WEBVTT
00:00:08.480 -->
00:00:09.839Hello, I'm Philip Stoughton.
00:00:09.839 -->
00:00:14.960I'm at Fermionics in the UK, and I'm with Will Patrick, who's the managing director.
00:00:14.960 -->
00:00:15.439Hiya.
00:00:15.439 -->
00:00:16.480Great to see you, Will.
00:00:16.480 -->
00:00:25.839Uh really fascinated by the business because I think this is a story that's happening in other places, but it's kind of unique the way it's happened here.
00:00:25.839 -->
00:00:27.760You're actually the third generation.
00:00:27.760 -->
00:00:31.519Your grandfather started the business decades ago.
00:00:31.519 -->
00:00:35.200You taken over recently from your father.
00:00:35.200 -->
00:00:40.880Talk me through a bit of the history and tell me what the new broom is sweeping through.
00:00:41.119 -->
00:00:48.799Yeah, I mean it started from my my granddad, but my nan as well, um, 53 years ago now, which is crazy.
00:00:48.799 -->
00:00:53.359They my granddad was a silkscreen artist, then moved into PCBs.
00:00:53.359 -->
00:00:59.920Then my dad took it on maybe 20 years ago, moved from PCBs into assembly, and now here we are.
00:00:59.920 -->
00:01:02.880So about two years into being managing director now.
00:01:02.880 -->
00:01:05.359So into that sort of journey of the third generation.
00:01:05.359 -->
00:01:05.680Yeah.
00:01:05.680 -->
00:01:06.959And yeah, it's exciting.
00:01:06.959 -->
00:01:07.680A lot's happened.
00:01:07.920 -->
00:01:09.840Yeah, I mean it's fascinating, isn't it?
00:01:09.840 -->
00:01:30.159Because as as you have those generations and clearly pivotal changes, revolutions each time, your revolution seems to be a bit of a digital revolution, and you've really changed the way the business operates, and you've really built something that is capable of scaling much more quickly and is a bit perhaps more agile for your customers.
00:01:30.159 -->
00:01:31.760What was the strategy behind that?
00:01:31.760 -->
00:01:38.239Did you know you were going to do that when you came in, or did you just see as soon as you came in, those were the changes that you needed to make?
00:01:38.560 -->
00:01:40.159I saw the gaps, I think.
00:01:40.159 -->
00:01:41.519I knew what we did really well.
00:01:41.519 -->
00:01:43.920I knew we serviced our customers really well.
00:01:43.920 -->
00:01:48.000But I also looked at our revenue for the about 10 years before that, and it was completely flat.
00:01:48.000 -->
00:01:50.239And I tried to understand why that was.
00:01:50.239 -->
00:01:52.480I have no interest in keeping it flat.
00:01:52.480 -->
00:01:55.040I want to grow, so I need to know how we did that.
00:01:55.040 -->
00:02:01.280Um, and quite quickly I walked in and thought, you know, I think I said earlier, I haven't got years of experience behind me.
00:02:01.280 -->
00:02:04.560I don't know every single part number like a lot of the staff did off the bat.
00:02:04.560 -->
00:02:10.639I needed a way to say, let's cut through all the mess as such, look at the top level.
00:02:10.639 -->
00:02:14.479Can I get this whole thing run on figures and dashboards and top level?
00:02:14.479 -->
00:02:16.080So loads of things changed that.
00:02:16.080 -->
00:02:25.199I mean, we got a new MES system, which was huge, much well, fully traceable, whereas the other one wasn't, which means everyone needs to add their data.
00:02:25.199 -->
00:02:26.639That's a massive cultural change.
00:02:26.639 -->
00:02:28.159We're about two years into it now.
00:02:28.159 -->
00:02:32.159I'd say we're probably about 85% of the way through there culturally.
00:02:32.159 -->
00:02:33.680People now understand it.
00:02:33.680 -->
00:02:38.879Then I got delved quite a lot into Power BI, so I now get all the top-level dashboards based on that.
00:02:38.879 -->
00:02:41.599If it doesn't work, we fix it at the core.
00:02:41.599 -->
00:02:45.280I can show customers a lot of their data from our system now.
00:02:45.280 -->
00:02:49.840If they find errors, you create a much more collaboration where they're, and that's not right.
00:02:49.840 -->
00:02:51.360And then we say, Oh no, it's not, you know.
00:02:51.360 -->
00:02:52.879Well, we need to fix that our thing.
00:02:52.879 -->
00:02:55.439So yeah, that pushed a lot of it.
00:02:55.439 -->
00:02:59.120Outside of that, I think we just needed a lot better quality controls.
00:02:59.120 -->
00:03:05.680So a lot of the investment we've done on machines and our infrastructure and factory helped that.
00:03:05.680 -->
00:03:15.840Fixing it really at the root, and it's a long-term strategy, obviously, because uh I hopefully have quite a long career left in me, so it's worth doing that now rather than and that's probably the best way to say it.
00:03:15.840 -->
00:03:18.400My dad had a sort of two, three-year plan.
00:03:18.400 -->
00:03:20.879I'm looking at a 20, 30 year plan.
00:03:20.879 -->
00:03:21.919They're just different, really.
00:03:21.919 -->
00:03:26.400And the step change is a lot of investment initially to give us the infrastructure to grow.
00:03:26.639 -->
00:03:30.400Yeah, and I think you know, the proof of the pudding is always in the Ethan and these things.
00:03:30.400 -->
00:03:36.240And and you said 10 years of static revenue, you've seen last two years of quite substantial growth.
00:03:36.400 -->
00:03:37.520Yeah, yeah, yeah, yeah.
00:03:37.520 -->
00:03:42.719I think a lot of that is we can give visibility to customers, they like that.
00:03:42.719 -->
00:03:52.960We've obviously invested, you know, hundreds of thousands in our facilities, we look a lot better, and we've cleared a lot of space by being or implementing operational improvements.
00:03:52.960 -->
00:03:58.879So now they're looking at a factory four where they can see a third of it's empty, ready for them to pull more work.
00:03:58.879 -->
00:04:04.159Before, like I said, we were reasonably profitable for the size, for the revenue we had.
00:04:04.159 -->
00:04:08.719So we needed to sort of strip that all back, start again, uh, put a bit more overheads in.
00:04:08.719 -->
00:04:10.240But yeah, we've got an infrastructure.
00:04:10.240 -->
00:04:17.360And I think also when you have customers around with a contract manufacturer, you're not you're not selling a product they might buy once and never come back.
00:04:17.360 -->
00:04:18.879You're having to sell a partnership.
00:04:18.879 -->
00:04:34.800So I think, hopefully, hearing a story of you know how I'm invested in this for the next 30 years is quite enticing to a business that they know there's stability there as well, and they know it's gonna continue, and I'm gonna continue to put a lot of effort into their product.
00:04:35.040 -->
00:04:36.399Yeah, I think that's really interesting.
00:04:36.399 -->
00:04:42.639And what and what you're bringing is you're bringing that future-proofing, but you're also drawing on what's happened in the past.
00:04:42.639 -->
00:04:48.319And you you said you know you've got a core ethic of customer service that was really strong.
00:04:48.319 -->
00:04:56.240So actually maintaining that and layering that up with tools that allow you to do that even better is is hugely advantageous.
00:04:56.240 -->
00:04:58.560What are your kind of ambitions for the business?
00:04:58.560 -->
00:05:00.079Where do you where do you see it going?
00:05:00.079 -->
00:05:02.240Do you see like a nice steady growth?
00:05:02.560 -->
00:05:03.439Yeah, I think so.
00:05:03.439 -->
00:05:07.759I think um probably created an infrastructure now.
00:05:07.759 -->
00:05:13.839We're around 10 million, created an infrastructure that's probably gonna 25 without having to increase that many overheads.
00:05:13.839 -->
00:05:19.040Each time I think we need more overheads in admin or people.
00:05:19.040 -->
00:05:21.360I think we've got a good team around us now.
00:05:21.360 -->
00:05:26.480I'm gonna lean quite heavily probably on AI and data, where we can do a lot of the admin jobs for us.
00:05:26.480 -->
00:05:27.199And that's my goal.
00:05:27.199 -->
00:05:33.680Sometimes you're gonna have to, but um, get to that 25 million mark in the same physical infrastructure we have now.
00:05:33.680 -->
00:05:38.959We might be able to run shifts, or there's probably ways we can run surface amount, uh increase capacity.
00:05:38.959 -->
00:05:44.160Um, and a lot of the admin work can be done in ways where you don't need a lot more people.
00:05:44.160 -->
00:05:44.560Yeah.
00:05:44.560 -->
00:05:46.800Um that's kind of my goal.
00:05:46.800 -->
00:05:49.519I think we could quite easily achieve that.
00:05:49.759 -->
00:06:01.920Yeah, and you've replaced kind of tribal knowledge, which you had before, you know, people just doing it that way because they'd always done it that way, and they knew almost a feeling of how the minds were tuned and everything with stuff that's really data-driven.
00:06:01.920 -->
00:06:02.800So that's huge.
00:06:02.800 -->
00:06:11.199You invested in MES, you've invested in the Koyang inspection system, which has given you the K Smart data, you've got the Luminovo system.
00:06:11.199 -->
00:06:19.120Does that reflect a strategy of like pooling as much data in into one resource as possible so you can actually drill down in that area?
00:06:19.439 -->
00:06:19.680Yeah.
00:06:19.680 -->
00:06:26.639Well, uh every action someone takes in this factory, sh there has to be a data point for it.
00:06:26.639 -->
00:06:30.079Um we're scanning on and off every single jobs where we weren't before.
00:06:30.079 -->
00:06:33.519Everything now goes through our main manufacturing system.
00:06:33.519 -->
00:06:36.319Luminovo, like I said, great bit of kit.
00:06:36.319 -->
00:06:42.319I need to get that with a better link API so it almost becomes one system in itself.
00:06:42.319 -->
00:06:51.680Um but yeah, and the main thing is you want to get a data point for every action that doesn't make it any slower for the employee or operator as well.
00:06:51.680 -->
00:06:57.439So there's things like in stores, we probably want to set up scanners so the whole thing just works a lot cleaner.
00:06:57.439 -->
00:07:05.839Um but we set the infrastructure now, we can see where there's gaps, and we can I've kind of stripped it all back, and now we need to build it all back up again based on that data.
00:07:06.079 -->
00:07:11.600Yeah, and you're basing it on new new data-driven foundations or a data-driven um backbone.
00:07:11.600 -->
00:07:24.480You talk about AI, and I think we're seeing AI being used in areas like maybe the Koyoung um feeding back to the printer and make helping make adjustments there and the stuff you're doing with Luminovo.
00:07:24.480 -->
00:07:37.040But I think as a leader, what you want is to be able to have an overarching system, some kind of agentic system where you can actually interrogate all those to make business decisions.
00:07:37.040 -->
00:07:41.120That's an interesting future, but we're away from it.
00:07:41.439 -->
00:07:42.000Yeah, yeah.
00:07:42.000 -->
00:07:44.560I think um let's see where it goes.
00:07:44.560 -->
00:07:52.480I think being a young MD, if it doesn't work, maybe the first time round or second time round, it's not a problem.
00:07:52.480 -->
00:07:54.959I think it's we need to learn how it works.
00:07:54.959 -->
00:08:02.240But I think things like an MRP or exception reporting, I just imagine where it's just lots of data points and lots of pluses and minuses.
00:08:02.240 -->
00:08:07.360If you can get something like AI to just scan through that and say this is the one I think might be the problem.
00:08:07.360 -->
00:08:08.399Who knows, yeah?
00:08:08.399 -->
00:08:12.959Like I said, I've focused on data first and then maybe AI at that point.
00:08:12.959 -->
00:08:14.560But we're kind of stable where we are now.
00:08:14.560 -->
00:08:21.680If we got a hell of a lot more sales in and we needed a hell of a lot more overhead, maybe I'd start to sort of implement it at that point.
00:08:21.680 -->
00:08:23.360But pretty happy where we are now.
00:08:23.600 -->
00:08:34.960Yeah, you start with data, you layer on insight on top of that, and then you look at where you can deliver value for you, but also deliver value for your customers, which is uh yeah, it's a huge part of it.
00:08:34.960 -->
00:08:36.559Will, thanks so much for your time.
00:08:36.559 -->
00:08:38.000Thanks for sharing around the factory.
00:08:38.000 -->
00:08:39.120It looks very impressive now.
00:08:39.120 -->
00:08:39.759You're welcome.
00:08:39.759 -->
00:08:41.200Um, look forward to our next chat.
00:08:41.759 -->
00:08:42.000Cheers.
00:08:42.000 -->
00:08:42.960Thanks.
