Joe Leech

S1 E9 Meri Williams CTO of AI healthcare startup Healx, former CTO of Monzo

Meri Williams. CTO of AI healthcare startup Healx, former CTO of Monzo, Print on demand Moo and Marks and Spencer. Chair of the fantastic lead developer conference. We talk collaboration with Product & Tech, overlap of disciplines, inputs to decision making and more.

Apple Podcast Spotify Google Podcast Overcast RSS Feed

Show notes


[00:00:00] Meri: [00:00:00] And we really democratize data. Once I, almost anybody could go in and build something and look at her to understand something and more details. It wasn’t the realm of analysts or only product managers, but anybody could go and understand how things were going and spot things as well. And you, don’t not just swapping for half a way to highlight them so that they keep discussed by relevant.

[00:00:26] Joe: [00:00:26] Hello, and welcome to making better decisions to podcast for digital product people. I’m your host, Mr. Joe Leach, and I’m a product coach, speaker writer, and author. Each episode, I talk to a product leader from a startup scale-up or large enterprise about one thing, how they make

[00:00:49] this time. I’m joined by Mary Williams, CTO of AI healthcare startup helix. Now he’s also the former CEO of Monzo. Print on demand company move as well as marks and Spencers. Now he’s [00:01:00] also the chair of the fantastically develop a conference to tell us about, about you. What Y you know, you’re a little bit of your history where you are working.

[00:01:07] Now, tell us a bit about you. Yeah, sure. So I’m Mary Williams. I’m a CTO these days, but I started out as a hardware hacker in South Africa as a kid, but my first computer from broken parts and that kind of thing, and early on was much more into, into hardware and stuff. So I kind of weird plans, Famer bill of sophomores first satellite when I was a teenager.

[00:01:26] So my, my excellent parenting advice to parents is don’t let your kids do it. Anything cool. When they young. Cause I sold something that went into space and it’s basically still been downhill since then absence. But these days I need the technology teams at the moment I’m at him, which is a AI powered drug discovery startup.

[00:01:45] So we find drugs that can be repurposed for rare diseases because there are 70,000 plus spreadsheets in the world and millions of people affected by them. But because it’s. By kind of pharmaceutical company standards, like a small markets. Um, it [00:02:00] is not very invested in, and so we’re using artificial intelligence, particularly NLP and deep learning to find drugs that we can reuse for some of these rare diseases.

[00:02:11] But prior to helix, I was at Monzo for a while and grew the team quite significantly there from about 50 to 250 to all just while the company went from under a million customers to over 4 million before that, because it moved. I proud that I’m going to sit still. I was part of the team that they’ll come to UK in the early days of becoming a digital service.

[00:02:30] And I started my prayer Procter and Gamble’s, I had a very kind of corporate staff and then I’ve gotten more and more t-shirt and jeans wearing ever since

[00:02:41] it feels to me, like sometimes that that’s a sort of a sanity check for any company. If you can work in a t-shirt jeans generally, you know, they’re going to be all right. Don’t you that’s it. They’ve got a dress code. It’s. Putin. And so in my experience, definitely. All right. So on the podcast and we talk a lot about making decisions.

[00:02:58] So [00:03:00] can you talk and think back to maybe a, sort of a large decision that you and your team have had to make previously and sort of talk us through what was happening in the context and what were you trying to achieve? The biggest one, we’re in the midst of making at the moment, actually, so that the history of helix has had a bunch of super smart people, but not really anybody who had come from a product management background, a way of thinking, or even service design, user experience, design, product design on that, on a thinking tool.

[00:03:26] And so we’ve just done a big kind of. They exploratory discovery discovery project with Nazi Oliphant studio, by helping us out a little bit, for me to pretty much figure out what our product strategy should be, what products should we even have? So what we had already was a series of tools. I would call them that have been developed the kind of way that AI stuff tends to get developed.

[00:03:47] And in an academic sense, all my stuff going, we’ve got a, we’ve got a problem. We think we can solve it this way. Let’s build a prototype. That’s proved that it works. And not taking it from that, turning it into a true product for our scientists, our [00:04:00] pharmacologists and curiousness to use. And so that’s the biggest decision that we’ve been making recently is to really invest in product management.

[00:04:06] So think of these as products, even if the primary users are all internal, but you know, I, one of my early conversations with team was, you know, what do you think of as production? And they were, Oh, we don’t have production. We don’t have any external users. No. Okay. Let me ask this a different way. Who can’t do that job if your system’s down and they pointed to.

[00:04:24] The whole rest of the company, I’m like, cool. That’s what we’re going to call production from now on is anything that somebody else can’t do their job if it’s not working. And so, yeah, our biggest, uh, biggest decision has been to invest in product management techniques and hiring support managers. Our first one starts in a couple of weeks time, which we’re very excited about and to rarely.

[00:04:45] To start operating in that much more user focused kind of way, and to have proper strategies and measurement around this. And so that’s probably more matter than, than you intended to ask about Joe, but [00:05:00] right now is that decision to invest in doing things in a product play in the first place. I really liked that idea.

[00:05:10] So the idea of course, that, you know, because like you say, you’re not directly out there to consumer style users, users, the product is the business itself. That actually, that can almost be an afterthought. And now we see this quite a lot, don’t you? That, Oh, it’s just the team using the product. We don’t have to spend any effort in making it robust or even nice to use they’ll cope, you know?

[00:05:30] And the reality is I liked the way you talked about is actually, well, If they can’t cope, then the business has got a bit of a problem. Hasn’t it? Really? So I left that way as well. I’m sorry. I think, I think my. When, what a little choppy there for a moment, I probably realized it’s this the one challenge of having two people on video calls on the same connection, I suppose.

[00:05:53] But yeah. And as much as anything that, you know, we’ve got these amazing curators and clinicians and [00:06:00] pharmacologists, they’re all super smart. They’ve got such expertise. You want them working on, you want them using that expertise to its fullest on these rarely gnarly, hard problems. You don’t want them worrying about, you know, Where that button’s gone and, or having to.

[00:06:17] Sometimes the best interfaces are the ones that disappear away while you focus on the luck. And I think that, you know, anytime where somebody is having to write themselves notes about how to operate the menus or which buttons do you, you know, that you’ve definitely not gotten where you needed to from it from an experience side point of view.

[00:06:35] I mean, if, just time. So I’m not being unfair, that the team that was already better than an amazing job, they just didn’t have anybody on staff who was a experienced designer. There was one front-end engineer who was particularly UX minded and rarely wanted to do the right thing, but knew himself. He wasn’t a specialist in that.

[00:06:54] And so it’s already made such a difference for us to just investigate these resets and such design thinking and [00:07:00] some, and some product management thinking and what we’re really looking forward to having kind of. Best two product designers, not best department managers on staff, so that they can be amassed in this kind of quite complex and difficult area.

[00:07:13] Day-to-day that making, making the complex. Tangible and useful is, is, is really important, but we are serving expert users. We don’t have to make it so anybody can walk in off the streets and use it. We need to make it. So it’s supercharging the capabilities of our scientists, which is a different kind of service design challenge than, than you often get in, in the roles I’ve been in previous fan anyway, which were much more about making it.

[00:07:40] So you don’t have to ever be trained or become expert in the hanging in order to get at them. So I liked the way you talk about it. Basically you’re designing it for the experts, which is quite a nice place to be. And that you’ve, I suppose you’ve got them there. So, you know, really in essence, what they kind of need and what they expected a bit more, you just got to make sure the product fits fits them [00:08:00] better than not, you know, say struggling with buttons.

[00:08:02] Cause you know, scientists are not, I suppose the cheapest of resources to have on hand and you don’t want them to spend ages, you know, written the manual to figure out what they’ve got to be doing. They should just be getting on with doing their job. So I really liked the way you talk about that. And as part of this kind of moving towards sort of being more user experience focused and product focused, if you sort of put any processes in place for that yet, have you got anywhere about how that’s sort of working in terms of that, of processes around that?

[00:08:27] Not a huge amount yet, because as I mentioned, our first product, it’s part of manageable join in, I think in two weeks time where we’re very, very excited to have him having have him arriving soon. And we already had teams that were following a sort of agile ish versus let’s say not, not, I mean, I think, I think, you know, we went on a hope to, no, I’m not, I’m not a zealot.

[00:08:45] I’m not like a sprint. Oh, it’s from all die kind of I Julissa anyway. So I think the key thing that. The key investment thing we’re going to make is in having [00:09:00] a better articulated product, vision, and backlog, and then having refinement sessions. So the biggest opportunity I’ve spotted is. Quite often the discussions about what the feature needs to be or how things might work or what even, you know, fundamentally what the problem is that we’re trying to solve when those are happening in the middle of a, of a planning session.

[00:09:21] I think they’re not as effective as if you can separate that from, from your regular planning and. Be able to go in depth and be able to do kind of, you know, been in teams. The coolest story workshops have been these refining workshops, but just helping people outside of the context of needing to estimate and deliver really think about what, what problem are we trying to solve?

[00:09:42] What more do we need to know? So you can front load any research. You can look ahead a little bit, but not get distracted from what you’re delivering right now. That’s interesting. So I like that way you talked about it, that it’s sort of not having a lot of those conversations initially in that, in that.

[00:09:59] Problem solving [00:10:00] workshop where you’re right. You know, estimating and doing the stories as you do the work ahead of that. So I’m guessing the decisions are already mostly made before you hit that story workshop in your experience, then, you know, maybe from some of the other, your previous roles, how what’s worked well in that regard, what’s kind of worked well in that sort of upfront piece of research to help frame the decision, I suppose, isn’t it maybe the best way to think about it?

[00:10:21] What, what advice could you give people around that, that. So I think, I think the most effective thing I’ve seen, and this is true at MNS and Mo and, and to some extent at Monzo as well was getting within the team that cool collaboration between your designer, your product manager and your, your tech lead, or, you know, Uh, lead engineer, whatever the, which I, you know, for the record, I think a tech lead or lead engineer or lead that because it’s a hat, someone was, it’s not, it’s not a title.

[00:10:48] Right. It’s just that, it’s just that there’s the person who gets to go to more meetings and be more of the upfront discussions about, about what, what the problems are and what questions we need to answer. I think getting that collaboration to work [00:11:00] really, really well. That’s what makes the biggest difference.

[00:11:02] So then you can have your product, product folks understanding the needs and. Steins think about how they would measure improvements. You’ve got design folks involved early enough that they can say, I think we need to do some, some research here, or I’d like to think about trialing a couple of different options, or maybe we can AB test this or whatever else.

[00:11:22] And then the, you know, the, the tech lead, they’re not, not, I don’t need to contribute on what’s technically feasible, but also to contribute that their ideas and the other engineers ideas. And I think. That’s the real joy of true multi-disciplinary team working, right. Just that you get it. W you sacrifice a little bit, the super, super clear delineation of roles, but you, you gain so much in the churn.

[00:11:46] The ration almost always happens in the overlap disciplines, not in single disciplines alone. And I, I remember, I remember learning this really early in my career, actually. Uh, I grew up at Proctor and gamble, as I mentioned down here, which is the world’s [00:12:00] biggest consumer goods company. And they had realized that almost every major scientific breakthrough they had came from an overlap between disciplines.

[00:12:10] It was when the chemists who focused on dishwashing liquid, and, you know, the folks focused on a completely different area from a completely different discipline got together that the most interesting stuff, that stuff was invented my favorites of which I believe this is true rather than apart. Well, I was told that in a fairly like formal setting at, at P and G when I was first, there was that they used to run these sessions called the connections, develop where they figure out how innovation from one area might be useful in the, in the rest of the company.

[00:12:39] And apparently the reason that P and G invented sprinkles, which is not another one of their brands, that’s Kellogg since, but they invented Pringles because. When you make pumpers, you get very good at making a bunch of stuff into a par, a perfect parabola shape. And somebody went, we’ve developed Randy Good technology for [00:13:00] taking some stuff and turning it into a parabola shape.

[00:13:03] What else could that be useful for? And so fundamentally, almost the same technology and manufacturing wise is used to. Take care, take Pringles style and form it into the perfect shape that, you know, keeps keeps all of your, all of your Pringles intact as was used and in nappies over the years when, when making pumpers and that’s, that’s a really wonderful example of this kind of overlap of disciplines coming up with something new, right.

[00:13:32] That’s wonderful that I loved the way that overlap. Yeah. That, that technology from like diapers and nappies and Pampers can be rated to bring goals. But what I really like the way you talked about that overlap with it, of the disciplines, then that kind of, that design, that product in that tech is that all, all of them.

[00:13:47] You know, there are, yeah. That them coming together is the way to make the decisions. Not effectively beavering away independently, each other, but together they make they’re better at coming up with these ideas and solutionizing than they are [00:14:00] individually in terms of doing that. I love that idea and that, that way I think about it.

[00:14:04] Brilliant. And so you talked a bit about inputs there as well, and discussing that any experiences on kind of some of the. Again, you’re from your career about inputs to making these choices where, where that comes from. So we talked a bit about user research, but is there any other experiences or any other inputs for making decision making that you kind of advocate and will turn to in decision-making?

[00:14:26] I mean, I’m, I’m a, I’m a big. Data nerd. Right? So, so I love to have plenty of plenty of data. Ideally, ideally, some decent graphs to look at as well. So in, in very commercial roles, like, you know, for instance, marks and Spencer and the end up, it made a huge difference to have really good data analysts and folks doing everything from data, data analysis, all the way up to data science, which I know that people can get very wound up about the difference between those two.

[00:14:53] But I generally would summarize it as like, When you’re doing data analysis, you know what your question is? And you’re looking for an answer when [00:15:00] you’re doing data science. Sometimes you don’t know what the right question to ask is when you want the data to provide you some answers that you then work out what the right questions might have been.

[00:15:07] Right. And again, there’s a famous example of this from the consumer goods world. And the, and I don’t know if you’ve had this before, but do you know what the best, like. You know, at the end of the supermarket aisle, there’s this thing called the golden Dilla end. That’s what the technical name is for the bit at the end of the aisle.

[00:15:24] And there’s always some kind of promotional thing there. Right. And one of the ways that I don’t know if you’ve ever noticed this, but at the end of the nappies aisle always has Bayer. There’s always some kinds of fear related thing at the end of that app you saw. And that’s because when the supermarket chain started to track with loyalty cards, what people were buying at the same time, they found that men almost are almost guaranteed to buy beer at the same time that they’re buying nappies.

[00:15:53] And I’ll, I’ll leave it to you to hypothesize about why that might be. But there was pretty clear data that showed that in, in their [00:16:00] baskets, if they were, if they were sent, uh, or they were going to. So buy some nappies with type business. You pointed out as the more international pin for it. Then we’re also going to buy, going, buy Bayer and the thing, those kinds of correlations and, and, and that kind of stuff can be really interesting on the, on the micro level.

[00:16:15] That kind of thing is used for store design now in grocery stores. But you also, you know, in a, in a commercial setting, you want to know what, what your, you know, Conversion rates are how many people are finishing the journey, all of those kinds of things. That’s how you start to look at where, what the right, what the right place to, to look at is somewhere like Monzo had invested immensely in great data from the very beginning.

[00:16:39] It’s probably the most impressive. Real time data set up that I, that I’ve ever had the joy of experiencing. It was just awesome to be able to know exactly how all of our key metrics, where we’re looking. And we really democratized data at Monzo, almost anybody go in and build a, fill, something in blocker to, to understand something more and more.

[00:16:59] If [00:17:00] you tell us it, wasn’t the realm of an analyst. So only product managers that, you know, anybody who could go and understand how things were going and spot things as well. And, you know, not just what things, but have a way to highlight them so that they can be discussed by relevant folks. And so I think that kind of data and performance kind of kind of thing is, is one input.

[00:17:20] The research that helps inform you what, what your customers want, any, anything else that helps you understand customer or user needs, I think is, is relevant. You know, depending on what situation you’re in, you’ve got surveys, you’ve got sometimes actual evidence of how they’re interacting with your, with your product, if you’re lucky.

[00:17:38] And then there’s other inputs. I think people forget about sometimes having, having now working in med tech and having been in banking previously, regulatory matters a lot. And I think that there’s often a. One of the things I really liked about, about Monzo as well was how they took regulatory concerns and really built it in to the product.

[00:17:59] So it wasn’t [00:18:00] an afterthought that made something that otherwise was a lovely experience. Then kind of feel like you’ve had the extra stuff tacked on. It was built in as part of that part of the experience. I think that’s really important because fundamentally. The regulators want to keep people safe.

[00:18:14] That’s what they’re there for. They’re trying to, they’re trying to protect individual users and consumers. And I think when you take that and view it as a. A value added things for your customers, but they’re going to be safe. And then you build it in to how you’re going to approach things. I know that, I mean, this was before I joined ones.

[00:18:34] I said, it’s not, it’s not me bragging or it’s me bragging about the team, but obviously nothing I’m taking any, any credit for, I think the way that three skill was implemented by moms, I was just a much nicer experience than almost any other, other, other bank, because they just went all. This is about making people feel.

[00:18:53] Fail and be safer. So let’s make it as useful and experience and as kind of [00:19:00] clean and frictionless and experience as possible. And so when you do 3d scale, you’re not remembering some password that you caught, you know, somebody forced you at no notice to create, because you’ve wants to finish a transaction and it wouldn’t let you not register for freebies.

[00:19:16] Just comes a little pop-up in your mom’s lap and says, is this you, is this something you want it to be happening or not? And then you it’s like, it happens above a set, certain the transaction level. It will send you a text and, you know, it’s almost like a two factor rolls cup approach. I think that that kind of that remembering those.

[00:19:32] Requirements. Aren’t just some onerous lawyer driven tick box exercise that they’re there for a reason that would make people feel and be safe. And so if you view that as just as important to your customers, as it feeling joyful and easy, and them feeling in control of their money, then you know, then you really do center.

[00:19:52] I was always very impressed with, uh, the user research team with the design team among the, the trip led by Hugo, by everyone at, at [00:20:00] really taking all of that. Properly seriously and wanting to, to do the right thing by, by building all of those needs and as well. Well, I love the idea that the idea that you embrace the regulatory way of doing things, rather than like you say, treating it as a hurdle to get over, it’s like, well, let’s, how can we do this?

[00:20:18] The spirit of what the government or the regulatory authorities are trying to do. Let’s let’s do that. Let’s work with that. And that can work really well. And especially that example you gave there 3d secure, which is that such an annoying thing at points, isn’t it? When you’re trying to do a credit card transaction, and it kind of asks you for a password that you may be.

[00:20:35] I’ve never seen before, or haven’t thought about doing, or, you know, you’ve got confirmed your birthday in a really weird eye frame or something like that. It’s just the timing of it is the worst thing, you know, it’s just right at the end when you think I’m done. Yeah. So I lived the way that you talked about that, embracing that and making that much more of a natural experience because it’s a regulatory thing rather than just sort of sighing and getting on with it, but doing something great around it.

[00:20:57] I also like the way you talked about that idea that the [00:21:00] Monzos kind of the data that they had and the way it was democratized and that kind of thing as well, and that anybody could access it and do something with it. Could you talk any more? Did anything come from that? Did any ideas come from this kind of the data, the data world that you can, you can remember from your experiences?

[00:21:16] I’m just trying to think about, about some that have been written about maybe, I mean, in some ways the, the, the work that. Is obviously very, very broadly known about the gambling block. Some of that came from, from, from data, but also from advocacy from the team that at the time was called core customers team.

[00:21:36] So they, they looked at people who are, who are at risk or in financial trouble or, or similar. And, but then they, they didn’t just go. They didn’t jump to the solution is what I’m trying to say. So that’s a vulnerable customer team, approached them, some folks and said, we think that there’s something that we could and should do to help people who are dealing with a gambling addiction.

[00:21:58] And we, we know that we can [00:22:00] identify vendors and transactions that are gambling because there’s a, there’s a kind of category. That comes with the information about, about it. But then they also went and looked into actual addiction research of what, what, when people are trying to, to, to, to stop. And that went into the design of gambling, blocking it in a really big way.

[00:22:19] So I, I remember just being so impressed by how thorough. Um, it was how well they had consulted actual people with gambling addiction. They weren’t, they weren’t designing something for their assumptions about how gambling addiction worked, people, what they wanted. They genuinely went and did proper research with Samantha and her team who leads the user research team that Monza done did a bunch of in-depth interviews.

[00:22:45] They, Randy did understand a standard and they looked at the research around how to have to deal with eviction. And so there were things in that design, like you can turn the gambling block on with one tap, [00:23:00] but if you want to turn it off again, you have to get in touch with customers, which you do in one I through chat.

[00:23:05] So it’s not a huge hurdle, but it is a, you have to do something active and that’s one extra opportunity for you to guide you already while they do this. And then you get in touch with them and you say that you want that you want it turned off again, please. And I go shot, but there’s a 24 hour waiting period.

[00:23:23] I says, I’ll wait, thank you. Wait 24 hours. And it’s the same principle, I suppose this, you know, but you’ve got someone who’s an alcoholic. She’s trying, trying to stop drinking. They’ll empty the house of booze and that makes it harder to go. And if they’re going to want strength, they’re going to have to leave the house and buy some on alcohol and then bring it back and then drink it.

[00:23:42] And that ex just those extra little bits of friction and not a huge friction, you know, Chatting in the app with, with someone and confirming that you want this turned off and waiting 24 hours. It’s not, it’s not huge. It’s not, it’s not my honeymoon. I signed in blood or take the passport down to, [00:24:00] to somewhere or anything like that.

[00:24:01] But it’s enough to let someone think, not just twice, but three, four times about whether they really want to fall off this wagon and, and it, yeah. Well, the feedback we got was that it, that it rarely rarely helped, but another, another example would be. There was some folks who realized that from, from the data, that there was a particular subset of our customers that were, were traveling internationally much, much more than others.

[00:24:28] And obviously bonds gives you great rates for the device. Basically give you the most Cod right, the best possible right. Foot for transaction in a different currency that. John trending on top of that. And they realized that maybe there was an opportunity to, to do, uh, uh, a special kind of account for, for that type of customer.

[00:24:46] And so with the recent launches of mums, I pass them on our premium. You can see some of that thinking coming through that they’re all different customer segments, evidence in, in, not in how. We imagined our customers to be, but how our customers actually [00:25:00] were behaving, certainly in pre COVID times, obviously with COVID, I think people’s travel and work and everything else patterns are quite different than they might have been before.

[00:25:08] But, but it’s again, been really nice to see that. Data centered approach to understanding customer segments rather than what in a blunt they I’d often seen as a very marketing center central for the, the personas people wanted to sell to you rather than the actual customers and the way that they. Truly behaved was often where it’s at, where it came from in some other organizations that I’ve been a possible.

[00:25:34] Um, the other great example of, of data being used was at, uh, the government digital service. When we first did come to decay, the folks who worked from accessibility, highlighted really early on that we needed to be able to see in our metrics when assistive devices were being used, if that was at all possible and it made it much, much.

[00:25:55] Easier to advocate for some of the improvements, but with that, from an accessibility point of view, [00:26:00] um, my favorites, which was actually that, I don’t know if you know the story of the WK fondant, but there’s a, there’s a, there’s a font that was developed by a particularly wonderful set of, I won’t get the name wrong for people who develop on school.

[00:26:13] They’re not economical for us all. There’s something else. Anyway, designers type design, it feels like I spent two and I’m just, it’s partly escaped from my head. But very famously that the font that’s used on most waistlines and so on in the UK is one that was specially designed and specially tested for being legible in the middle of driving rain in the dark with just your headlights of your car or your motor bike lighting the sign up, it’s still legible.

[00:26:44] And now they’re amazing. And one of the things that the design team in the early desk of UK decided was that they would commission it. And an updated and digital version of this to be developed. And then what the folks who were specializing on the accessibility side who’s they [00:27:00] organize some extra rounds of user research that were specifically focused on whether this fund was rarely helping people.

[00:27:06] So not just theoretically, was it more legible or easier for folks maybe who were dyslexic as well or had dyscalculia also those who were partially sites at all, or had any kind of vision impairment? And, and they did, they, they were looking at like, from how, almost like an optometrist that, you know, how far away can you read that sentence?

[00:27:28] And is it clear what the heading says from here and so on? And it was, it was quite interesting research to watch being performed because you had, there were asking people to, you know, go to the line at the back of the room, please. And now read the screen from all the way back then it was like, Almost almost literally like a, an optician, but that was also really, really interesting, interesting use of it to be, to be sure that we were setting not just some successes, but Charlie.

[00:27:54] Well, I was, I was very proud of that, but you specialized in that one? [00:28:00] No, I left the idea that when you talk about, I guess that almost the data identifying something that needs to be. Two more work to be looked at, you know, from the gambling block through to the, to the use of assistive technologies. But then you kind of, I suppose then you overlay that with some further user research or, you know, interviews or any quite sensitive way to, to understand really what the problem is, and then look at what the solution could be to that alongside it.

[00:28:23] So rather than, you know, taking that data and leaping to a solution, I think you talked about, which is never necessarily the right way to do it, but to take your time. Consulting almost with, with your users to figure out together with that and what, what solutions could have work best for them in it.

[00:28:37] That’s, that’s really interesting. Both of those examples. Talk about that there as well. Great. So we’ve talked about some good decisions then that have happened in your experience that you’ve been working with. Can you talk about, or think about where anything may have gone a little bit. Not as well or quite a bit wobbly in terms of anything that’s happened with you in the world.

[00:28:55] You’ve worked in you’re frozen when the wobbles already happened and be the one asked to [00:29:00] fix it. I’m Randy plums, facet. I mean, if I mislead Marx and spend some move from being on what was effectively, Amazon white label, and then stopped making their white label thing available and they moved, they did a four year long, very expensive project to move to the new website and conversion hall at the point that they did say.

[00:29:19] I don’t think I was speaking out of school, mentioning it. I think he was pretty public at the time. And so I was, I was brought into to help, I suppose, with the recovery, from that, with making things a bit more, a bit more useful again, I suppose I, I don’t have many huge disasters or I, I have a few close, but let’s talk about the microscopic close call Spencer one.

[00:29:39] That’s interesting. How did you deal with that? How do you know conversions? Harvey? They call it right. Mary, can you sort this out with, where do you start? The first one called for it to be shown back. We’ll call them again and see, he brought me in. Then he very much started the journey to recover already.

[00:29:56] And I mean, it was very plastic stuff, right. You know, waterfall [00:30:00] and not super customer centric and both focused on getting it done and knowing what the, what the results were going to be. And so, you know, a lot of what we, what we did was to look at why customers are finding it difficult, look at how the customer service folks are finding it difficult as well.

[00:30:15] My first interaction to M and S was actually, if I was. Running a little product studio at the time. And we were brought in to build a tool for online ordering for the customer service folks. Because they were trying to use the same website, um, as, as consumers with, to do ordering which isn’t, isn’t terrible on the face of it.

[00:30:32] Right. Let’s make it. So the customer service folks see the same thing that can see the customer’s pain, but then when a customer rings up, which many of them, and as customers do with a list of literally the code names like catalog shopping, That call up with the code for the things that they wanted. I want four packs of tea, one, two, three, seven, six, eight, please.

[00:30:52] And when you, when somebody is reeling off these numbers to you and you’re having to go just second, what, what is the actual product called? I [00:31:00] can’t search for that code. I can only search for the name of the thing. And so we literally just went one time and did a load of. To you sitting in the call center, listening to how difficult these calls were watching people, trying to try and use it.

[00:31:15] Brilliant designer, who, who worked with me at the time and carrier did just a really brilliant job of going. Ordering taking a phone order is very different to doing an online order. And really what people want is to be able to reel off the 10 things that they want. And then for those 10 things to come up for the agent to just select the number that they want to, each of them put it all in the box at once without the screen manager refresh without things, things change.

[00:31:44] And that was what we built. It’s actually that. That first little team, but that was really a contractor team, but that went in, we all then kind of went in house. So over time as we got convinced to join permanently. And that’s when I took that, the more, the, more of a leadership role, but, you know, I think [00:32:00] we.

[00:32:00] I think we reduced call times by 40% or something like that, which one you’re taking, taking orders over the phone, which is the difference in demographics, feminists, customers. Many of them were pretty comfortable to kind of browse on their iPads, but they didn’t want to actually put credit card information in and finalize the sale on that.

[00:32:20] And so, so yeah, it was, it was quite interesting. I’m just reflecting on why I don’t have all that many. Yeah, I’ve actually, I do. All right. It’s kind of major incidents. I have survived kind of chats that, speak to this, but yeah, luckily not, I suppose I’m a big fan of the style of decision making, where you consider how easy it is to undo or change a decision.

[00:32:41] And so anything where it’s very permanent. Lots of upfront work, be ready. Sure. It’s the right thing. Everything else, if you try and make reversing the change or changing your mind as uncomplicated, as on as least painful as possible, the kind of Socratic method of decision making, which [00:33:00] is just, is it good enough for now and safe enough to try and.

[00:33:04] If it turns out to have been the wrong, wrong move, then we’ll adjust, we’ll reflect and we’ll adjust. And almost everything you can do that there’s, there’s few things that you can’t find some way to make it more iterative in the way that you deliver it. And the way that you test the water. And I show that the thing’s going to be okay.

[00:33:24] Now that, that way you talk about how easy is it that decision to reverse, you know, that gives you a clear risk, I suppose, the scope of it and the danger that could propose if it goes wrong. I think both the examples you used of marks and Spencer’s there. And the, you know, the, the phone lines are too.

[00:33:38] Decisions that you can’t easily reverse on day. So it’s like that’s given you that clue that it’s not that easy to reverse, you know, we’ve, we should probably put some more time and effort really making sure this is absolutely the right thing to be doing rather than leaping in, in terms of doing all that great.

[00:33:54] Well, we’ve pretty much run out of time today, Mary. So thank you so much for your time today. Is there anything in particular you want [00:34:00] to share with the audience? You know, you’re looking to recruit any job offerings or anything you want to talk about? I mean, yes. W we are recruiting at helix, I’ve got a product manager roles, a user research roles, a bunch of leadership roles.

[00:34:11] I’m sorry. Various applied science bioinformatician and engineering roles. So anybody who’s interested have a look at helix, the IO slash bris. Um, and then I’m on I’m Mary Williams on Twitter. I’m a geek underscore manager, the poor guy who has econ manager without the underscore probably gets a load of questions that he doesn’t particularly want to answer.

[00:34:32] So remember the underscore and yeah, I suppose just stepping the other thing that I’m involved in that I should, of course mention is I help. Chad, let me develop a conference. So for folks who are in a technical roles where they are, they are leading, whether that does need to be as a manager, you can also just be as a, you know, a leader.

[00:34:49] Who’s an individual contributor. The lead dev is a, is a great conference focused exactly on folks in that situation and on all the kind of skills and knowledge that you need. Uh, [00:35:00] Excel in a typical leadership. So come and check that up if you haven’t it. Great. Thank you so much for your time today. Mary has been wonderful.

[00:35:06] Thank you. Excellent. You’ve been listening to making better decisions, podcasts for product limits, to get show notes, transcripts, and more. Go to Mr. forward slash podcast. Otherwise please leave a rating or a comment in your podcast app. It really helps me and it really helps the show. See you next time.

Leave a comment

Leave a Reply

Your email address will not be published.


I'm Joe Leech and I 10x leaders, the business and the products and services they create. I bring 15 years in tech, $20b in revenue, experience with 30+ startups & FTSE100 / Fortune100 giants. Together we can do great things. Work with me.

Host of

Keynote Speaker