Episode Transcript
[00:00:03] Speaker A: Define the business case and make an impact. So rather than trying to tackle 10 of your people processes and build an AI framework around it, pick the one where there's the biggest return on investment.
And we know our people, our team members in store, their interactions with the customer or the quality of the coffee they make is integral to the success of our company.
Getting the recruitment right, getting the right behaviors is definitely where we put the business case for the AI change.
[00:00:32] Speaker B: Welcome to a new episode of the what Works podcast series. Cathy and Dari sits down with David Hughes, global talent Director at Costa Coffee, about their journey using AI to reinvent how they hire Frontline workers. David shares how Costa is finding the right people among hundreds of thousands of applicants by looking for behavioral fit instead of resumes. And why getting candidate experience right matters more than ever. Let's get to it.
David, welcome to the what Works podcast.
[00:01:01] Speaker A: Hello. It's great to be here. Thanks for inviting me.
[00:01:03] Speaker B: Thanks for talking with me. I'm really looking forward to exploring with you what you're doing with AI and supporting Frontline hiring. But before we get into all of that, tell us a little bit more about yourself and your company.
[00:01:16] Speaker A: So I'm David Hughes. I'm the global talent director for Costa Coffee. I'm Costa Coffee, or a global coffee company operating across 50 markets. Now our home markets here in the UK, over 2,000 stores and employ about 16,000 people. So we are part of the Coca Cola company. They acquired us from Whitbread a few years ago and they've been helping us on our journey to expand into new markets and new territories. And it's a very exciting growth journey and one that we're pleased to be part of.
[00:01:46] Speaker B: Love it. I used to live in London and I always loved Costa Coffees when it was there. Every time I go back to the uk, I always go there. So really good brand and really good coffee as well. So being such a big retail company, of course there's lots of challenges. I mean, you have high turnover, right? You probably have to hire lots and lots of people. What's your talent challenges in in this segment?
[00:02:09] Speaker A: So if I talk specifically about Frontline. So for us it's baristas, we hire about 5,000 people per year on of which we, we probably get about 250,000 applicants. So we've got a high volume hiring and probably one of the biggest employees in the in the uk, which my team, my team look after. I think with, with any sector like this, whether it's retail or hospitality, you've got to expect Turnover, you know, and most companies, most big brands, will sit between 30 and 60% turnover. So there's constant talent challenges, constant turnover, constant and constant need to hire brilliant talent. Then what we've learned over recent years is, is to remove the focus on hiring top talent and much more into relevant talent. So I think it's a bit of a myth when you talk about let's go and hire the best of the best in the market. That may be true, but how does that fit into your company, your company's morals, values and culture? We needed to do through assessment is find relevancy and find fit to firstly interact better with our customers and secondly build careers here at Costa, I love that.
[00:03:15] Speaker B: So it's really not about the best talent, but the best fit for you. I think there's a really good way of thinking about it because sometimes we hear, oh, we have to hire the best and the brightest and all of that. But top talent is really depends on who you are as a company and what your values are, what your company is.
[00:03:32] Speaker A: Well, I think what's worth mentioning is especially in a service driven environment where you're interacting with customers and consumers. You know, customers want a friendly face, especially the community stores, which they may be going to most days and therefore turnover is really important to maintain and to control.
We can't always hire the best of the best of the best that maybe not fit to our culture or our values. We need to hire people that live and breathe Costa as a culture and can interact with the consumers day after day after day because that will drive mps which drive sales and just build, build a better business for the customers and for the team.
[00:04:10] Speaker B: That's so key to think about the frontline, even specifically to not just hire bodies, right? Not just hire people that can come in and do the job because they're really carrying your brand. I mean, they are your brand ambassadors, they're closest to the customer.
[00:04:24] Speaker A: I think when we talk about this more about the behavior. Fit is everything for us at the minute. So hiring on them, softer skills and communication and teamwork, et cetera, is much more important for us than just one's experience and cv. It is worth noting and it would apply to many of our competitors for the frontline roles.
Candidates that are in the market for vacancies may apply for seven or eight roles all at the same time. It's different at the most senior level in the head office, where maybe you're targeting one particular role that you think is right for you. In Frontline, you may look at three or Four different businesses or even sectors and apply for a lot of them. So be able to narrow that field down to people that are right for you. It is hard. This is where AI comes into play.
[00:05:09] Speaker B: So how do you know what's right for you? What is the right fit for you? What are the behaviors that are really right for your company?
[00:05:17] Speaker A: So I think it's a combination. So if I start with our values, you know, we're proud of our values. They've evolved a little bit over the last couple of years, but for all intents purposes, they're very similar. And they're part of our culture, part of our DNA. It's part of who we are. And we're trying to build, build that and maintain it as we, as we grow, especially globally. But secondary, we have to look at the behaviors which we know works for a frontline team member. So how do they interact with the customer? How do they interact with one another? Because in any store there's maybe three, four, 10 people. They have to work together like a really effective unit. And that's born out of behaviors, not about experience. I think that how they work as a team is the kind of success rate of that store. Behaviors over experience every time and also over skills.
[00:06:01] Speaker B: Right. Because technical skills are not that important. Right. You can probably train them pretty easily on how to make the coffee, I would imagine. And you probably prefer that anyway because even if they've worked in a different gene, you probably are a different brand. It's probably different process. Right. So probably the skills are not that important either, I would imagine.
[00:06:21] Speaker A: I mean, I don't say I'm biased, but, but, but cost to make the best coffee in the world. It's not easy to be a barista. I should say that. You know, we put them through amazing training to, to learn to make these coffees flat whites, cortados, cappuccinos, they go through a lot of tr.
Make the very best. But they're skills you can teach. Whether it takes you a month or six months, you'll get there them softer skills, morals, teamwork, communication, they're so, so important. They're not things you can train overnight. A lot of the time you either got it or you haven't. And that's what we will look for when we're sifting through, you know, great talent.
[00:06:53] Speaker B: So how do you sift through all that talent? So you have 250,000 to 300 applications a year and obviously you probably don't have that many recruiters. How do you do that?
[00:07:03] Speaker A: If I talk a Bit about the journey we've been on and I've been at Costa for five years now and when I came in we had no assessment. So it was purely based down on a CV review and then a hiring manager or store manager interview. So littered with inconsistencies and biases. We then brought in another, sorry, our first assessment provider and that focus a lot more on situational judgment tests. So in this situation, what would you do? And actually for the early stages that was helpful for us because we went from no assessment to a more robust assessment. We found over the years though is again, we allowed a bit more bias and inconsistency to creep in and it focused a bit more on the skills than the experience and the behaviors which we needed. Therefore, we went to market to look for a new innovative and AI tool and thankfully we came across Sepia Australian company who's working with some amazing brands already and we wanted to be one of the first to bring them to the uk. So, so we went on a long journey with them to do the, to do the research and to build out some competency models and chat based interviews and we are super confident with our end product. Now every single one of them, 250,000 applicants, goes through the SAPIA chat interview and it's the easiest conversation you'll ever have. It's five simple questions which you can either text, type or talk into your phone and then the magic happens behind the scenes.
We've, we've mapped in our competencies and our behaviors into the SAPIEN process and it pulls from the AI and the algorithms to kick out a score which essentially is a fit with COSTA score. It's not what's your cv, what's your experience, it's how do you fit the costa? And that's been the game changer for us.
[00:08:42] Speaker B: So basically the, every single One of these 250,000 to 300,000 people get a chat based kind of assessment, right. And takes what, 15, 20 minutes, 30 minutes?
[00:08:55] Speaker A: Not even that, not even that. 10 minutes.
[00:08:56] Speaker B: 10 minutes. Okay, so 10 minutes they go through that. It's probably not like tedious, it's probably nice to do. It probably also speaks different languages I'd imagine too. Right. I mean if it's, if it's AI based.
[00:09:10] Speaker A: So simplicity is one of the key bits here. So as I said, five, five questions you can text into your phone, you can talk into your phone. I think about 60% of people do it on mobile, so it's on the go and it's super easy and they're Quite generic questions, you know, what do you think you will bring to Costa? It's the easiest open question you can ask, but where they go with that? We plugged all of our behaviors into. We want to see people talk about collaboration and multitasking and teamwork. So the AI does the, the, the solution in the background that you need an army of recruiters to, to sift through and it's just not, it's not possible.
[00:09:49] Speaker B: Absolutely. I mean that the volume that you can accommodate, but then also the candidate experience. Right. Because the candidate get feedback basically. They get how they performed on that and how good the fit was. And it's not a good or bad thing about them. It's just about fit with the cost though, right?
[00:10:04] Speaker A: Yeah. So candidate experience actually was one of the key business cases for change here. Yeah. We wanted to make sure that everyone went through a fair, consistent and innovative experience. I think part of the reason for this is of all the applicants we get, nearly all of them will also enjoy a latte or a cappuccino. At Costa, there are consumers as well. We have to give them the right experience. Now one of the beauties of working with AI is what you can give back to the, to the candidates. So every single one of them applicants gets a personalized insight report about how they did, what are their strengths and what are their opportunities. In a world where people are applying for jobs and getting nothing back. And we hear it all over LinkedIn and social media about being ghosted in the Costa process. They're going to bespoke two page reports. Here's how you did, here's where we think your strengths are, here's some opportunities. So giving them something back almost. Thanks for taking the time or 10 minutes to apply for it. And in return we'll give you this. And it's gone down incredibly well. The MPS of the minute on the interview is 9 out of 10. 9 positive responses to out of every 10. It's amazing. Assessments naturally are. That was a bit tedious. That took a while. What we're hearing is that was simple and wow. I got something back as well.
[00:11:21] Speaker B: Wow. And it's amazing because most of the people didn't get the job and still they said this was a good experience.
I enjoyed it, it was simple and I got something back and it was worth my time probably. Right. Because they invest 10 minutes of time and then they get, get a two page summary back. So people always like learning about themselves as well.
[00:11:41] Speaker A: Indeed. There's definitely something about self exploration, especially in Frontline where maybe this is a first job or so they're applying for, so they're not used to getting feedback. The addition on this, I think, is worth noting is here in the uk, a lot of the baristas we employ would fall in the Gen Z camp. And in between 18 and 24, we were looked at. Now, that's a generation which are much more comfortable with AI and technology than my generation would have been. If anything, they expect it. So they're going through these processes and they're looking for, where's the innovation, where's the digital agenda? Because this is their first interaction with Costa as a brand. We want it to be a positive one. We want them to go, wow, Costa are embracing the use of AI. That positive sentiment is felt throughout the whole experience.
[00:12:24] Speaker B: What was difficult about this whole journey was anything that was challenging or difficult.
[00:12:29] Speaker A: So I think twofold. I mean, firstly, it's the first time we really embraced AI and brought it into Costa. There was a lot of people that needed to be involved in that and for all the right reasons.
So we worked with our talent teams, the legal teams, the tax teams. Anyone that had a vested interest in making sure we went through the right process was part of the project. The data teams, input input in particular. So we took a while longer than we thought actually to go from start to finish and first conversation with Tapia to implementation. It took longer because we hadn't prepared ourselves properly to align all the stakeholders we needed. We know for next time what a process should look like the first time you do this. My advice is just be patient and have a little bit more, more time on your side. Secondly, it was the adoption. So whilst we built this really efficient and innovative process with candidates are saying they love going through store managers who do a lot of the back end of the process to the face to face is new for them and we have to get them on the journey of why are we bringing AI in? What does it give you? What does it give the candidate? And you don't flick a light switch and see this play through. You've got to just take your time and go through the stages. And I use this word again, but patience is really important right now. Not everybody adopts AI at the same pace, but being patient and respectful of people's journeys and change curves is super important.
[00:13:51] Speaker B: So, David, this is what you just said that really reminded me of a couple of things. So the first thing was it was very easy for the candidates, you said, because they are actually the younger generation actually looking for you to innovate. And to use the AI tools. But the manager needed a little bit more support. They needed a little bit more opportunity also to voice their concerns probably on all that. So that change aspect, really, really important. The other thing that I also thought was interesting is you said this was the first time that Costa Coffee was using AI overall. Is that right? Or is it in the HR area?
[00:14:27] Speaker A: No, in the HR talent space.
[00:14:29] Speaker B: In the HR area. Okay. So you probably used it in other areas as well, but in hr. Yeah.
And HR I think is always challenging, of course, because you have really high stakes kind of interactions. Right. If somebody gets a job or doesn't get a job, it's a big consideration.
[00:14:46] Speaker A: So I think it's a bit of a, a strange time at the minute because if you go to a lot of networking events or you talk to your peers straight away, they want to know, you know, what's your AI roadmap for the future or what are you doing the next few years? And it's so easy to look at your entire end to end process and think, well, let's AI this, this, this and this. What I've learned is AI has to be purposeful to find a business case and make an impact. So rather than trying to tackle 10 of your people processes and build an AI framework around it, pick the one where there's the biggest return investment.
And we know our people, our team members in store, their interactions with the customer or the quality of the coffee they make is integral to the success of our company. Getting the recruitment right, getting the right behaviors is definitely where we put the business case for the AI change.
[00:15:37] Speaker B: I love that we call this falling in love with a problem, not falling in love with the solution. Right. Because it's easy to be an amateur with all the AI tools and they can do anything really.
[00:15:48] Speaker A: It's a lot of the time people see AI and they just go to it's efficiency, you know, it's quicker and easier. What we're trying to say is it can give us a bit above that.
You would need an army of recruiters to try and assess 250,000 behaviors and values and morals. We can do that. With AI, it's not about effectiveness, it's about quality. And more important than that, I should stress this, it's about inclusion.
Everybody gets the same experience, the same report, the same opportunity, and we're confident with the right tool.
It's a game changer.
[00:16:21] Speaker B: Yeah, no, absolutely.
What are some of the.
Any kind of metrics or any kind of results you can share right now?
[00:16:30] Speaker A: So I Think it's a bit early at the minute to talk about turnover. We'll need to look at that after 12 months rather than kind of four or five months. But what we can look at is kind of scale and experience.
So I think of the.
I think so far we've done 150,000 assessments and we have a completion rate of 85% which is way above what we've seen previously. Barely. We're getting no dropout.
I think we've got a candidate satisfaction score as I've shared earlier, of 9 out of 10, which is incredibly high in this space. What I find interesting actually is um, we also allow candidates the opportunity to give us some feedback and a bit of a optional comments box which normally if you apply for a job you walk straight past 81% people are leaving those comments. The feedback on the process in through which I think is staggeringly high. It's not part of the assessment, it's not part of the process to help them get a job. Little optional text box which normally you walk straight past. But we're seeing them so engaged in the, in the tool, enjoying the experience. They want to share their feelings and most of the time it is super positive and about how simple it is, how personable it is. And they've always articulated it as it's like talking to a friend because it's a chat based, two way conversations. It's brilliant. And I should share with you what I remember the data I mentioned about the, the candidate report and giving them something back. And again I link this to consumers. We've got to take them on the right process to ensure they maintain, you know, their loyalty to Costa. I think we saw that 94% of people found this either helpful or very helpful, which I mean you almost, you're smudge below 100% there, which I think is wow.
[00:18:08] Speaker B: Yeah, those are phenomenal results. And if you think about like compare it to the old school process, I would imagine it would be much, much lower. Right.
[00:18:18] Speaker A: I think the only addition I'll put on that is and we should always remember the diversity lens. So the campus experience lens we know is working. But what about the diversity lens? And we started to look at some of the metrics of change. So we haven't done the deep dives yet into things like ethnicity and disability. But we did look at age and gender, which are two really important things for us in stores. And we looked at the end to end process from who applies for a role, who is interviewed, who is offered, who starts and with age and we looked at the Gen Z bracket and the slightly older bracket and it was negligible change throughout the process.
The deep dive we did actually was on gender. Obviously trying to get gender equality is important to everyone in a minute. And the end to end process we saw a 0.4 variance which in data terms means there was no variance.
[00:19:07] Speaker B: Fantastic. Fantastic. Wow.
Did you also cut the hiring time or the time to bill? I'd imagine that gets faster too.
Is it too early to ask?
[00:19:18] Speaker A: We haven't looked at that specific. I think what we're more interested in is could cutting the time taken to take an assessment. So I looked to see how long it takes to do a situational judgment test and it was about 20 minutes which it's too long.
This takes between 5 and 10 and most people sit somewhere in the middle. So the candidate experience, it's super simple, super quick for them. And then we're straight into the inbox for how we start rank a shortlist. So it's, it's efficient. Sure.
[00:19:45] Speaker B: So I mean what a great success. Where are you taking this next? What are your next steps?
[00:19:50] Speaker A: I think what we, because of what we've seen, especially on candid experience, we want to see what the scale is.
So if I talk about the baristas in the uk, that's a portion of our business.
We have global stores where we can start to think where could AI or SAPIA play a role. In addition to that you're here in support center or head office if you will. We've not even scratched the surface yet with assessments and AI and innovation. So I think that the quick answer to your question is scale. Where can we take it next is we have opportunities, we have many, many proof points. I think once we get into the 12 months review and tick off all the business case, we'll start to look at how we get bigger and better with it.
[00:20:32] Speaker B: What's one piece of advice that you would have for somebody who is considering this journey?
[00:20:38] Speaker A: Be brave. Because I think you can look at AI and you can look at all the reasons not to do it because you worry about jobs and legislations and this and this.
Sometimes you've got to be brave as a leader. Take a step forward and embrace, embrace the unknown. You know, this was the first time we had brought AI into the talent space.
So it was on my shoulders whether it went right or wrong and it could have gone both ways.
And I, and I dwell on the point I made earlier. Patience is you won't know everything at the start. You won't know anything through the process you will by the end. But be patient in how quick you try and do this and go about it the right way and bring the right people on the journey with you.
[00:21:19] Speaker B: Wow. Well, congratulations David. I mean this has been of course a brave endeavor, but also congrats to the patience that you had and to getting all these results in. So really appreciate your time. It was also a pleasure talking with you, David. Thank you so much.
[00:21:35] Speaker A: No worries. Thank you. Have a great rest of day. Bye for now.
[00:21:39] Speaker B: And that's a wrap for our conversation with David Hughes, the Global Talent Director for Costa Coffee. Costa Coffee hires about 5,000 frontline staff annually from 250,000 applicants.
David discussed the challenges and strategies in frontline hiring, emphasizing the importance of hiring relevant talent that is culturally aligned.
David shares his insights on leveraging AI to streamline the recruiting process, reduce bias, enhance the candidate experience, and ensure a good cultural fit. His advice is to focus AI transformation not just on technology, but on patience and stakeholder alignment, purposefully addressing specific business needs.
Thanks for tuning in to the what Works podcast. Until next time, keep pushing the boundaries of what Works in your world.
