The effects of the pandemic have cost hundreds of thousands of jobs, with many more still at risk. Redundancies have especially been a problem in parts of the private sector where unionisation rates are very low. As a result, most workers facing uncertainty at work won’t have a union to turn to. And unions don’t have channels to reach them.
There may be an opportunity for unions to reach out to non-union workers and provide valuable advice content to them. In doing this we could give these workers an initial experience and understanding of unions, learn about their needs, and find other useful ways to engage with them.
At the Digital Lab, we set out to test the use of digital advertising channels to find workers facing uncertainty and sign them up to receive advice content from the TUC. We tried two different advice formats – a webinar with experts and a digital publication. For both content formats, users had to register with the site, allowing us to deliver the content to them and giving us a chance to offer them onward engagement.
We used Google and Facebook ads to recruit targeted users to this set up, hoping to attract workers who had no existing connection to the TUC or unions.
In doing this we had three objectives:
- Find out how to deliver useful advice to non-members that we did not yet have contact channels to.
- Position unions as experts on your side to workers who have no experience of them to date.
- Learn more about engagement with non-union workers and test viability of developing a channel and data that could lead on to organising.
The infrastructure
As this was a pilot project only, we didn’t want to develop any more than we needed to in terms of functionality and design. The website was built as a stand-alone mini site using the basic site-builder service Squarespace.com, and running on a separate domain to the TUC. We followed a methodology developed from an earlier Digital Lab project, to structure the content to appeal to different audiences.
We embedded a Typeform into the site for the survey. This was structured with logic jumps to keep all questions shown to users as relevant as possible (between 9 and 11 questions).
We integrated submissions from Typeform into Mailchimp, to do email communications (using a separate account from the TUC’s main account to reduce data complexity). We also integrated submissions via Zapier into the CallHub platform to handle SMS messaging.
Depending on users’ answers, they were automatically registered on an email journey, an SMS journey, or both. They received an automatic confirmation of their registration by email (including a request to share it with colleagues) or SMS, and a scheduled announcement by the same medium before the event started.
The webinar was delivered using Zoom webinar, using an external supplier to provide speech-to-text captions.
After the event we sent follow up communications by email and SMS, with links to the archive video of their session, a downloadable handout doc, and a link to Typeform feedback survey, offering a prize draw incentive for completion.
As well as the webinar test, we tested offering a downloadable document using the same set up, though this time only using email for registrations, and sending users the download link in an automatic email response.
For both, we used Facebook ads and Google search ads to drive traffic, making use of Facebook pixel and Google tracking codes to gauge conversion rates.
Given this project was seeking to communicate with new audiences, and use some new providers, we did a Data Protection Impact Assessment before starting, and we had to log a minor revision to the TUC privacy policy to reflect new data processing suppliers.
Results from the webinar test
We offered two live webinar session options, a weekday daytime slot and a weekday evening. For the first week, our signup rate was initially very low – over 90% were dropping out at the first question (“which event are you registering for?”). We revised this to introduce the third option of a recording if they couldn’t make the dates. We also moved the form from a separate registration page onto the home page itself, so people could see the three options immediately. These two measures increased signup rate by around 4.5 times.
Most registrations used an email address, though the SMS route was still significant, as was the number wanting both options. The SMS journey was simple and inexpensive to offer through CallHub, so it was worthwhile to add into the options.
73% of users registering gave consent to be told about further advice materials. This suggests that offering perceived value to them is a realistic way to build an ongoing engagement channel.
Who were the users?
We focused particularly around three audiences on Facebook. A group of all-ages retail workers, one of all-ages hospitality workers, and a group of younger workers (20-35) from across the private sector.
We had hoped to be able to reach non-union members and this was largely correct, with 72% telling us they weren’t a union member.
On the registration form we asked optional fields on industry and region, as well as an open field for more detail on where people worked, leaving it up to them to decide what this meant. Some gave us an employer, a job title, a more defined sector, or a town location (there’s more detail on how we structured and revised the forms and questions here).
Questions on region and industry were answered by 95% of people registering, and the employer/job detail free question was answered by 82%.
Crucially 32% of respondents gave us the name of their employer. Of this we could identify a few potential groups forming, either where a large employer is likely to be having problems, or also where people may have been spreading the word amongst their colleagues. The largest employer group was 15 people from the same employer.
This was really encouraging. It suggests that it is possible to get useful data to segment workers for organising, if they are offered something of value and an explanation of why the data is needed. Coupling this kind of activity with a larger promotional spend or publicity campaign could help develop a cluster of workers to engage with on collective solutions as well as providing individual advice.
Users were overwhelmingly using mobile when they found us (84% of users were on a phone). This is really worth considering in structuring landing pages and advice resources, as well as the signup process. Asking for more information than we did might have made it harder to register these mobile users.
Ads:
Ads on Facebook cost us about 78p per landing page view overall. Facebook pixel suggested 1 in 17 people made it through from landing to sign-up since the changes we made. That gives an average acquisition cost of around £13 per user actually signing up. That cost was lower for our audiences of retail workers and hospitality workers – and higher for the audience of younger workers.
Google ads cost us 61p per clickthrough once we had optimised them around Google’s initial feedback. Google’s stated cost per conversion was much higher than Facebook at around £20 per person. Even though these people were actively searching on redundancy topics they were less likely to convert. As a hypothesis, it might be that searching for something means you want an answer immediately, and so you may be less inclined to wait when you find out a webinar is days or even weeks away.
We also believe these costs are high in part because we were bidding against employment lawyers who hoped to make a lot of money out of each of their conversions. Rates like this are obviously too high for us to consider viable at large scale on a project without a defined return on investment, such as union membership. But it points to something that could be worthwhile in a collective campaign with a way to involve them more in a deeper engagement going forward.
Webinar:
TUC Education did a brilliant job running a webinar, working with Labour Research Department employment rights advisor Nerys Owen as legal expert presenter. The webinar was delivered in the style of a Q&A, with a mix of planned questions and genuine audience questions.
Many of the webinar viewers came online expecting individual advice about a workplace issue they had, and it was important for us to offer an element of interactivity. On the first webinar, we had more questions posted that we had viewers, so there is an obvious need for this type of guidance. The content we offered was very valuable, but there was clearly still a need for more.
The evening session was less popular than the daytime one, also with a greater drop off in terms of numbers joining. This chimes with a separate project we ran in 2020, where a late evening session did less well. Perhaps as unpredictable family life made people make a last-minute choice not to join, or just forget.
Results from the advice guide test
A few months after the webinar trial, we ran this test again, using instead a new published resource on redundancy and change at work that the TUC had developed. We wanted to see if people would view a packaged document as being of similar value to screen time with an expert.
This time, Facebook ads to download the guide performed broadly at the same cost as for the webinar, but Google ads were significantly cheaper at 39p per clickthrough. It seems that the offer of a downloadable resource resonated better with people looking for an answer from online search.
The conversion rate was also improved from the webinar, in terms of the number completing the form to access the content. On average, conversions here cost £7.50. Again this is expensive for offering free advice, but it may be much closer to a viable return on investment for recruitment.
However we noticed significantly fewer people agreeing to further contact from us – only 42%. Our hypothesis is that users found the content proposition less convincing so were less likely to commit to receiving more of the same at the point of sign up.
Feedback
Feedback from our follow up surveys in both tests was positive (average 4/5 rating for being worth their while engaging), though it was a small sample and very few users offered feedback, despite the prize draw incentive.
Users were interested in receiving content on wider topics ongoing, from other immediate problems such as Covid-safety to more careers focused issues such as the possibility of retraining. Unions will have a lot of experience to be able to offer onward content, providing it can be refocused to feel relevant from the target audience’s perspective.
We found that those who responded fell into two similarly sized groups between those who faced an immediate problem with redundancy, and those who were concerned about the future of their employer longer term and were doing their research.
This is obviously a difficult situation for unions to offer follow up practical assistance to non-members with an existing problem. And as such it limits the potential for useful follow up work in organising and recruitment.
However there may be two qualifiers here. We did demonstrate that it was possible to find clusters of workers with the same employer, and this may offer an avenue for unions in supporting a group of relevant workers rather than isolated individuals. Also there is a difference between those who have an existing problem and those who have a concern that they may collectively experience a problem soon. The latter may be easier for an organiser to support in achieving a result for the group.
Learning points
Positive:
- It is possible for unions to find relevant users online if we have the right product for them, even if we don’t yet have any prior contact or brand awareness.
- If users have a logical explanation for the request and perceive value in what’s offered, they will be willing to provide useful data with potential for organising engagement.
- If users perceive potential ongoing value, they will be willing to subscribe and enter into a longer-term content relationship with a union. The more valuable the initial offer, the more likely users are to view an ongoing relationship as valuable to them.
- Live events are perceived as high value content but are much less convenient to access than downloads, so ultimately reached fewer users.
- Gathering larger numbers of people around problems they are aware they may be about to face at work could be a viable way to generate clusters of interest that would make engagement more viable for unions.
Challenges:
- More work would need to be done to establish viable return on investment to justify advertising to a significant volume of users this way. So far we only know how much it costs to start engaging with these workers, not to ultimately recruit any into membership.
- “Webinars” were maybe not a concept understood or valued by many users (possibly why we had a high proportion of office-based workers registering despite our targeting). Research with users on how we position a webinar with them could be useful.
- Clusters of users generated don’t always match initial targets (eg despite targeting for retail and hospitality, we acquired a significant cluster in a financial services company). Unions may need to be open to working with each other to better support such workers.