WEBVTT
00:00:04.160 --> 00:00:09.759
I'm excited to welcome Tara Button, best-selling author and founder and CEO of Buy Me Once.
00:00:10.080 --> 00:00:19.199
Once a recovering shopaholic, Tara built out a movement to find the longest-lasting products on the planet and change our throwaway culture forever.
00:00:19.440 --> 00:00:26.480
Her business idea went viral in 2016, and her book, A Lifeless Throwaway, became an Amazon bestseller.
00:00:29.839 --> 00:00:34.960
It had got to a point where I've been running this business for like eight years.
00:00:35.520 --> 00:00:37.759
We'd raised a whole lot of cash.
00:00:37.840 --> 00:00:40.960
You know, we had over time raised like three million pounds.
00:00:41.119 --> 00:00:43.759
The team grew to 14 at one point.
00:00:44.240 --> 00:00:49.359
And through COVID, obviously being in the e-commerce business, everyone's at home buying frying pans.
00:00:49.520 --> 00:00:50.240
Brilliant.
00:00:51.119 --> 00:01:00.640
But you know, over time, you know, the mechanics of the business changed, and we just weren't getting enough revenue in to cover our costs.
00:01:00.960 --> 00:01:13.359
And uh the point that I spoke to you at, we were facing not being able to raise any more capital and running out of money at at an alarming rate.
00:01:13.680 --> 00:01:20.239
And um I couldn't really figure out what my choices were at that point.
00:01:20.400 --> 00:01:22.719
I think that was part of the issue.
00:01:22.879 --> 00:01:31.920
I had so many competing thoughts and ideas in my head, and the idea of letting that mission go was abhorrent to me.
00:01:32.079 --> 00:01:36.239
Um, and at the same time, I felt this huge responsibility for shareholders.
00:01:36.400 --> 00:01:45.200
I also knew that, you know, the stress I was putting my family through wasn't great at that point, but I couldn't see through the mess to figure out what to do.
00:01:45.519 --> 00:02:08.719
I realized at the end of that conversation that the only choice was, if I was serious about fulfilling that mission, which is the clarity you gave me, was to let go of everyone in my company and run the entire business myself, which is what I then did that day.
00:02:08.879 --> 00:02:31.199
I had a meeting and I said, look, everyone, this isn't what I would have chosen, but the choice is either the business closes completely and I lose everything, or I let everyone go, semi-stopped paying myself pretty much, and just try and build again from the ground up something that can support itself.
00:02:31.840 --> 00:02:39.840
And so obviously that was a huge decision, and then the practicalities of that like kicked in, and that was pretty big as well.
00:02:40.240 --> 00:02:40.560
Wow.
00:02:40.639 --> 00:02:44.159
So I mean that I mean that is a that is a huge moment and a huge change.
00:02:44.479 --> 00:02:44.800
Yeah.
00:02:45.120 --> 00:02:50.080
I I think it's the reason for that that is, I think the choice probably would have been taken away, wouldn't it?
00:02:50.159 --> 00:02:57.840
It's like that sometimes we tell ourselves that there is a choice, and there is a choice unless we don't make it, and not making a choice is still a choice.
00:02:58.080 --> 00:02:58.479
Yeah.
00:02:58.639 --> 00:03:02.719
And that was what I had been doing for you know the months before that.
00:03:02.879 --> 00:03:12.159
And or not quite seeing what the choices were that I had to make, and so that moment of having that clarity, you know, changed everything.
00:03:12.240 --> 00:03:19.039
And I'm yeah, so grateful that I did that because you know, it's been incredibly difficult.
00:03:19.520 --> 00:03:24.719
However, you know, the team being these amazing people that they were were completely supportive.
00:03:24.800 --> 00:03:27.439
They knew that if I could have kept them, I would have done.
00:03:27.680 --> 00:03:36.639
And so there was no sense of resentment or uh, you know, child you know, churlishness.
00:03:36.800 --> 00:03:38.879
They were like, How can we help?
00:03:39.280 --> 00:03:45.280
How can we help make this okay for you to run the business completely single-handedly?
00:03:45.439 --> 00:03:55.199
And they taught me everything that they did, and they put in systems and um and culled things to make it work.
00:03:56.240 --> 00:04:05.280
I think I think this is really a fascinating sort of thing to dive into because you took a business with sort of 14 people and took it down to yourself.
00:04:05.520 --> 00:04:15.919
And uh a lot of a lot of people look at their business and go, oh, there's no like there's there's there's no like everyone's needed, everyone's necessary, um, and we really focus on scaling.
00:04:16.000 --> 00:04:22.240
And this idea of when to downsize feels almost like a step back or a mistake or a failure.
00:04:22.399 --> 00:04:23.040
Yeah.
00:04:24.079 --> 00:04:26.800
But I'd be I think it's really interesting to dive into like how you did this.
00:04:27.040 --> 00:04:29.680
If you can take from 14 to 1, yeah.
00:04:29.920 --> 00:04:33.199
What what what allowed you to to do that?
00:04:33.680 --> 00:04:34.720
Spreadsheets.
00:04:36.399 --> 00:04:37.680
You think I'm joking.
00:04:37.920 --> 00:04:57.600
Um no, I um I built a spreadsheet and I got everyone to write down every single task that they did, and whether it was daily, weekly, monthly, yearly, how much time it took them and how um needed it was.
00:04:57.920 --> 00:05:08.720
I also then you know went through and I was like, well, these are the completely business critical operations that have to happen, whether it's daily or weekly.
00:05:08.879 --> 00:05:12.480
Without this, the business will crumble.
00:05:12.720 --> 00:05:37.439
Like, you know, if the customers stop getting their products, if the suppliers stop getting paid, if customers stop being replied to, very soon you have one Star Trust pilot review saying, You're a scam, very soon you'd have suppliers pulling out, and very soon, you know, obviously if you don't send a product, you'd probably be facing some sort of legal action.
00:05:37.680 --> 00:05:40.480
So those were the things that I had to solve first.
00:05:40.879 --> 00:05:51.680
And so I very quickly went about building automations, essentially robots that could replicate those tasks.
00:05:52.720 --> 00:05:58.160
But before we jump into the sort of road to the automation, I'd like I mean that's a fascinating process too.
00:05:58.319 --> 00:06:02.480
I mean, fascinating moment with the right word, but for that spreadsheet, I'm curious.
00:06:02.560 --> 00:06:08.639
Like, having done that, what percentage of the tasks do you reckon you were left with as critical?
00:06:08.879 --> 00:06:14.720
Because often we build businesses, they they morph into like more and more actions and projects and stuff.
00:06:14.800 --> 00:06:14.959
Yeah.
00:06:15.120 --> 00:06:19.680
I'm really curious, like, what percentage of those things were you left with when you had to cut it down?
00:06:19.920 --> 00:06:24.879
And picturing the spreadsheet, now I think it was surprisingly fewer than I thought.
00:06:25.759 --> 00:06:35.120
And but I would like to caveat that in that these were the operations without which the business might begin to fail within a couple of weeks.
00:06:35.920 --> 00:06:53.920
Uh then there was the kind of second tier where if I didn't do any of this, the business would just die a very slow death because no marketing emails are being sent out, no social media is being sent out, no new products are being put on the site, you know.
00:06:54.160 --> 00:06:59.120
So you might not die immediately, but it would be a slow death over months.
00:06:59.439 --> 00:07:06.399
So I dealt with the first tier first, and then I had to kind of work on the second tier.
00:07:06.480 --> 00:07:14.319
But it is it did surprise me going through the whole list, being like, nope, nope, nope, nope, nope.
00:07:14.560 --> 00:07:16.319
Yeah, it it was surprising.
00:07:16.879 --> 00:07:20.000
So gut feel, let's say the first list die in two weeks.
00:07:20.399 --> 00:07:22.079
That was about maybe 10%.
00:07:22.639 --> 00:07:25.279
And then the sort of the slow slow death pile.
00:07:25.600 --> 00:07:32.399
The slow death pile was probably another that was more like 40 to 50 percent.
00:07:32.560 --> 00:07:38.959
That was quite substantial stuff, but it didn't all have to happen at once.
00:07:39.120 --> 00:07:52.000
It was like some things might mitigate other things, but they would still be in that pile as long as enough of that stuff was happening, then you'd probably mitigate like degrowth, if that makes sense.
00:07:52.480 --> 00:07:55.680
So kind of you could roughly half what was being done before.
00:07:56.000 --> 00:07:56.160
Yeah.
00:07:56.399 --> 00:07:59.600
That still takes down to about seven people, let's say.
00:08:00.800 --> 00:08:02.720
Then you've got to you've got to make yourself the seven people.
00:08:02.800 --> 00:08:06.160
So yeah, sounds like automation was the route that you chose.
00:08:06.959 --> 00:08:20.720
Yeah, and I think that's an important point to make is that if we had had that conversation a year pr earlier than that, so we had it in March 2024, if it had been March 2023, I don't think my choices would have been the same.
00:08:21.519 --> 00:08:26.959
And that's because of AI and automation and what it allows you to do.
00:08:27.279 --> 00:08:43.279
Uh when I realized that AI can code for you, and essentially the ability to code means that you can build yourself little apps and automations that can essentially do anything as long as it's in a digital form.
00:08:43.440 --> 00:08:45.440
And obviously, we're a digital site.
00:08:45.679 --> 00:08:50.879
That opened the doorway to me believing that this was possible.
00:08:52.159 --> 00:08:57.120
But open AI hadn't, I mean, because I remember I remember this conversation, I remember so the trick trigger.
00:08:58.240 --> 00:09:02.240
Now it's kind of like a take taken as red thing that you can do this stuff.
00:09:03.679 --> 00:09:14.480
OpenAI wasn't talked about when we were, and I know it's more than OpenAI this, but like it feels like you were ahead of the curve on this because it hadn't really hit public consciousness in the same way that it has now.
00:09:14.799 --> 00:09:16.240
Yes, I think that's true.
00:09:16.399 --> 00:09:27.919
I think I'd just begun, I think maybe even even in like the couple of weeks before we talked to try and figure out solutions and to try and figure out what was possible.
00:09:28.159 --> 00:09:53.840
And I was only just beginning to scrape the surface, but I'd seen a couple of YouTube videos that had slightly blown my mind in terms of the automation space, and I and that had kind of sent my brain into a kind of like, oh well, that's possible, then this could be possible, and if that's possible, then that could be possible, and that I think gave me the I guess the courage to take the plunge.
00:09:54.399 --> 00:10:08.879
Otherwise, I I think if it had been a choice between um closing the business and running everything myself manually, it wouldn't have been the same choice.
00:10:10.399 --> 00:10:11.600
Nowhere close.
00:10:12.879 --> 00:10:17.039
So you so you you've got then this this sort of seven perils worth of tasks.
00:10:17.200 --> 00:10:25.679
You start on it, how how do you for other people that maybe you think about doing this, you've got your spreadsheet, you've got your task, what's what now?
00:10:26.399 --> 00:10:27.840
So I I went through the list.
00:10:28.000 --> 00:10:32.240
So the first thing was um inventory, I think.
00:10:32.399 --> 00:10:47.679
So if products are out of stock with a uh you know, with a brand, but we don't know that they're out of stock, then we're selling products that don't exist, which is obviously gonna very quickly start to cause some issues.
00:10:47.919 --> 00:11:02.799
And so trying to build something that would take our inventory um and automate it to the point where I didn't have to touch it was probably one of the biggest challenges.
00:11:03.039 --> 00:11:21.840
Um the way that we received inventory uh from our brands was in such a weird and wonderful variety of ways, from somebody sending a spreadsheet, someone sending a CSV, someone just sending an email saying, hey, we've only got two of these left.
00:11:22.080 --> 00:11:24.559
So, but it was generally all through email.
00:11:24.639 --> 00:11:48.159
So I built an email scraper for the brands that sent inventory in that way, which allowed it to scrape my email, take all of the information, no matter what format it was in, and put it into a Google Sheet that then normalized it against all of the products we have.
00:11:48.320 --> 00:12:03.840
And then as soon as that was done, we had the correct information, and then I used um make.com uh to automate that spreadsheet to take any change in inventory directly into Shopify.
00:12:04.399 --> 00:12:14.720
So that took I want to say a week and a half to build, and it was very frustrating because I can't code at all.
00:12:15.200 --> 00:12:48.879
So um I would get ChatGPT to write the code and it would half do it, uh, but then there would be errors, and I would the the process was um write the code, it would write the code, I'd copy and paste it, run it, it would half do something, but probably error out, and then I'm copy and pasting the errors back into ChatGPT and saying it's doing this, and then it'd be like, I will rewrite the code for you, and then it would, and then I'd paste it back, and it was this.
00:12:49.039 --> 00:12:59.200
And sometimes what would happen is that it would do something perfectly, but then there'd be another error, and then I'd paste it back, and then it would stop doing the thing perfectly.
00:12:59.600 --> 00:13:11.039
And that was the most frustrating thing, is it because I can't code, so I have no idea what's changed, and I'm for you know completely reliant on it doing the right thing.
00:13:11.759 --> 00:13:15.440
And you know, it wasn't an easy process.
00:13:15.600 --> 00:13:27.519
Um, it was quite addictive, I would say, because you never knew which, you know, the moment that it worked, it felt like magic.
00:13:27.600 --> 00:13:42.320
It was like, you know, when I could see like all of the spreadsheets and all of the data coming in perfectly and pinging into Shopify, and I'm like, I never have to do this manually again.
00:13:42.639 --> 00:13:46.639
That moment, if it feels like a miracle.
00:13:47.120 --> 00:13:57.279
You're having this time pressure looming over you of like just the doing, and then you've got to commit all this time to plan and design and go through this process.
00:13:57.679 --> 00:13:58.320
Yeah.
00:13:58.639 --> 00:13:59.840
It feels like a lot.
00:14:00.159 --> 00:14:01.679
It I mean, it was insane.
00:14:01.840 --> 00:14:05.840
I was working 14, 60 nowadays.
00:14:05.919 --> 00:14:07.679
I lost all the skin on my fingers.
00:14:07.840 --> 00:14:15.440
I had to wear gaming gloves because I was on my laptop so much it was I lost about a stone and a half.
00:14:15.519 --> 00:14:18.080
Uh so all the baby weight was gone, so that was quite handy.
00:14:18.240 --> 00:14:20.159
Uh, but not in a healthy way.
00:14:20.720 --> 00:14:22.720
I mean, it was it was bad.
00:14:22.879 --> 00:14:33.120
I didn't want to sugarcoat it and say, like, oh, you know, I'm, you know, got rid of all my stuff and you know, I didn't have robots from my company, and it's all hunky-dory.
00:14:33.360 --> 00:14:37.440
It was incredibly painful and a lot of work.
00:14:38.399 --> 00:14:46.000
The breakthroughs and the victories when they happened were spectacular and gave me the strength to carry on doing it.
00:14:46.159 --> 00:14:48.799
Um, but it wasn't without sacrifice.
00:14:48.879 --> 00:14:58.159
I have a two-year-old and a four-year-old, so sleep was at a minimum, and yeah, it was it was a big pressure on yeah, the whole family.
00:14:58.240 --> 00:15:01.200
And balance balancing was there was no balance.
00:15:01.360 --> 00:15:02.240
There was no balance.
00:15:02.320 --> 00:15:11.600
And I think I had to be okay with that, and I had to make sure that my partner was okay with it too.
00:15:11.840 --> 00:15:25.200
And luckily, he realized he had to be, uh, and was actually incredibly supportive and let me have no balance for the first couple of months at the very least.
00:15:25.759 --> 00:15:26.080
Yeah.
00:15:26.559 --> 00:15:30.639
So you've taken sort of the company back, it's been super stressful.
00:15:30.960 --> 00:15:40.960
It's like you had these these sort of euphoric wins amongst sort of no sleep and long hours and sort of the consequences of this.
00:15:41.840 --> 00:15:47.200
But are you seeing at this stage, are you starting to see the business impact of these changes?
00:15:48.159 --> 00:16:08.879
Well, what was interesting is that it took a really long time to see any sort of impact because, as you say, I was spending so much of my time just trying to build the mechanisms that were going to allow me to run it by myself, that a lot of balls were being dropped dropped.
00:16:08.960 --> 00:16:14.399
So that 40% of the things that would grow the business weren't really happening.
00:16:14.960 --> 00:16:22.320
And you know, so throughout the whole of the summer, so the last member of staff left in June.
00:16:22.639 --> 00:16:41.919
And yeah, running the business myself, June, July, August, um, it was kind of under or hovering around the point where it needed to be to break even after losing all of the staff.
00:16:42.799 --> 00:16:53.600
And uh I remember October particularly, um, September, September was kind of okay, and then October was really bad for sales.
00:16:53.679 --> 00:17:03.519
And in October, I was like, you know what, I'm not sure this is gonna work because you know, I have stripped this company down to the bare bones.
00:17:03.679 --> 00:17:05.680
I got rid of so many apps.
00:17:05.839 --> 00:17:18.079
I got so like every cost I looked at, and I was like left with what, like insurance and the Shopify, you know, app itself, you know, I was like, and I'm barely paying myself.
00:17:18.160 --> 00:17:26.000
So I was like, I don't think I can strip it back anymore, and we're still not breaking even.
00:17:27.119 --> 00:17:39.519
And then um the one thing I had managed to keep going in in all of that time is to plan the um the Christmas sales and the and and and all of that.
00:17:39.680 --> 00:17:46.400
And as soon as I turned those on in November, things went completely bananas.
00:17:46.720 --> 00:17:51.200
And the you know, the sales absolutely shot up.
00:17:51.519 --> 00:18:09.039
I ran a really good um Black Friday campaign, and it it it completely flabbergasted me because Q4 2024 was up on Q4 2023 when I had all of those staff members.
00:18:09.440 --> 00:18:10.079
Wow.
00:18:10.720 --> 00:18:12.400
In terms of revenue.
00:18:12.799 --> 00:18:13.440
Wow.
00:18:13.680 --> 00:18:33.519
And we made a profit for the first time ever, and that was such a huge moment and just made me so happy just to know that the business could was capable of supporting itself just felt really huge.
00:18:34.319 --> 00:18:38.400
Um and and was a yeah, a moment of real pride.
00:18:38.559 --> 00:18:43.039
Um I was still working crazy, crazy hours.
00:18:43.200 --> 00:18:50.160
So then the the challenge became a different one, I think, at the beginning of the year, and it's still one I'm working through.
00:18:50.240 --> 00:18:57.920
So it's like, can the business grow and do well without me killing myself?
00:19:00.960 --> 00:19:15.440
And I think the answer is yes, but it needs more of that automation work to build more than just the stuff that's going to keep the business sticking along.
00:19:15.599 --> 00:19:30.240
It's now much more nuanced and actually much more complex automations that are needed, the ones that are going to really get the marketing side humming and get new products on the site and and all of that.
00:19:30.559 --> 00:19:36.079
And um, while that work has begun, there's more to do.
00:19:36.480 --> 00:19:44.240
I think for sort of founders listening, if I was like you you took this decision we talked about beginning, like almost the last moment possible.
00:19:44.400 --> 00:19:47.200
Like if you hadn't taken took it, there wouldn't have been a runway.
00:19:47.519 --> 00:19:47.839
Agreed.
00:19:48.160 --> 00:19:55.759
If you'd taken it maybe three months, six months earlier, you maybe couldn't have done because AI didn't exist in the same form.
00:19:55.920 --> 00:19:58.720
But let's say you were in that position today.
00:19:58.880 --> 00:20:05.039
And other founders are thinking about, okay, I've got a team of 14, 20, 25, like they're in the scale up phase.
00:20:05.440 --> 00:20:10.960
And there's this sense of, okay, am I am I scaling with adding more people in?
00:20:11.200 --> 00:20:14.799
Do I take the time to like think about automation?
00:20:15.039 --> 00:20:16.400
How would you do it?
00:20:16.640 --> 00:20:22.240
What would be your message to other founders if they were thinking about exploring this path?
00:20:22.799 --> 00:20:25.599
Don't underestimate the power of having people.
00:20:25.759 --> 00:20:28.480
I miss my people desperately.
00:20:28.960 --> 00:20:33.359
Um and I really wish I still had my team.
00:20:34.559 --> 00:20:38.480
However, uh there is so much you can do.
00:20:39.519 --> 00:20:55.200
Um and if those team members, as well as the founder, aren't looking into ways of increasing their productivity and you know, automating the things that computers are just great at.
00:20:55.440 --> 00:21:00.079
I I just feel like let the computers do the things that computers are good at.
00:21:00.319 --> 00:21:08.079
So taking data from one place, processing it in some way and putting it into another place, that's a computer's job.
00:21:08.400 --> 00:21:10.319
That that shouldn't be any human's job.
00:21:10.400 --> 00:21:11.920
It's boring, right?
00:21:12.160 --> 00:21:19.839
So I'm like, let AI take care of any and all of that immediately.
00:21:20.000 --> 00:21:24.960
And what you'll find is that you know, that frees up time for your team.
00:21:25.039 --> 00:21:28.400
And I I don't want to come in and be like, yeah, fire all your stuff.
00:21:28.480 --> 00:21:31.759
Um, it's great, uh, because that's that's not the case.
00:21:32.079 --> 00:21:43.759
But um there is so much that you can do, and you know the rate of progress, even since I started doing this, is immense.
00:21:44.000 --> 00:21:52.240
And the power of AI agents now, I just started building my own agents, which is slightly different from automations.
00:21:52.880 --> 00:21:59.759
Um, and what what they can do is is insane, and I think the rate.
00:22:00.559 --> 00:22:06.000
Of how good they are is doubling every month.
00:22:06.400 --> 00:22:06.960
Wow.
00:22:07.680 --> 00:22:08.960
I mean, it's insane.
00:22:09.200 --> 00:22:24.640
So I think that every founder will see cost savings, efficiency savings, like if you can keep your team due, but just scale much faster and more efficiently with AI.
00:22:24.799 --> 00:22:32.720
If you can't keep your team, then there are options for you, I guess, is what I'm saying.
00:22:33.440 --> 00:22:40.720
Because you did it in probably the most extreme way possible, which is like short sort of at the last minute, all team, all automation.
00:22:40.960 --> 00:22:41.039
Yeah.
00:22:41.200 --> 00:22:44.160
So it feels like you can do it earlier better.
00:22:44.480 --> 00:22:45.039
Yes.
00:22:45.359 --> 00:22:48.000
But it's it sounds like even if you're doing it earlier, it still can be disruptive.
00:22:48.240 --> 00:22:50.160
You've got the whole the whole team on board.
00:22:50.400 --> 00:22:51.039
Yeah.
00:22:51.440 --> 00:23:04.079
This process of or sort of conflict almost doing the day-to-day well and spending time, assigning time to automate to think it through.
00:23:04.480 --> 00:23:06.319
It feels like a quite a big piece of work.
00:23:06.400 --> 00:23:07.119
So I'm just wrong.
00:23:07.519 --> 00:23:13.599
That's the hardest thing, and it's that probably my biggest challenge currently is that there are things I still want to build.
00:23:13.759 --> 00:23:18.240
I have a list of post-its on my wall of automations that I want to build.
00:23:18.319 --> 00:23:28.240
And I'm kind of quietly confident that once those that list is built, that I'll have a semi-self-running business.
00:23:28.480 --> 00:23:34.079
However, each one takes a couple of weeks.
00:23:34.799 --> 00:23:38.400
And I still have to do everything else at the same time.
00:23:38.720 --> 00:23:46.000
So I would say that yeah, it it is about eking out that time.
00:23:46.079 --> 00:23:52.160
And it might be a question of being like, right, this is the automation that we're going to work on next.
00:23:52.240 --> 00:23:55.279
And this is, yeah, this is the time we're going to allocate to it.
00:23:55.359 --> 00:24:00.799
And it's going to take time out of the everyday, and something else might have to be dropped to make it happen.
00:24:01.039 --> 00:24:07.839
But it's about thinking forward to that, well, now we never have to do this again.
00:24:08.079 --> 00:24:10.960
And how much time saving does that take out of our week?
00:24:11.279 --> 00:24:13.440
And then how much more can we do with that time?
00:24:13.680 --> 00:24:19.039
So every automation takes you closer to being able to do more with less.
00:24:26.319 --> 00:24:41.920
But it sounds like you really feel this is a founder's sort of almost a key part of founder's role now to start making the time to explicitly allow team to take the time to think about how you can make how you can scale smarter with AI.
00:24:42.799 --> 00:24:54.240
I do think that if a founder isn't thinking about that, they're losing out massively because there will be other founders in their space who are, and it will allow them to scale faster and cheaper.
00:24:54.559 --> 00:24:57.599
So and then they'll be able to undercut them.
00:24:58.160 --> 00:25:09.839
And if you were to give like sort of a percentage of time that you'd allow for team or even for themselves, to go day-to-day AI, would you would you almost sound like you almost almost would ring fence some time?
00:25:10.160 --> 00:25:10.559
Definitely.
00:25:10.880 --> 00:25:13.279
What substantial time would you think you'd ring fence?
00:25:13.759 --> 00:25:21.440
I would say it depends on what that person's role is, um, you know, and how technical perhaps they are.
00:25:21.599 --> 00:25:53.119
You know, if, for example, I had a finance person and and they were strategic, it would be about kind of trying to download their their strategic thinking and their and their um assumptions and the way that they think about money and how they you know manipulate it with the their current tools to try and understand whether AI can take anything kind of manual or processing off their plate to give them more instantaneous insights and more instantaneous answers.
00:25:53.279 --> 00:26:14.880
So that, you know, for a you know, it could be just a question of a couple of hours of deep um interviewing of your CFO to try and really understand how they think about the money in your uh in your company and what would make their life quicker.
00:26:15.039 --> 00:26:23.279
So it might be that that person doesn't need to ring fence any huge amount of time and start building automations in their spare time.
00:26:23.440 --> 00:27:07.680
But then it might be that, you know, uh another member of staff, say, for example, um a researcher who's finding products to put to add onto the site, maybe they're the one that both show the interest and um and ability to figure out ways to get AI to help them and actually start to build the automations or agents that can go out into the world and and do that scraping and research for them to get, you know, to get the data to a stage where then that researcher takes those final steps to say, yes, this is a product that we want to bring on the site.
00:27:08.319 --> 00:27:21.920
Um, and so then maybe that's a question of saying, like, right, okay, so you, I now want you to, and I will, you know, help you to do it, I want you to watch these YouTube videos and I want you to build something that will do that.
00:27:22.319 --> 00:27:30.480
Um and then and then maybe say, look, I want you to re-re-ring fence you know, 20 to 30% of your day on to do this.
00:27:30.960 --> 00:27:47.359
And actually that could be quite significant, if not more, because you know, if there aren't lots of new products being added to the site all the time, that's not necessarily going to make the business slow down significantly.
00:27:47.839 --> 00:27:59.759
But if they build an automation that can speed up the uh onboarding of uh products by you know ten times, then that's incredibly significant.
00:28:01.359 --> 00:28:06.000
It sounds like part of this is you've got as a fan, you don't know, you don't know what you don't know.
00:28:06.319 --> 00:28:06.640
Of course.
00:28:06.880 --> 00:28:11.200
And like going back to this time, you you sort of you were ahead of the game because you're interested in it.
00:28:12.000 --> 00:28:18.880
If a fan was looking at this now, where would you suggest they go to to start getting their knowledge and raising it up?
00:28:19.680 --> 00:28:23.920
Um so I listen to the AI Daily Brief every day.
00:28:24.079 --> 00:28:26.960
It gives you the latest insights in AI.
00:28:27.039 --> 00:28:31.279
Um, I follow a lot of AI podcasts, and actually some of them are super fun.
00:28:31.759 --> 00:28:43.279
And what they'll do is they'll break down every time there's a new innovation in AI, you'll have people like live checking them out, showing you what they can do in your space.
00:28:43.359 --> 00:28:49.359
So, you know, AI for marketing podcasts, AI for e-commerce podcasts, and they'll show you what's possible.
00:28:49.920 --> 00:29:01.359
Um, on top of that, I would encourage everyone to look into either make.com, which is what I use, or Zapia, which is an automation platform.
00:29:02.000 --> 00:29:06.160
And that just allows the big bits of tech in your life to talk to each other.
00:29:06.240 --> 00:29:13.519
And for me, it still feels like magic because I could, you know, I built my own fitness app with it because it was fun, you know, and it's all just run off the spreadsheet.
00:29:13.599 --> 00:29:24.880
I put in what my fitness level was now and what I want it to be, and I'm a massive geek, so I created like 90 levels, uh, and they're all Harry Potter characters and Game of Thrones characters.
00:29:24.960 --> 00:29:28.559
So I start out as a pygmy puff and I end up as Balerian the Black Dread.
00:29:28.720 --> 00:29:50.720
And every morning my automation sends me an email with a workout that's only 15 minutes because that's all I have, and it gives me specific exercises, and you know, and it's silly and fun, but it it's genius what it allows you to do, because that's just spreadsheet talking to an AI module, talking to an email.
00:29:52.400 --> 00:30:01.680
And it's that interplay of data between the big bits of tech that you have in your life that allows for almost anything to happen.
00:30:02.400 --> 00:30:08.240
Like, because that could be your Shopify talking to ChatGPT, talking to Instagram.
00:30:08.319 --> 00:30:15.039
It could be your zero accounting talking to a spreadsheet, talking to Slack.
00:30:15.680 --> 00:30:17.519
It could be anything.
00:30:17.920 --> 00:30:26.400
And as soon as you've got an AI module in there to interpret the data going between bits of tech, then you have anything.
00:30:26.480 --> 00:30:28.559
You know, it could it can be anything.
00:30:28.880 --> 00:30:30.079
It's wonderful.
00:30:30.640 --> 00:30:32.799
I'm just really curious what your current character is.
00:30:33.119 --> 00:30:35.200
I am currently mowing metal.
00:30:35.680 --> 00:30:36.079
Yeah.
00:30:36.240 --> 00:30:43.119
I'm still on a on a bit of a low level, but um, yeah, wispy and you know, floating above a toilet.
00:30:43.200 --> 00:30:45.440
Um Where you want to be.
00:30:45.680 --> 00:30:50.559
But yeah, it gets me doing push-ups in the mornings, anything that does that is is a bonus.
00:30:51.039 --> 00:30:56.960
I think what what sounds here is there's there's there's almost so much stuff that you could do.
00:30:57.119 --> 00:31:03.359
What I really like about this is kind of you've managed to do it when you when you went through this this big change.
00:31:03.440 --> 00:31:06.160
You came into it with no coding experience.
00:31:06.480 --> 00:31:11.200
So you've managed to make it happen on the back foot.
00:31:12.000 --> 00:31:22.880
And a lot of people listening will have sort of will still have the team, they'll be on the front foot, and they'll have the capability to do it, and that they'll probably have less excuses to not do it.
00:31:23.200 --> 00:31:24.079
Yeah, I agreed.
00:31:24.319 --> 00:31:31.039
And it sounds really nice, like there are ways quite easy that you can educate yourself, there are ways that you can involve the team.
00:31:31.279 --> 00:31:42.640
It doesn't and actually it doesn't mean that you have to rip everything up, it's just really dedicating the time to understand what's possible and then to make sure you're having those conversations.
00:31:42.960 --> 00:31:43.440
Definitely.
00:31:43.599 --> 00:31:52.799
I mean, I think you know, the one thing I'd say is like sit with your team and do what I did when I realized I was gonna have to let everyone everyone go.
00:31:53.039 --> 00:31:55.920
What are they doing and how long is it taking them?
00:31:56.400 --> 00:31:58.079
And can it be automated?
00:31:59.279 --> 00:32:09.119
Because if it can, then it takes it takes something off their plate, and then they can do something more useful that only a human can do.
00:32:10.240 --> 00:32:12.880
And there are still those things that exist.
00:32:13.200 --> 00:32:30.240
So um, although I'm slightly worried that we're all gonna end up as hairdressers because at this rate, you know, pretty much you know, anything's gonna be um AIable, but it's um there's so much that can be done, and I, you know, people don't enjoy copying and pasting.
00:32:30.480 --> 00:32:36.640
Let the computers do that, at the very least, and then work up from there.
00:32:36.799 --> 00:32:49.759
You know, maybe make it fun, set a weekly little automation challenge, you know, find something that's really annoying in your working life and build an automation that does it for you.
00:00:04.160 --> 00:00:09.759
I'm excited to welcome Tara Button, best-selling author and founder and CEO of Buy Me Once.
00:00:10.080 --> 00:00:19.199
Once a recovering shopaholic, Tara built out a movement to find the longest-lasting products on the planet and change our throwaway culture forever.
00:00:19.440 --> 00:00:26.480
Her business idea went viral in 2016, and her book, A Lifeless Throwaway, became an Amazon bestseller.
00:00:29.839 --> 00:00:34.960
It had got to a point where I've been running this business for like eight years.
00:00:35.520 --> 00:00:37.759
We'd raised a whole lot of cash.
00:00:37.840 --> 00:00:40.960
You know, we had over time raised like three million pounds.
00:00:41.119 --> 00:00:43.759
The team grew to 14 at one point.
00:00:44.240 --> 00:00:49.359
And through COVID, obviously being in the e-commerce business, everyone's at home buying frying pans.
00:00:49.520 --> 00:00:50.240
Brilliant.
00:00:51.119 --> 00:01:00.640
But you know, over time, you know, the mechanics of the business changed, and we just weren't getting enough revenue in to cover our costs.
00:01:00.960 --> 00:01:13.359
And uh the point that I spoke to you at, we were facing not being able to raise any more capital and running out of money at at an alarming rate.
00:01:13.680 --> 00:01:20.239
And um I couldn't really figure out what my choices were at that point.
00:01:20.400 --> 00:01:22.719
I think that was part of the issue.
00:01:22.879 --> 00:01:31.920
I had so many competing thoughts and ideas in my head, and the idea of letting that mission go was abhorrent to me.
00:01:32.079 --> 00:01:36.239
Um, and at the same time, I felt this huge responsibility for shareholders.
00:01:36.400 --> 00:01:45.200
I also knew that, you know, the stress I was putting my family through wasn't great at that point, but I couldn't see through the mess to figure out what to do.
00:01:45.519 --> 00:02:08.719
I realized at the end of that conversation that the only choice was, if I was serious about fulfilling that mission, which is the clarity you gave me, was to let go of everyone in my company and run the entire business myself, which is what I then did that day.
00:02:08.879 --> 00:02:31.199
I had a meeting and I said, look, everyone, this isn't what I would have chosen, but the choice is either the business closes completely and I lose everything, or I let everyone go, semi-stopped paying myself pretty much, and just try and build again from the ground up something that can support itself.
00:02:31.840 --> 00:02:39.840
And so obviously that was a huge decision, and then the practicalities of that like kicked in, and that was pretty big as well.
00:02:40.240 --> 00:02:40.560
Wow.
00:02:40.639 --> 00:02:44.159
So I mean that I mean that is a that is a huge moment and a huge change.
00:02:44.479 --> 00:02:44.800
Yeah.
00:02:45.120 --> 00:02:50.080
I I think it's the reason for that that is, I think the choice probably would have been taken away, wouldn't it?
00:02:50.159 --> 00:02:57.840
It's like that sometimes we tell ourselves that there is a choice, and there is a choice unless we don't make it, and not making a choice is still a choice.
00:02:58.080 --> 00:02:58.479
Yeah.
00:02:58.639 --> 00:03:02.719
And that was what I had been doing for you know the months before that.
00:03:02.879 --> 00:03:12.159
And or not quite seeing what the choices were that I had to make, and so that moment of having that clarity, you know, changed everything.
00:03:12.240 --> 00:03:19.039
And I'm yeah, so grateful that I did that because you know, it's been incredibly difficult.
00:03:19.520 --> 00:03:24.719
However, you know, the team being these amazing people that they were were completely supportive.
00:03:24.800 --> 00:03:27.439
They knew that if I could have kept them, I would have done.
00:03:27.680 --> 00:03:36.639
And so there was no sense of resentment or uh, you know, child you know, churlishness.
00:03:36.800 --> 00:03:38.879
They were like, How can we help?
00:03:39.280 --> 00:03:45.280
How can we help make this okay for you to run the business completely single-handedly?
00:03:45.439 --> 00:03:55.199
And they taught me everything that they did, and they put in systems and um and culled things to make it work.
00:03:56.240 --> 00:04:05.280
I think I think this is really a fascinating sort of thing to dive into because you took a business with sort of 14 people and took it down to yourself.
00:04:05.520 --> 00:04:15.919
And uh a lot of a lot of people look at their business and go, oh, there's no like there's there's there's no like everyone's needed, everyone's necessary, um, and we really focus on scaling.
00:04:16.000 --> 00:04:22.240
And this idea of when to downsize feels almost like a step back or a mistake or a failure.
00:04:22.399 --> 00:04:23.040
Yeah.
00:04:24.079 --> 00:04:26.800
But I'd be I think it's really interesting to dive into like how you did this.
00:04:27.040 --> 00:04:29.680
If you can take from 14 to 1, yeah.
00:04:29.920 --> 00:04:33.199
What what what allowed you to to do that?
00:04:33.680 --> 00:04:34.720
Spreadsheets.
00:04:36.399 --> 00:04:37.680
You think I'm joking.
00:04:37.920 --> 00:04:57.600
Um no, I um I built a spreadsheet and I got everyone to write down every single task that they did, and whether it was daily, weekly, monthly, yearly, how much time it took them and how um needed it was.
00:04:57.920 --> 00:05:08.720
I also then you know went through and I was like, well, these are the completely business critical operations that have to happen, whether it's daily or weekly.
00:05:08.879 --> 00:05:12.480
Without this, the business will crumble.
00:05:12.720 --> 00:05:37.439
Like, you know, if the customers stop getting their products, if the suppliers stop getting paid, if customers stop being replied to, very soon you have one Star Trust pilot review saying, You're a scam, very soon you'd have suppliers pulling out, and very soon, you know, obviously if you don't send a product, you'd probably be facing some sort of legal action.
00:05:37.680 --> 00:05:40.480
So those were the things that I had to solve first.
00:05:40.879 --> 00:05:51.680
And so I very quickly went about building automations, essentially robots that could replicate those tasks.
00:05:52.720 --> 00:05:58.160
But before we jump into the sort of road to the automation, I'd like I mean that's a fascinating process too.
00:05:58.319 --> 00:06:02.480
I mean, fascinating moment with the right word, but for that spreadsheet, I'm curious.
00:06:02.560 --> 00:06:08.639
Like, having done that, what percentage of the tasks do you reckon you were left with as critical?
00:06:08.879 --> 00:06:14.720
Because often we build businesses, they they morph into like more and more actions and projects and stuff.
00:06:14.800 --> 00:06:14.959
Yeah.
00:06:15.120 --> 00:06:19.680
I'm really curious, like, what percentage of those things were you left with when you had to cut it down?
00:06:19.920 --> 00:06:24.879
And picturing the spreadsheet, now I think it was surprisingly fewer than I thought.
00:06:25.759 --> 00:06:35.120
And but I would like to caveat that in that these were the operations without which the business might begin to fail within a couple of weeks.
00:06:35.920 --> 00:06:53.920
Uh then there was the kind of second tier where if I didn't do any of this, the business would just die a very slow death because no marketing emails are being sent out, no social media is being sent out, no new products are being put on the site, you know.
00:06:54.160 --> 00:06:59.120
So you might not die immediately, but it would be a slow death over months.
00:06:59.439 --> 00:07:06.399
So I dealt with the first tier first, and then I had to kind of work on the second tier.
00:07:06.480 --> 00:07:14.319
But it is it did surprise me going through the whole list, being like, nope, nope, nope, nope, nope.
00:07:14.560 --> 00:07:16.319
Yeah, it it was surprising.
00:07:16.879 --> 00:07:20.000
So gut feel, let's say the first list die in two weeks.
00:07:20.399 --> 00:07:22.079
That was about maybe 10%.
00:07:22.639 --> 00:07:25.279
And then the sort of the slow slow death pile.
00:07:25.600 --> 00:07:32.399
The slow death pile was probably another that was more like 40 to 50 percent.
00:07:32.560 --> 00:07:38.959
That was quite substantial stuff, but it didn't all have to happen at once.
00:07:39.120 --> 00:07:52.000
It was like some things might mitigate other things, but they would still be in that pile as long as enough of that stuff was happening, then you'd probably mitigate like degrowth, if that makes sense.
00:07:52.480 --> 00:07:55.680
So kind of you could roughly half what was being done before.
00:07:56.000 --> 00:07:56.160
Yeah.
00:07:56.399 --> 00:07:59.600
That still takes down to about seven people, let's say.
00:08:00.800 --> 00:08:02.720
Then you've got to you've got to make yourself the seven people.
00:08:02.800 --> 00:08:06.160
So yeah, sounds like automation was the route that you chose.
00:08:06.959 --> 00:08:20.720
Yeah, and I think that's an important point to make is that if we had had that conversation a year pr earlier than that, so we had it in March 2024, if it had been March 2023, I don't think my choices would have been the same.
00:08:21.519 --> 00:08:26.959
And that's because of AI and automation and what it allows you to do.
00:08:27.279 --> 00:08:43.279
Uh when I realized that AI can code for you, and essentially the ability to code means that you can build yourself little apps and automations that can essentially do anything as long as it's in a digital form.
00:08:43.440 --> 00:08:45.440
And obviously, we're a digital site.
00:08:45.679 --> 00:08:50.879
That opened the doorway to me believing that this was possible.
00:08:52.159 --> 00:08:57.120
But open AI hadn't, I mean, because I remember I remember this conversation, I remember so the trick trigger.
00:08:58.240 --> 00:09:02.240
Now it's kind of like a take taken as red thing that you can do this stuff.
00:09:03.679 --> 00:09:14.480
OpenAI wasn't talked about when we were, and I know it's more than OpenAI this, but like it feels like you were ahead of the curve on this because it hadn't really hit public consciousness in the same way that it has now.
00:09:14.799 --> 00:09:16.240
Yes, I think that's true.
00:09:16.399 --> 00:09:27.919
I think I'd just begun, I think maybe even even in like the couple of weeks before we talked to try and figure out solutions and to try and figure out what was possible.
00:09:28.159 --> 00:09:53.840
And I was only just beginning to scrape the surface, but I'd seen a couple of YouTube videos that had slightly blown my mind in terms of the automation space, and I and that had kind of sent my brain into a kind of like, oh well, that's possible, then this could be possible, and if that's possible, then that could be possible, and that I think gave me the I guess the courage to take the plunge.
00:09:54.399 --> 00:10:08.879
Otherwise, I I think if it had been a choice between um closing the business and running everything myself manually, it wouldn't have been the same choice.
00:10:10.399 --> 00:10:11.600
Nowhere close.
00:10:12.879 --> 00:10:17.039
So you so you you've got then this this sort of seven perils worth of tasks.
00:10:17.200 --> 00:10:25.679
You start on it, how how do you for other people that maybe you think about doing this, you've got your spreadsheet, you've got your task, what's what now?
00:10:26.399 --> 00:10:27.840
So I I went through the list.
00:10:28.000 --> 00:10:32.240
So the first thing was um inventory, I think.
00:10:32.399 --> 00:10:47.679
So if products are out of stock with a uh you know, with a brand, but we don't know that they're out of stock, then we're selling products that don't exist, which is obviously gonna very quickly start to cause some issues.
00:10:47.919 --> 00:11:02.799
And so trying to build something that would take our inventory um and automate it to the point where I didn't have to touch it was probably one of the biggest challenges.
00:11:03.039 --> 00:11:21.840
Um the way that we received inventory uh from our brands was in such a weird and wonderful variety of ways, from somebody sending a spreadsheet, someone sending a CSV, someone just sending an email saying, hey, we've only got two of these left.
00:11:22.080 --> 00:11:24.559
So, but it was generally all through email.
00:11:24.639 --> 00:11:48.159
So I built an email scraper for the brands that sent inventory in that way, which allowed it to scrape my email, take all of the information, no matter what format it was in, and put it into a Google Sheet that then normalized it against all of the products we have.
00:11:48.320 --> 00:12:03.840
And then as soon as that was done, we had the correct information, and then I used um make.com uh to automate that spreadsheet to take any change in inventory directly into Shopify.
00:12:04.399 --> 00:12:14.720
So that took I want to say a week and a half to build, and it was very frustrating because I can't code at all.
00:12:15.200 --> 00:12:48.879
So um I would get ChatGPT to write the code and it would half do it, uh, but then there would be errors, and I would the the process was um write the code, it would write the code, I'd copy and paste it, run it, it would half do something, but probably error out, and then I'm copy and pasting the errors back into ChatGPT and saying it's doing this, and then it'd be like, I will rewrite the code for you, and then it would, and then I'd paste it back, and it was this.
00:12:49.039 --> 00:12:59.200
And sometimes what would happen is that it would do something perfectly, but then there'd be another error, and then I'd paste it back, and then it would stop doing the thing perfectly.
00:12:59.600 --> 00:13:11.039
And that was the most frustrating thing, is it because I can't code, so I have no idea what's changed, and I'm for you know completely reliant on it doing the right thing.
00:13:11.759 --> 00:13:15.440
And you know, it wasn't an easy process.
00:13:15.600 --> 00:13:27.519
Um, it was quite addictive, I would say, because you never knew which, you know, the moment that it worked, it felt like magic.
00:13:27.600 --> 00:13:42.320
It was like, you know, when I could see like all of the spreadsheets and all of the data coming in perfectly and pinging into Shopify, and I'm like, I never have to do this manually again.
00:13:42.639 --> 00:13:46.639
That moment, if it feels like a miracle.
00:13:47.120 --> 00:13:57.279
You're having this time pressure looming over you of like just the doing, and then you've got to commit all this time to plan and design and go through this process.
00:13:57.679 --> 00:13:58.320
Yeah.
00:13:58.639 --> 00:13:59.840
It feels like a lot.
00:14:00.159 --> 00:14:01.679
It I mean, it was insane.
00:14:01.840 --> 00:14:05.840
I was working 14, 60 nowadays.
00:14:05.919 --> 00:14:07.679
I lost all the skin on my fingers.
00:14:07.840 --> 00:14:15.440
I had to wear gaming gloves because I was on my laptop so much it was I lost about a stone and a half.
00:14:15.519 --> 00:14:18.080
Uh so all the baby weight was gone, so that was quite handy.
00:14:18.240 --> 00:14:20.159
Uh, but not in a healthy way.
00:14:20.720 --> 00:14:22.720
I mean, it was it was bad.
00:14:22.879 --> 00:14:33.120
I didn't want to sugarcoat it and say, like, oh, you know, I'm, you know, got rid of all my stuff and you know, I didn't have robots from my company, and it's all hunky-dory.
00:14:33.360 --> 00:14:37.440
It was incredibly painful and a lot of work.
00:14:38.399 --> 00:14:46.000
The breakthroughs and the victories when they happened were spectacular and gave me the strength to carry on doing it.
00:14:46.159 --> 00:14:48.799
Um, but it wasn't without sacrifice.
00:14:48.879 --> 00:14:58.159
I have a two-year-old and a four-year-old, so sleep was at a minimum, and yeah, it was it was a big pressure on yeah, the whole family.
00:14:58.240 --> 00:15:01.200
And balance balancing was there was no balance.
00:15:01.360 --> 00:15:02.240
There was no balance.
00:15:02.320 --> 00:15:11.600
And I think I had to be okay with that, and I had to make sure that my partner was okay with it too.
00:15:11.840 --> 00:15:25.200
And luckily, he realized he had to be, uh, and was actually incredibly supportive and let me have no balance for the first couple of months at the very least.
00:15:25.759 --> 00:15:26.080
Yeah.
00:15:26.559 --> 00:15:30.639
So you've taken sort of the company back, it's been super stressful.
00:15:30.960 --> 00:15:40.960
It's like you had these these sort of euphoric wins amongst sort of no sleep and long hours and sort of the consequences of this.
00:15:41.840 --> 00:15:47.200
But are you seeing at this stage, are you starting to see the business impact of these changes?
00:15:48.159 --> 00:16:08.879
Well, what was interesting is that it took a really long time to see any sort of impact because, as you say, I was spending so much of my time just trying to build the mechanisms that were going to allow me to run it by myself, that a lot of balls were being dropped dropped.
00:16:08.960 --> 00:16:14.399
So that 40% of the things that would grow the business weren't really happening.
00:16:14.960 --> 00:16:22.320
And you know, so throughout the whole of the summer, so the last member of staff left in June.
00:16:22.639 --> 00:16:41.919
And yeah, running the business myself, June, July, August, um, it was kind of under or hovering around the point where it needed to be to break even after losing all of the staff.
00:16:42.799 --> 00:16:53.600
And uh I remember October particularly, um, September, September was kind of okay, and then October was really bad for sales.
00:16:53.679 --> 00:17:03.519
And in October, I was like, you know what, I'm not sure this is gonna work because you know, I have stripped this company down to the bare bones.
00:17:03.679 --> 00:17:05.680
I got rid of so many apps.
00:17:05.839 --> 00:17:18.079
I got so like every cost I looked at, and I was like left with what, like insurance and the Shopify, you know, app itself, you know, I was like, and I'm barely paying myself.
00:17:18.160 --> 00:17:26.000
So I was like, I don't think I can strip it back anymore, and we're still not breaking even.
00:17:27.119 --> 00:17:39.519
And then um the one thing I had managed to keep going in in all of that time is to plan the um the Christmas sales and the and and and all of that.
00:17:39.680 --> 00:17:46.400
And as soon as I turned those on in November, things went completely bananas.
00:17:46.720 --> 00:17:51.200
And the you know, the sales absolutely shot up.
00:17:51.519 --> 00:18:09.039
I ran a really good um Black Friday campaign, and it it it completely flabbergasted me because Q4 2024 was up on Q4 2023 when I had all of those staff members.
00:18:09.440 --> 00:18:10.079
Wow.
00:18:10.720 --> 00:18:12.400
In terms of revenue.
00:18:12.799 --> 00:18:13.440
Wow.
00:18:13.680 --> 00:18:33.519
And we made a profit for the first time ever, and that was such a huge moment and just made me so happy just to know that the business could was capable of supporting itself just felt really huge.
00:18:34.319 --> 00:18:38.400
Um and and was a yeah, a moment of real pride.
00:18:38.559 --> 00:18:43.039
Um I was still working crazy, crazy hours.
00:18:43.200 --> 00:18:50.160
So then the the challenge became a different one, I think, at the beginning of the year, and it's still one I'm working through.
00:18:50.240 --> 00:18:57.920
So it's like, can the business grow and do well without me killing myself?
00:19:00.960 --> 00:19:15.440
And I think the answer is yes, but it needs more of that automation work to build more than just the stuff that's going to keep the business sticking along.
00:19:15.599 --> 00:19:30.240
It's now much more nuanced and actually much more complex automations that are needed, the ones that are going to really get the marketing side humming and get new products on the site and and all of that.
00:19:30.559 --> 00:19:36.079
And um, while that work has begun, there's more to do.
00:19:36.480 --> 00:19:44.240
I think for sort of founders listening, if I was like you you took this decision we talked about beginning, like almost the last moment possible.
00:19:44.400 --> 00:19:47.200
Like if you hadn't taken took it, there wouldn't have been a runway.
00:19:47.519 --> 00:19:47.839
Agreed.
00:19:48.160 --> 00:19:55.759
If you'd taken it maybe three months, six months earlier, you maybe couldn't have done because AI didn't exist in the same form.
00:19:55.920 --> 00:19:58.720
But let's say you were in that position today.
00:19:58.880 --> 00:20:05.039
And other founders are thinking about, okay, I've got a team of 14, 20, 25, like they're in the scale up phase.
00:20:05.440 --> 00:20:10.960
And there's this sense of, okay, am I am I scaling with adding more people in?
00:20:11.200 --> 00:20:14.799
Do I take the time to like think about automation?
00:20:15.039 --> 00:20:16.400
How would you do it?
00:20:16.640 --> 00:20:22.240
What would be your message to other founders if they were thinking about exploring this path?
00:20:22.799 --> 00:20:25.599
Don't underestimate the power of having people.
00:20:25.759 --> 00:20:28.480
I miss my people desperately.
00:20:28.960 --> 00:20:33.359
Um and I really wish I still had my team.
00:20:34.559 --> 00:20:38.480
However, uh there is so much you can do.
00:20:39.519 --> 00:20:55.200
Um and if those team members, as well as the founder, aren't looking into ways of increasing their productivity and you know, automating the things that computers are just great at.
00:20:55.440 --> 00:21:00.079
I I just feel like let the computers do the things that computers are good at.
00:21:00.319 --> 00:21:08.079
So taking data from one place, processing it in some way and putting it into another place, that's a computer's job.
00:21:08.400 --> 00:21:10.319
That that shouldn't be any human's job.
00:21:10.400 --> 00:21:11.920
It's boring, right?
00:21:12.160 --> 00:21:19.839
So I'm like, let AI take care of any and all of that immediately.
00:21:20.000 --> 00:21:24.960
And what you'll find is that you know, that frees up time for your team.
00:21:25.039 --> 00:21:28.400
And I I don't want to come in and be like, yeah, fire all your stuff.
00:21:28.480 --> 00:21:31.759
Um, it's great, uh, because that's that's not the case.
00:21:32.079 --> 00:21:43.759
But um there is so much that you can do, and you know the rate of progress, even since I started doing this, is immense.
00:21:44.000 --> 00:21:52.240
And the power of AI agents now, I just started building my own agents, which is slightly different from automations.
00:21:52.880 --> 00:21:59.759
Um, and what what they can do is is insane, and I think the rate.
00:22:00.559 --> 00:22:06.000
Of how good they are is doubling every month.
00:22:06.400 --> 00:22:06.960
Wow.
00:22:07.680 --> 00:22:08.960
I mean, it's insane.
00:22:09.200 --> 00:22:24.640
So I think that every founder will see cost savings, efficiency savings, like if you can keep your team due, but just scale much faster and more efficiently with AI.
00:22:24.799 --> 00:22:32.720
If you can't keep your team, then there are options for you, I guess, is what I'm saying.
00:22:33.440 --> 00:22:40.720
Because you did it in probably the most extreme way possible, which is like short sort of at the last minute, all team, all automation.
00:22:40.960 --> 00:22:41.039
Yeah.
00:22:41.200 --> 00:22:44.160
So it feels like you can do it earlier better.
00:22:44.480 --> 00:22:45.039
Yes.
00:22:45.359 --> 00:22:48.000
But it's it sounds like even if you're doing it earlier, it still can be disruptive.
00:22:48.240 --> 00:22:50.160
You've got the whole the whole team on board.
00:22:50.400 --> 00:22:51.039
Yeah.
00:22:51.440 --> 00:23:04.079
This process of or sort of conflict almost doing the day-to-day well and spending time, assigning time to automate to think it through.
00:23:04.480 --> 00:23:06.319
It feels like a quite a big piece of work.
00:23:06.400 --> 00:23:07.119
So I'm just wrong.
00:23:07.519 --> 00:23:13.599
That's the hardest thing, and it's that probably my biggest challenge currently is that there are things I still want to build.
00:23:13.759 --> 00:23:18.240
I have a list of post-its on my wall of automations that I want to build.
00:23:18.319 --> 00:23:28.240
And I'm kind of quietly confident that once those that list is built, that I'll have a semi-self-running business.
00:23:28.480 --> 00:23:34.079
However, each one takes a couple of weeks.
00:23:34.799 --> 00:23:38.400
And I still have to do everything else at the same time.
00:23:38.720 --> 00:23:46.000
So I would say that yeah, it it is about eking out that time.
00:23:46.079 --> 00:23:52.160
And it might be a question of being like, right, this is the automation that we're going to work on next.
00:23:52.240 --> 00:23:55.279
And this is, yeah, this is the time we're going to allocate to it.
00:23:55.359 --> 00:24:00.799
And it's going to take time out of the everyday, and something else might have to be dropped to make it happen.
00:24:01.039 --> 00:24:07.839
But it's about thinking forward to that, well, now we never have to do this again.
00:24:08.079 --> 00:24:10.960
And how much time saving does that take out of our week?
00:24:11.279 --> 00:24:13.440
And then how much more can we do with that time?
00:24:13.680 --> 00:24:19.039
So every automation takes you closer to being able to do more with less.
00:24:26.319 --> 00:24:41.920
But it sounds like you really feel this is a founder's sort of almost a key part of founder's role now to start making the time to explicitly allow team to take the time to think about how you can make how you can scale smarter with AI.
00:24:42.799 --> 00:24:54.240
I do think that if a founder isn't thinking about that, they're losing out massively because there will be other founders in their space who are, and it will allow them to scale faster and cheaper.
00:24:54.559 --> 00:24:57.599
So and then they'll be able to undercut them.
00:24:58.160 --> 00:25:09.839
And if you were to give like sort of a percentage of time that you'd allow for team or even for themselves, to go day-to-day AI, would you would you almost sound like you almost almost would ring fence some time?
00:25:10.160 --> 00:25:10.559
Definitely.
00:25:10.880 --> 00:25:13.279
What substantial time would you think you'd ring fence?
00:25:13.759 --> 00:25:21.440
I would say it depends on what that person's role is, um, you know, and how technical perhaps they are.
00:25:21.599 --> 00:25:53.119
You know, if, for example, I had a finance person and and they were strategic, it would be about kind of trying to download their their strategic thinking and their and their um assumptions and the way that they think about money and how they you know manipulate it with the their current tools to try and understand whether AI can take anything kind of manual or processing off their plate to give them more instantaneous insights and more instantaneous answers.
00:25:53.279 --> 00:26:14.880
So that, you know, for a you know, it could be just a question of a couple of hours of deep um interviewing of your CFO to try and really understand how they think about the money in your uh in your company and what would make their life quicker.
00:26:15.039 --> 00:26:23.279
So it might be that that person doesn't need to ring fence any huge amount of time and start building automations in their spare time.
00:26:23.440 --> 00:27:07.680
But then it might be that, you know, uh another member of staff, say, for example, um a researcher who's finding products to put to add onto the site, maybe they're the one that both show the interest and um and ability to figure out ways to get AI to help them and actually start to build the automations or agents that can go out into the world and and do that scraping and research for them to get, you know, to get the data to a stage where then that researcher takes those final steps to say, yes, this is a product that we want to bring on the site.
00:27:08.319 --> 00:27:21.920
Um, and so then maybe that's a question of saying, like, right, okay, so you, I now want you to, and I will, you know, help you to do it, I want you to watch these YouTube videos and I want you to build something that will do that.
00:27:22.319 --> 00:27:30.480
Um and then and then maybe say, look, I want you to re-re-ring fence you know, 20 to 30% of your day on to do this.
00:27:30.960 --> 00:27:47.359
And actually that could be quite significant, if not more, because you know, if there aren't lots of new products being added to the site all the time, that's not necessarily going to make the business slow down significantly.
00:27:47.839 --> 00:27:59.759
But if they build an automation that can speed up the uh onboarding of uh products by you know ten times, then that's incredibly significant.
00:28:01.359 --> 00:28:06.000
It sounds like part of this is you've got as a fan, you don't know, you don't know what you don't know.
00:28:06.319 --> 00:28:06.640
Of course.
00:28:06.880 --> 00:28:11.200
And like going back to this time, you you sort of you were ahead of the game because you're interested in it.
00:28:12.000 --> 00:28:18.880
If a fan was looking at this now, where would you suggest they go to to start getting their knowledge and raising it up?
00:28:19.680 --> 00:28:23.920
Um so I listen to the AI Daily Brief every day.
00:28:24.079 --> 00:28:26.960
It gives you the latest insights in AI.
00:28:27.039 --> 00:28:31.279
Um, I follow a lot of AI podcasts, and actually some of them are super fun.
00:28:31.759 --> 00:28:43.279
And what they'll do is they'll break down every time there's a new innovation in AI, you'll have people like live checking them out, showing you what they can do in your space.
00:28:43.359 --> 00:28:49.359
So, you know, AI for marketing podcasts, AI for e-commerce podcasts, and they'll show you what's possible.
00:28:49.920 --> 00:29:01.359
Um, on top of that, I would encourage everyone to look into either make.com, which is what I use, or Zapia, which is an automation platform.
00:29:02.000 --> 00:29:06.160
And that just allows the big bits of tech in your life to talk to each other.
00:29:06.240 --> 00:29:13.519
And for me, it still feels like magic because I could, you know, I built my own fitness app with it because it was fun, you know, and it's all just run off the spreadsheet.
00:29:13.599 --> 00:29:24.880
I put in what my fitness level was now and what I want it to be, and I'm a massive geek, so I created like 90 levels, uh, and they're all Harry Potter characters and Game of Thrones characters.
00:29:24.960 --> 00:29:28.559
So I start out as a pygmy puff and I end up as Balerian the Black Dread.
00:29:28.720 --> 00:29:50.720
And every morning my automation sends me an email with a workout that's only 15 minutes because that's all I have, and it gives me specific exercises, and you know, and it's silly and fun, but it it's genius what it allows you to do, because that's just spreadsheet talking to an AI module, talking to an email.
00:29:52.400 --> 00:30:01.680
And it's that interplay of data between the big bits of tech that you have in your life that allows for almost anything to happen.
00:30:02.400 --> 00:30:08.240
Like, because that could be your Shopify talking to ChatGPT, talking to Instagram.
00:30:08.319 --> 00:30:15.039
It could be your zero accounting talking to a spreadsheet, talking to Slack.
00:30:15.680 --> 00:30:17.519
It could be anything.
00:30:17.920 --> 00:30:26.400
And as soon as you've got an AI module in there to interpret the data going between bits of tech, then you have anything.
00:30:26.480 --> 00:30:28.559
You know, it could it can be anything.
00:30:28.880 --> 00:30:30.079
It's wonderful.
00:30:30.640 --> 00:30:32.799
I'm just really curious what your current character is.
00:30:33.119 --> 00:30:35.200
I am currently mowing metal.
00:30:35.680 --> 00:30:36.079
Yeah.
00:30:36.240 --> 00:30:43.119
I'm still on a on a bit of a low level, but um, yeah, wispy and you know, floating above a toilet.
00:30:43.200 --> 00:30:45.440
Um Where you want to be.
00:30:45.680 --> 00:30:50.559
But yeah, it gets me doing push-ups in the mornings, anything that does that is is a bonus.
00:30:51.039 --> 00:30:56.960
I think what what sounds here is there's there's there's almost so much stuff that you could do.
00:30:57.119 --> 00:31:03.359
What I really like about this is kind of you've managed to do it when you when you went through this this big change.
00:31:03.440 --> 00:31:06.160
You came into it with no coding experience.
00:31:06.480 --> 00:31:11.200
So you've managed to make it happen on the back foot.
00:31:12.000 --> 00:31:22.880
And a lot of people listening will have sort of will still have the team, they'll be on the front foot, and they'll have the capability to do it, and that they'll probably have less excuses to not do it.
00:31:23.200 --> 00:31:24.079
Yeah, I agreed.
00:31:24.319 --> 00:31:31.039
And it sounds really nice, like there are ways quite easy that you can educate yourself, there are ways that you can involve the team.
00:31:31.279 --> 00:31:42.640
It doesn't and actually it doesn't mean that you have to rip everything up, it's just really dedicating the time to understand what's possible and then to make sure you're having those conversations.
00:31:42.960 --> 00:31:43.440
Definitely.
00:31:43.599 --> 00:31:52.799
I mean, I think you know, the one thing I'd say is like sit with your team and do what I did when I realized I was gonna have to let everyone everyone go.
00:31:53.039 --> 00:31:55.920
What are they doing and how long is it taking them?
00:31:56.400 --> 00:31:58.079
And can it be automated?
00:31:59.279 --> 00:32:09.119
Because if it can, then it takes it takes something off their plate, and then they can do something more useful that only a human can do.
00:32:10.240 --> 00:32:12.880
And there are still those things that exist.
00:32:13.200 --> 00:32:30.240
So um, although I'm slightly worried that we're all gonna end up as hairdressers because at this rate, you know, pretty much you know, anything's gonna be um AIable, but it's um there's so much that can be done, and I, you know, people don't enjoy copying and pasting.
00:32:30.480 --> 00:32:36.640
Let the computers do that, at the very least, and then work up from there.
00:32:36.799 --> 00:32:49.759
You know, maybe make it fun, set a weekly little automation challenge, you know, find something that's really annoying in your working life and build an automation that does it for you.
