How Anchor Trust Leverages Chat Bots to Drive Recruiting - HROS

HROS Hero_Bolt_Black_Zazzle.png

The Anchor Group is England’s largest not-for-profit housing association, providing housing, care and support to people over 55 years old.

Our initiative was to create a better recruitment process, while providing the best applicant experience possible.




We noticed a pattern emerging during the recruitment follow up process, with a high number of individuals that completed application forms then failing to engage with our in-house recruitment team despite attempts to reach them via phone and email.

Candidates’ busy lifestyle meant that they were often unable to respond to emails or answer the phone (except at very specific times of the day) and tended to shy away from completing extensive application forms. We also realised that there were highly-skilled candidates that were experienced in care but unable to translate this onto paper in the form of a CV.




The solution was a purpose-built recruitment Chatbot built into Facebook Messenger, with questions that will allow us to screen and engage with candidates instantly. 1.2 billion people currently use Facebook Messenger, and can access it straight from the mobile phone they keep in their pocket. This made it the perfect tool to build the new technology upon.

When someone clicks on an ad on Facebook, a conversation will immediately open in Messenger, allowing the applicant to be preliminarily qualified and book in a specific time for a call with an Anchor recruiter.




Using the Chatbot, we initiated automated one-on-one conversations with potential candidates. There are a variety of platforms available to build Chatbots, however we chose to build ours within FlowXO. The reason for this is because it has allowed us to hard code in our own features and integrations, so that it can be much more than a simple automated question and answer system.

The best example of this, is a feature whereby once the applicant has inputted their postcode, the Chatbot can then calculate how far they are from the nearest Anchor care home and then estimated commute times. Our ‘natural language’ Chatbot responses also ensured that we retained a personal feel.


In less than a minute, our recruitment Chatbot can:

  • Establish which role the candidate is interested in

  • Whether the location is within commutable distance

  • Whether the candidate has the required experience

  • Capture multiple data points such as name, email address and postcode


Facebook Messenger is also becoming a platform for customer service, so it was important this was considered when building the Chatbot. This resulted in building in functionality so that if the user had a specific question, they could be passed along to a member of our recruitment team while still on Messenger.

The technology alone, however, was not enough however to ensure success - the initiative involved building targeted social media campaigns that drove people to the Chatbot where they could then apply.

An example of our Facebook ads

An example of our Facebook ads

The great things about Facebook ads are that we have data on everything - click-throughs, engagement, applications, etc. It allowed us to A/B test and optimize our messaging. Through analysing several months’ worth of data, we were able to conclude what messages would appeal to our target audience the most. By delivering the right message to the right audience, we were able to drive effective traffic to the Chatbot to maximise applications.



  • Since launching the chatbot, we’ve attributed 86 hires directly to candidates who initiated contact with Anchor Trust through the bot.

  • The “traditional” recruitment method applying for a job via uploading a CV to a website yielded a conversion rate of 2.04%. The conversion rate for the Chatbot stands at an unprecedented 27.35%.

  • 1,062 total Chatbot conversations with potential applicants. All those who do not finish their application can be “retargeted” via our social campaigns to re-engage them.

  • The average cost per applicant for previous recruitment campaigns was £51. The Chatbot has reduced the cost/applicant by 64% which now stands at £18

  • The number of average monthly applicants has increased 82%




The main consideration should be the development time for building a Chatbot. Once the software has been built there is also continual process of analysing the data to see how it can be improved. If we were to create another Chatbot, we would allow for more time to test it before it goes live. There is always a risk with new technologies that it won’t work as flawlessly as hoped, so you can never test too much!

For example, over the course of the development of the Chatbot, we have learnt that the language used is vital in guiding the applicant through the process and keeping them engaged. We know that emojis are popular online, so we incorporated them into the Chatbot along with our informal language and GIF’s to make the application process as fun and engaging as possible.



  • Simplifying the application process has significantly increased the number of applicants, while decreasing the cost per applicant.

  • Data is king – continually analysing results and data has allowed us to improve and evolve the Chatbot to make the experience even better.

  • Simply building a Chatbot is not enough – how are people going to discover it? You need an accompanying online advertising campaign to drive the people you want to apply onto the Chatbot.

  • Anticipate the time and cost that goes into developing new technologies.



  •  FlowXO

  •  Google Analyitcs

  •  Facebook / Instagram Ads