble cast to the main character
Variations in the images used to promote Orange is the New Black on Netflix. Note how over three seasons the focus has shifted from an ensemble cast to the main character

Netflix is well-known for its data-driven approach to decision-making. It recently conducted a study to see what difference the choice of a film or TV programme’s artwork made on viewing figures. Through their work they discovered:

  • Images matters. Images constituted over 82% of people’s focus while browsing
  • Three’s a crowd. Images with fewer than three people in them were more more likely to get people to click that images with more people, such as the ensemble cast of Orange is the New Black
  • Emotion is everything. Images which clearly demonstrate emotion, such as close ups of people’s faces, generated more clicks than images where people’s expressions were harder to distinguish

You can read more about the findings on Social Media Today.

Organ donor register: finding the right call to action

Variations on the DVLA's call to action, encouraging people to sign up to organ donor register
Variations on the DVLA’s call to action, encouraging people to sign up to organ donor register

The stakes are high on when it comes to an organ donor campaign. Testing to improve the performance of a campaign ultimately means more donors and more lives saved.

Working with the DVLA, the Cabinet Office Behavioural Insights Unit and GOV.UK, the NHS Organ Register Service tested 8 different ‘call to action’ messages to encourage people to register as an organ donor, immediately after they had renewed their tax disc.

The service aren’t sure why one variation was more successful than the others but they extrapolated that using the most successful message would lead to around 100,000 extra donor sign ups over a 12 month period. You can read a really detailed case study on GOV.UK.

40 shades of blue

Google results page showing results for Google blue links
Could Google’s famous blue links soon be a thing of the past?

Google tested 40 shades of blue before settling on the right one to use for its Gmail advertising. Going to this trouble in a $200 million dollar advertising boost.

Right now, Google is currently testing black links on the main Google site. So far, it has faced some fairly negative feedback on Twitter. The ultimate test will be whether black links encourage more clicks and therefore more revenue. Will it make more people click, bring in more revenue and if so, will Google implement it?

What can you test?

What are you working on at the moment? Could you tweak and test variations of your content? It doesn’t have to be on the scale of these examples. It could be as simple as sending the same tweet at two different times of day, posting the same message with different images, or using different hooks to drive traffic to your website.

Facebook insights, Twitter analytics, and Google analytics will all help you to judge the response so you can compare, contrast, and improve your results. Let us know how you get on @helpfuldigital