Latest / eCommerce Podcast / Turn Boring Product Listings into James Bond Adventures with AI | Matt Anderson

Transcript

Turn Boring Product Listings into James Bond Adventures with AI | Matt Anderson · eCommerce Podcast

0:04
Well, hello, I'm Matt Edmundson and you are listening to the eCommerce Podcast.
0:10
Uh, now I've been around eCommerce since 2002, which I appreciate
0:13
is a very, very long time.
0:15
And during that time, you know what the biggest challenge I faced, it wasn't
0:19
finding products or building websites.
0:21
It was finding.
0:22
The right partners, the right people who could help me grow
0:26
beyond my own limitation.
0:28
And that's why these days I partner with eCommerce brands to
0:31
help them grow, scale and exit.
0:33
And if you'd like to know more about that and more about how that works,
0:36
just head over to the website eCommerce Podcast dot net 'cause it's all there.
0:40
Uh, come join it.
0:41
Come say, how's it now?
0:42
Today I'm joined by Matt Anderson from Dragonfly.
0:47
AI and we're gonna be looking at this fascinating place where e-commerce,
0:52
AI, and brand all collide together now.
0:56
Ai.
0:57
Think about it, right?
0:58
You've used it, you've used chat.
0:59
GPT, if you're like me, you're a big fan of Claude.
1:02
How many of you have done that thing where you've just asked chat, GPT or
1:07
Claude for a, for a piece of content, uh, and they give it to you and that's it.
1:13
You just use that.
1:14
There's no, there's no filter.
1:17
There's no quality control.
1:18
We all know that there should be one, right?
1:20
We genuinely do.
1:21
But we just don't do it mainly 'cause we're time conscious.
1:23
Well, that's what we're gonna get into, uh, today, all this kind of stuff.
1:27
Matt, I'm curious, when did you first sort of come across this as a
1:32
big problem that people were having?
1:35
Um, well thank you for, for having me on, first and foremost.
1:38
Um.
1:39
I think well o over the last sort of year to 18 months, um, I mean,
1:44
AI first and foremost rolled back.
1:46
AI has been around for a while now.
1:47
Um, you know, a,
1:48
a good decade or so, anyone that's used I. To auto correct
1:51
on, on their, on their phone or
1:53
Grammarly and those kind of things.
1:55
I mean, that's
1:55
all type
1:57
type of AI uh, you could say.
1:58
So, but I mean, I think, I guess really very much in the last sort of.
2:03
Six to 12 months, the world of sort of chat GPTs become on, on, on the,
2:08
the mainstream conscious really.
2:10
Um, and then all the other plethora of, of ai, has roll rolled on from that.
2:16
I know, we are at the point that Pandora's box is open.
2:20
Um, and perhaps I think people are starting to understand the power of ai.
2:25
Um, and I think, but also, you know, there's, there's skepticism
2:29
and, and also, um, some.
2:31
Somewhat of a fear and trepidation.
2:33
Anyone who's seen Trump's video recently about Garza know exactly
2:36
what I'm talking about, you know?
2:38
So there's that, issue there that not only, not only what, you know,
2:44
the content that we're pushing out, but the content we're consuming.
2:47
There's there's the lack of a filter really
2:49
at the moment.
2:50
I think, um, uh, governments and organizations are, are, are, you know,
2:54
I think it's fair enough, uh, are are playing catch up there regarding that.
2:57
Um.
2:58
What I would say is ai, um, and, and, and we're, and I'll talk about
3:04
different types of ai, but AI at the moment, um, is very much dependent on
3:09
how quality of the data that you're putting in and also the quality of
3:12
the prompt engineering and this term
3:14
about prompting.
3:16
How
3:16
good is the prompt that you're giving, giving the, the AI is very
3:20
much dependent on the quality of content you get, you're getting out.
3:24
Um, but also many organizations, um, and as they say, and the world of
3:28
sellers and, and, and, and eCommerce sort of small one man bands and
3:32
what have you or one woman bands
3:33
are,
3:35
uh, not putting the checks and balances in place about the contents coming out.
3:39
Um, 'cause certain AI can hallucinate, as they say in the traders though,
3:45
it'll come to a point where it can't really find the facts or, or understand
3:49
the subject and it'll make a fact up,
3:51
you know,
3:51
um, and a percentage up and what have you.
3:54
And certain AI out there is better at qualifying it,
3:57
like perplexity for example.
3:58
It'll reference,
3:59
you know, it'll give you a statin and give you a reference, but, you know,
4:03
that is a limitation of, of, of chat, GPT and, and, and also clawed as well, so.
4:10
We need to have those sort of checks and balances.
4:12
And the way I always refer to it's like, if you're using ai, AI as a copilot, it
4:17
is like a, um, for both fans by the way.
4:20
But I would say like, this is like having a, a, master's graduate
4:25
next year that's just entered your business, but they don't, they don't
4:28
know all the facts about your product and what you do and what you sell.
4:32
So you have to sort of have that feedback loop really with,
4:34
with with the ai.
4:36
Just to just double check really.
4:38
Um, So having sys good, good systems and sos in place is, is one thing really.
4:42
Um, um, I very much so to, to, to
4:45
be aware of.
4:49
Yeah.
4:49
isn't it, that we, um, because I was talking to someone earlier,
4:52
we're doing a website review
4:54
Hmm.
4:55
looking at the website and I'm kinda like, I'm reading the content.
4:59
I'm like, did AI write this?
5:01
You and,
5:02
and I feel like we've got to the phase where actually humans are now
5:07
starting to quite quickly and easily recognize when AI creates content.
5:13
Um, there, there's certain tells isn't there?
5:16
Like if you ask it to write a headline, it will always write a statement with
5:20
a colon followed by another statement and it it, it drives me absolutely nuts.
5:24
Right?
5:24
And we can, and you can see it straight off the bat.
5:27
And
5:27
Yes.
5:27
this case of whenever it was 18 months ago in chat, GPT
5:32
sort of exploded on the scene.
5:33
You're like, flip at it, can write all this amazing content.
5:37
And it did, but it's very AI content.
5:41
And actually now I think not only can computer spot AI content, but we
5:46
where humans are very good at sort of subconsciously spotting AI content.
5:51
that a bad thing?
5:53
yes, but I think possibly no.
5:56
But I think like the website I was referring to, uh, looking at this
5:59
morning, it just felt all a bit dry, if I'm honest with you, Matt, and all a
6:03
Yes.
6:04
and all a bit dull and I think I. This seems to be with
6:09
the default kind of prompts.
6:11
Like if you go to chat GP, T and C, give me some product copy for this widget.
6:17
It's just gonna give you fairly average content because it's taken
6:21
the sum of all the parts from the website to create that content.
6:24
I would've thought.
6:25
Yeah, I think it's, uh, it is referencing the, the content
6:28
within the large language model.
6:29
Essentially a lot of the, um.
6:33
Generative ai, which is like, if you look at step by step, by
6:37
step, it it's basically learning through, uh, machine learning.
6:41
So it is going, oh, well you said this and you said that,
6:44
so you most likely mean this.
6:46
Uh, or it's going step by step by step.
6:49
And that's hence why the prompts are so important because you can
6:52
say, okay, I want you to act like a e-commerce copywriter, and I
6:55
want write, I want you to write a copy for a female audience, not a
6:57
male audience.
6:58
And I want you to be, you know, and be very.
7:01
Specific about the audience and, and, and that, you know, subconsciously a
7:05
human copywriter that's been around the, around for a while will, will,
7:09
you know, make those decisions again.
7:11
Oh, right.
7:11
I'm writing for a female audience and it's about baking and I'm gonna be writing
7:14
about, and these kind of phrases are gonna resonate with my target audience.
7:19
Yeah.
7:19
Ai.
7:21
Doesn't know that to start off with.
7:22
So you have to sort of give it those prompts to sort of
7:24
steer it down the right track.
7:26
Um, and also you can refer a little tip for when using Claude, you can write a
7:31
piece of copy yourself and go, Claude, when now you've looked at like the
7:37
bare bones of this article, can you, this, this piece of text as as a tone
7:42
of voice and, and write in my tone of voice and you'll get a better output.
7:47
So giving, giving the.
7:50
The AI those steers are very important because I think, you know, we perhaps
7:56
most, most of it's coming into it, you think, oh, it's amazing.
7:59
There's this black box.
8:00
I can just ask you to do anything.
8:01
And it's amazing.
8:01
It'll just spit out, start spitting out content and doing work for me,
8:05
and we get all giddy and, and start firing off all sorts of prompts.
8:08
Um, but with, perhaps give it, give AI as it is at the moment, are too much, um.
8:16
Too much kudos about the, how intelligent it is at, at, at the
8:19
generative level.
8:21
Um.
8:22
but there's the next wave of ai.
8:25
Um, and, you know, um, I'd say like the better, the better ones out there.
8:31
The claw, the deep seeks the, um, rock is pretty impressive as well.
8:36
Um, are start starting to work towards a generative ai.
8:39
So generative AI is basically starts to use initiative.
8:44
so.
8:44
can imagine like your, your grad analogy, what, when you've had a grad
8:49
work for you for sort of 18 months, two years, they start to be able to
8:53
Oh,
8:53
right.
8:53
Well, they know the ropes, they know what's good and what's bad, and
8:56
they can make, they can use their own initiative.
8:59
And I think that's what, that's the next sort of leap within, within
9:03
a, um, AI that once that becomes main mainstream, it can get, it can
9:08
be really, really super powerful.
9:10
I
9:10
think.
9:11
Uh, um, and
9:13
I think that's where, where we're sort of moving to.
9:16
It's interesting, isn't it, because.
9:18
Um, it's the logical place, I think for AI to move to now.
9:22
Is, is that side of things.
9:23
I mean, the latest release of Claude, I can't remember what Claude
9:26
we're on sonnet something or other,
9:29
Yeah.
9:29
out a week or two ago.
9:30
Uh, here it is, Claude 3.7 sonnet.
9:33
I think it's,
9:34
Yeah.
9:34
it's just stepped up the game quite a bit, you know, I,
9:37
Yes.
9:38
where it was even previously.
9:40
So you can see that there are these rapid improvements.
9:43
Um.
9:44
And I get what you're saying about the prompting and, and telling it to the
9:50
sort of the style, the tone of voice and how important that is and getting that
9:54
right and not assuming that it will know.
9:57
I remember, uh, recently a friend of mine, he, he, he's released, he's
10:01
launched, uh, or he's acquired actually a, um, an airsoft business, you know.
10:06
Um, and I have a slight interest in Airsoft having two sons, uh,
10:10
who like to shoot each other, you know, with various plastic balls.
10:14
Yes.
10:15
so he asked me what I thought about the site, and I thought, well, again,
10:18
the product copies, you've just copied that from the manufacturer
10:21
and put it on your website.
10:23
And so within sort of a, I guess 10 minutes of playing around with some
10:27
prompts on, chat GPT or Claude, I can't remember which one we used,
10:31
'cause that was the one that he had.
10:34
We found that actually if we put into chat, GPT, um, can you please rewrite this
10:40
copy in the style of a James Bond novel?
10:46
And actually.
10:48
And, and by the way, you are aiming at people who've, who want
10:51
to be James Bond in their sort of twenties and thirties, right?
10:55
Who feel like they should belong to some secret service type affair.
11:00
And actually, I. did a really good first job.
11:04
So you, if you just take the copy and tell it to rewrite
11:08
it, it's still gonna be boring.
11:09
But given it this direction, given it that prompt, it was some of the best product
11:13
copy I've read, I've read for a long time actually, uh, when we introduced this
11:18
Yep.
11:18
James Bond and, and I think giving Claude and Chat GPT, that information just.
11:23
Really is helpful.
11:24
So I think what you said is super important.
11:26
Get the prompts right, understand your tone of voice, and I think
11:30
understand your tone of voice as it relates to your client.
11:32
I would've thought, you know, who, who your customers are
11:35
and what makes sense for them.
11:37
and I think that is, gonna have a big impact just on the copy on your website.
11:43
So now we're moving into generative ai.
11:45
And by the way, just to say ladies and gentlemen, uh, if you are
11:49
enjoying conversations like this.
11:51
Uh, then why not come and join our monthly Cohort calls?
11:56
We've got one in Australia, we've got one in the uk.
11:58
We're just about to start one in the us They're basically
12:00
free set, uh, free calls.
12:02
Basically we, we, jump on, uh, on Zoom, uh, as founders, uh,
12:08
e-commerce founders, we share ideas, share thoughts, um, and just.
12:13
Figure out how we can do eCommerce better.
12:15
So if you are in the eCommerce space, uh, and would like to join our Cohort, why
12:19
not come network, come meet with people.
12:21
Uh, more information at eCommerce Podcast dot net at breakover.
12:25
Uh, Matt, so
12:26
let's talk about generative ai.
12:30
This is, like you say this, I, I like again, your, your graduate.
12:34
This is where they've learned a little bit and now they're starting to be a lot
12:37
more creative and a lot more playful.
12:40
This is where I. People start to get a little bit more scared though, because if
12:46
I think of Cyberdyne systems, you know, from Terminator, it started to learn,
12:51
Yes.
12:51
and then Sarah Connor and, and all that sort of stuff had to happen.
12:57
so how have you, have you put all of this together?
13:00
Are you excited by it?
13:01
Do you think we're all going to, you know, get taken over by robots?
13:04
Um, well, I think that taking a step back and that obviously disclaimer, I'm, I'm
13:09
working in ai so obviously clearly I'm gonna be pro it, but I mean, I think, um.
13:15
Listen, you know, when, back in the day when the Ford Motorcar was being rolled
13:19
out and everyone was riding around, riding around in horse and cart, they
13:22
said everyone's gonna lose their jobs and it was gonna be bad for, for business.
13:25
You know, the, the motorcar was,
13:27
uh, people were quite skeptical of
13:29
it when it first came out.
13:30
And I think, you know, AI is, is another layer of technology, Um, will it
13:35
ever be to the level of like the Terminator?
13:38
No, uh, it won't, it won't do, um.
13:41
but Uh, on the, on the flip side, I think it could actually have a lot
13:45
of, uh, good, uh, power and powerful use within the world of research is
13:50
one, one area.
13:51
For example, um, you know, I've got AstraZeneca down the road
13:54
and, you know, the use of ai when
13:57
Reviewing clinical tests, huge, massive data models, which would take, you know,
14:02
hundreds of grads years to, to decipher.
14:06
You know, AI could actually do a, a,
14:08
ton of heavy lifting there.
14:10
Um, if you think of bringing, bring it into the world of e-commerce, things
14:13
like keyword research and understanding buying intent and how and why people
14:19
are searching for a products online.
14:21
Um,
14:22
now that's
14:22
a certain passion of mine.
14:25
Uh, Coming from my sort of performance marketing background as
14:28
well, but having an AI that can actually, um, do an awful lot of that heavy lifting
14:32
and give really in really good insights into demographics, uh, um, uh, where
14:38
people searching from worldwide, you know, really deep diving into that information
14:44
and giving you really clear actionable insights at a rapid pace.
14:48
I mean, that's, that's.
14:50
Fantastic.
14:50
You know, so just think the amount, amount of years I've spent probably
14:53
sat staring at Excel spreadsheets, and you're probably the same.
14:56
It's, um, you know, it's probably, probably a good thing.
14:59
Uh, but, um, so I think that that's it, you know, it has actually got
15:04
a lot, a lot of good that you can actually bring to the world and
15:07
actually reduce a lot of mundane work.
15:11
Um, um, and there'd be
15:12
new, new kinds of jobs and new kinds of kinds of work.
15:15
That's opened up because of it basically.
15:18
Um, So, you know, I think there in, in the world of, in the world
15:21
of research, advanced engineering medicine, uh, you know, all those
15:27
sort of areas that, um, The, the, they'll actually open up the jobs or
15:31
make things more accessible basically.
15:34
Um,
15:35
and
15:35
those kind of
15:36
companies like, you know, the world Yeah.
15:39
And and eCommerce as well.
15:40
Yeah.
15:40
Um, so you think that, you know, there's the, there's a
15:44
whole sort of eCommerce brand.
15:46
uh, A sort of ecosystem, which was sort of exploded on the back of the
15:51
Amazon aggregators a few years back.
15:53
Uh, God rest their souls at
15:57
uh, made a lot of very wealthy people.
15:59
But essentially, rather than having a, having to be a massively
16:04
well-funded aggregator like
16:06
Thia was, uh, you know, having individuals or small groups of people
16:11
launching, uh, brands and being able to.
16:14
sell a lot of product very quickly and, you know, get to market faster than
16:18
the Proctor and Gambles of this world.
16:20
Um, you know, there's, there's, there's a real opportunity for, for, for people
16:24
to, to, to, to, you know, make, make money on the back of, um, AI rather than just
16:29
thinking around, I'm gonna lose my jobs.
16:31
You know, it's, it
16:32
is, an opportunity for, for people that are.
16:37
Willing to learn and willing to upskill and, um, you know, use the latest
16:41
tools versus some of the, you know, the bigger organizations in, in, in
16:45
AI are adverse to using it because they're worried about, you know, data security
16:50
and, and ethics and those types of things.
16:52
So,
16:53
Well, I, I like what you said there
16:56
yeah.
16:56
um, in that, um, like with everything, you know, we've got the, at the moment I've
17:02
been writing on LinkedIn about tariffs.
17:03
There's a lot of.
17:05
I dunno when this episode's coming out, but as at the time of recording,
17:08
there's a lot of press about the tariffs, the, you know, the Trump tariffs and
17:11
how they're sort of shaking things up.
17:13
And what, what happens is when things like this happen is people panic and they
17:17
make decisions, reactionary decisions rather than strategic decisions.
17:21
And I think when there are leaps, like in ai.
17:26
Actually sitting back and making a strategic decision is really important
17:29
and un so understanding the tools, understanding what it can do for you.
17:32
And I love this idea that you, you know, you said you can now using these
17:36
tools in effect, you've got a research scientist working for you full time.
17:41
Uh, they don't complain.
17:42
They just, they go away.
17:43
They, it's, it's, it's a brilliant thing.
17:47
And so taking that data, if you know how to use the tools.
17:50
Because you are not Proctor and Gamble, because you are not the Titanic.
17:54
You're a speedboat.
17:55
You can turn, you can be super nimble and bring something to
17:58
market way quicker than they can.
18:01
I. take advantage of the opportunities that the bigger companies can no longer
18:05
do because they're, they're, well, they've always had this issue, haven't they?
18:08
They've always been constrained by their size and their
18:11
processes and how they do things.
18:13
and I think in some respects what you're talking about then is generative AI
18:17
becomes a great leveler, doesn't it?
18:19
For,
18:19
Yeah,
18:20
entrepreneurs.
18:21
a absolutely a hundred, A hundred percent.
18:23
And I think if you, you layer on that concept with, um, AI
18:27
agents that, uh, are working
18:29
within the consumer journey already.
18:31
Ru Rufuss being one of them with
18:33
within, um, Amazon World, uh, you know, um, you know, I, I. Hey, I, I'm, I'm,
18:39
I'm a guy and I'm, I need a razor, but I'm, I'm looking for a waterproof
18:42
razor that's, you know, I can use, uh, on, on the go when I'm, I'm traveling
18:47
and, and,
18:47
and, I can ask the AI that and go, go off and find me.
18:50
Here's my budget.
18:51
Go off and find
18:53
the, the best ones in the market.
18:54
Um, and, and I want it to be sustainable and I'm not gonna buy, you know, from
18:59
the US 'cause the tariffs, are gonna buy Germany or whatever it may be.
19:02
You know, the consumer has lots of fickle, um, ways means to buy a product and.
19:08
You can, get, you can ask an AI agent like rufuss,
19:12
those questions that'll go off and turn the product.
19:15
So it's a great level where it's um, you know, the proctors and gambles and
19:20
the Unilevers and the, and the rackets of this world are, are there dominating
19:24
the physical shelf in the store?
19:26
You know,
19:26
they're, they're, they're, they're using their, their sourcing and.
19:30
Heft in, in a marketing sense to, to, to be there and keep
19:34
the little guys and girls out.
19:36
Um, and it's the same in the world of eCommerce, isn't it?
19:38
You know, they're out spending people on Google.
19:41
They, they, they're, they're spending a ton on organic SEOs.
19:44
they're outranking you.
19:45
And sometimes as a, you know, the one man band, you, you, you're
19:49
thinking, well, how, you know, how how am I gonna compete in this niche?
19:53
Whereas the AI will be a great lever of that.
19:55
Okay.
19:55
And actually, well, no, I'm gonna go off.
19:57
I'm gonna, I'm gonna be agnostic to, to brand.
20:00
I'm not gonna follow the big logos or anything like that, but
20:03
I'm gonna go off and find the best product for you consumer.
20:06
So the consumers have got the research agents as well.
20:10
So
20:10
I think you've got like the entrepreneurs and the consumers are, well, hopefully,
20:14
you know, and we all spoken about to our friends the, you know, the latest product
20:19
that you've just found in God, you know, only if I'd found this thing earlier,
20:23
guys, but it's, it's absolutely brilliant.
20:25
You know, I've just
20:25
found Whatever it may be.
20:28
Um, you know, the Dysons of this world before, before, uh, uh, you know,
20:34
vacuum, vacuum cleaner bags were a good thing that, you know, di Dyson was,
20:38
was nowhere to be seen because the, the big boys in in that market were,
20:44
were
20:44
pushed pushed them out.
20:46
So, but now AI would be a great leveler with, with entrepreneurs like Dyson
20:51
because they would've been able to get
20:52
way To market way faster through
20:55
prototype
20:56
testing and also be found by the consumer
20:59
faster as well.
21:01
Yeah, it's fascinating, isn't it?
21:02
I, I like this idea.
21:04
I am intrigued actually how AI is now starting to impact
21:08
how we search for products.
21:11
Hmm.
21:12
to be if you wanted something from any website, Amazon, or
21:14
whatever it was, you kind of had to know what it was you wanted.
21:18
Yes.
21:19
Um, and if you didn't know what you wanted, like when we had the beauty
21:22
site, we had blog posts after blog posts, after blog posts to try and
21:25
help you understand if you have this skin condition, you might wanna try
21:29
these products to bring that education.
21:31
And we did it through blogs, did it through YouTube videos
21:34
Yep.
21:35
all that sort of stuff to help consumers figure it out.
21:38
Now, um.
21:41
I go to perplexity and I say to perplexity, um, I was after some, uh,
21:46
what product was I after the other day?
21:48
Um, that I bought, uh, the electrolyte drinks, right?
21:52
So
21:52
Yes.
21:52
the gym, I wanted an electrolyte.
21:53
I. Supplement that I could take.
21:56
What's the best per supplement that I should take for electrolytes?
21:59
This is my exercise.
22:00
This is what I do.
22:00
Go away, find out, come back to me and tell me.
22:02
And it, and it goes away.
22:03
And it comes back.
22:03
Now, whether it's the right one for me is another story entirely.
22:06
But the fact is that we are now outsourcing and with Amazon, you know,
22:11
with, with Rufuss, like say a friend of mine, he developed Ross Heatherington,
22:15
um, who owns Shopify Marketing.
22:17
He developed a search engine where you would put in your problem and
22:22
using AI it would find products that would solve your problem.
22:26
Right.
22:27
he, I've seen the prototype and he is like, there's no
22:30
point in taking it any further.
22:31
'cause Amazon seemed to have now resolved this problem with Rufuss and
22:33
it, it seems to be that that is a really interesting avenue now to go down and to
22:41
understand as e-commerce entrepreneurs that actually the way we're gonna
22:45
search, I think is probably different.
22:47
Yeah,
22:48
absolutely a hundred percent.
22:49
It, it's going, I mean, um, uh, before I used to work at an agency
22:55
before and, and opt optimize on an, a Marketplace in, in Amazon world, and
23:00
we, I always used to describe it to the, the teams there, as you know, to.
23:06
To understand how we should be optimizing a product, we need to
23:09
think about the buying missions.
23:10
So what, are the different ways that you're gonna buy a snack bar
23:14
and like eat Naturals, a brand that we were looking at at the time
23:17
and you think, well, let's see.
23:19
Well, it could be a snack bar and I'm just hungry and you know, kid, I've
23:22
got teenagers in the house and they
23:23
just need snack bars to fill 'em up One, one by mission.
23:27
Another one is, I'm a gym monkey and I'm I, and I want protein
23:31
bars and I want something with protein so I can, I can stack up on
23:34
protein after I've been in the gym.
23:35
Another way is actually, it's also a vegan bar, I'm a vegan
23:39
and I don't want anything.
23:40
So there's,
23:41
you know, consumers
23:41
come at a product in different
23:42
angles to like their, their needs and wants is really important.
23:47
And I think that, um, you know, if you can, if you as an entrepreneur
23:51
and as an e-commerce entrepreneur, this is where I was talking about
23:56
like brand and e-commerce and AI really meet in a perfect match.
23:59
It's like I I, and I'm launching my own brand later this year, by the way.
24:03
Uh, so what's.
24:04
So, um, but you know, I'm, I'm, I'll take you through the thought process.
24:07
I'm, I'm, I'm looking at, okay, well, I've spotted a niche that I think I can
24:11
get in in, it's not too much of a large niche, but, you know, I think I can, um,
24:16
uh, if I, if I, if I hits it right, I could, you know, generate enough income
24:20
to make, to get me to the next level and then expand that, which is great.
24:24
That's a good tick, right?
24:25
Let's, then, Then I can deploy my a AI agents and go, okay,
24:29
well let's look at this niche.
24:30
What, what do people like about products?
24:32
What don't they like about products?
24:34
What, what pain points have they got?
24:36
What Issues?
24:36
Have they got, you know, um, and then figure out why, why people would
24:42
want a product or need a product.
24:44
And it's that need, need point It's really like, what, what, what's it?
24:49
Is this not being scratched by the competition or the big boys?
24:53
And I think it's really, that's at a point where, you know, as an
24:56
entrepreneur you can go, okay, well
24:58
If
24:59
that product doesn't exist.
25:00
Well, let's go off and create that product, and then we know
25:03
how to market it in that niche
25:04
uh, and then go after it from that point of view.
25:07
Um, or I can create a different category altogether because there's
25:11
needs and wants from a consumers here that are not being met by them.
25:16
So I think that's really super interesting that you can
25:18
actually develop stroke source
25:21
products Based on, on the needs of what's in a category that perhaps are even,
25:25
even being fulfilled by the competitors.
25:27
Um.
25:28
and, and that, and then, and then adapt your, your marketing
25:31
to really target them as well.
25:34
Um, I think that's about, it's a really interesting space, um, particularly
25:39
if you can get AI has to go off it.
25:40
Haven't quite nailed it quite yet, but I, I spotted an ai, um, the Wonder
25:44
Family, I dunno if you've seen that.
25:46
It's called Steve.
25:48
Um,
25:48
And, um, there, there's, they've just literally just launched it.
25:52
So there's an AI that you can, you can give it a niche that you want
25:55
to target, and then you can sort of decide on a product and it'll go off
25:59
and then, uh, source that product and send out the emails to suppliers and
26:05
set up a competitive contender and then get it, get it all for you as well.
26:09
So,
26:09
in theory, I haven't, you know, that's something that's being
26:12
launched at the moment, but in theory, you know, you could actually.
26:17
Get a very slick system in place, basically.
26:19
So I think that's sort of, um, uh, and so there's some really,
26:22
really cool tools out there.
26:26
It's really interesting, isn't it?
26:27
How people are using AI now to help them.
26:30
It's like, what's the most boring part of your job?
26:31
Finding the products, getting the tenders and all that sort of stuff.
26:33
So it's just here, just go do that.
26:37
I, I, I want to come back to
26:38
your
26:38
Yes.
26:38
if that's okay, but I, I guess this is a good point to ask.
26:43
What tools then are you using to launch your brand, and what tools
26:49
should e-commerce entrepreneurs be?
26:51
I mean, we've touched on GPT as an obvious one.
26:53
Claude I think is much better for content than GPT, but that's
26:57
just, you know, my, my personal
26:59
Agreed.
27:00
um.
27:00
Yes.
27:01
Well, we agree this, this is a beautiful thing.
27:02
So what else should we be thinking about?
27:05
What else should, what tools have you found that you think,
27:07
actually, this is a really clever thing that they're doing right
27:09
Well, you, me, you, you mentioned one, um,
27:12
or, or carry, carry
27:12
on the, the, the agreement is there's perplexities, it's a great
27:16
tool for product research and,
27:17
and, um, you know, if you wanted to dig into.
27:21
the Demographics and, uh, just provide a quick overview of a,
27:25
of a category, so that's great.
27:27
Um, um, and, um, also, um, shameless plug of Dragonfly, which
27:34
where I'm working at the moment.
27:35
Um, uh, and that'll allow, uh, the user to scan any content, physical or online,
27:42
uh, to 89% accuracy of the human eye.
27:45
Um, and decide about which, which, which product or which image.
27:50
Will, uh, gain more attention and you're more likely
27:53
to click on it before you launch it, basically.
27:57
Um, and that works on the physical shelf as well, because obviously, you know,
28:01
picking off the product off the shelf, uh, that, that standout on shelf is
28:04
really important as well for packaging.
28:07
Um, so that's, that's a really good tool.
28:11
Um.
28:12
there's, um, tools like Pfu, uh, out there, which has been a while around
28:16
for a while, but they've got some good, interesting AI capabilities in that.
28:20
So you can, when launching your product, you
28:23
can have different logo types or different.
28:27
Concepts and then launch 'em out to small groups of people, get some feedback.
28:32
But the AI will summarize their overall feedback as well.
28:35
Um, so that's, that's really, really cool.
28:37
I, there's, there's some, some good, good starting points
28:41
there really, uh, to look at.
28:43
Um, and there's some good tools like make.com, which help.
28:47
It's a bit like,
28:49
Yeah.
28:49
connecting
28:50
various things.
28:51
isn't it really?
28:52
Yeah, basically.
28:53
I think that's, that's a good.
28:54
A good tool really.
28:55
But I would, I would say, is, um, you know, we're going back to sort of
29:00
opening up Pandora's box and people being absolutely swamped with lots of different,
29:04
um, products out there that you can use.
29:07
And I haven't even touched upon visual and, and video ai, which is, some of
29:12
it is just absolutely mind blowing.
29:15
I dunno if you've seen any of it, like cling and, and, and render and the like.
29:20
Um, it's really just.
29:21
Start with start small,
29:23
start what they say, what is that really boring?
29:26
Because if, if you're not passionate about doing something, you're never
29:29
gonna, A, be really good at it, or b, make a lot of money at it, in my
29:31
view.
29:32
So, so let's find that really boring thing that's taking up a lot of time in
29:36
your business and just focus on that one thing with, and just get a system that's
29:40
nicely polished and, and you, you've got a, you know, human being checking
29:45
it before it goes out sort of thing.
29:47
Um.
29:48
You know, and, and, and then, and then move on from there would
29:51
be my advice because it, it, it
29:53
can be overwhelming just for the amount of time it takes from
29:56
learning these new tools, obviously.
29:58
But also, uh, you know, the costs start ramping up as well.
30:02
Um, you know,
30:03
before, you know, you're spending like sort of five or six grand a month on
30:08
It's easy, isn't it?
30:09
It's easy to do.
30:10
yeah.
30:10
And I think there's an opportunity cost isn't there?
30:12
Because for me, I've, I've spent a lot of time understanding at the start,
30:18
I spent a lot of time understanding GPT and Claude when they, you
30:21
Yeah.
30:22
out understanding how AI works.
30:24
And I think I can use those, those both relatively quite well to get out
30:30
of it what I want to get out of it.
30:33
Um, but I'm also aware that how I use Claude now is very different to how I use
30:39
c. Used it a year ago, you know, there's this sort of perpetual learning isn't
30:42
there and, and staying on top of things.
30:44
'cause it's not like, it's not like stationary, it's not like Microsoft Word,
30:47
where once you've figured it out, you're fine for the next 20 years, you know,
30:50
Yes.
30:50
uh, it's, you may be fine for the next 20 minutes.
30:53
So there's this perpetual learning.
30:55
So there's the actual cost of the software, but there's also the opportunity
30:58
cost of having to stay on top of it.
31:01
So I, I do like what you said.
31:04
You know, you've gotta pick your tools wisely, haven't you?
31:07
Otherwise you're just gonna get swamped and overwhelmed with the whole thing.
31:11
Yes.
31:12
what's having the biggest impact on your business?
31:15
I think,
31:17
that.
31:19
That makes a lot of sense.
31:20
You know, pick sort of two or three and then just get really
31:22
good at them and then add, it's a bit like social media, isn't it?
31:25
I suppose you can, you could be on every social media platform, but
31:28
it's probably not good for you.
31:30
Um,
31:31
Yeah.
31:31
do one or two really Well,
31:33
Yeah, and I think there's, um, there's about, so.
31:36
18 months ago, I think, you know, I, I, I've sort of hit a moment where
31:41
I, I was back in the day, sort of 2007 on Twitter and, and, and I was,
31:47
uh, doing digital marketing on Facebook at sort of 20 10, 20 12 and,
31:52
and, and, and, just, you get absolutely, um, new, it's a new
31:56
shiny thing, but you having to absolutely immerse yourself in it.
31:59
And I was just, um, I I, I got a bit burnt out with social media to be
32:03
honest, because it was just, it was just,
32:05
you know.
32:06
Too, too immersed in, in all the different platforms too,
32:10
too overwhelmed with it all.
32:11
Um,
32:12
and I think that's just the pace of technology now is, this, this so, so far.
32:16
So I think you are right.
32:17
It's just sort of taking a step back, Focus, focus on where you can add the
32:22
best value and, and, and where's the best opportunity cost in your business.
32:25
And, and, um, and, and, and master that you know, you, you can't
32:29
be good at every martial art.
32:30
Just pick one.
32:32
Be a black belt, and it basically has been my advice.
32:34
Yeah, very true.
32:36
Very true.
32:37
Although I st and I'm the wrong person to say that 'cause I still
32:39
did quite a few different ones.
32:41
Um, but.
32:43
I, I
32:43
do get the point.
32:45
Yeah.
32:48
Yeah.
32:48
Yeah.
32:48
And it's usually me to be fair, mate.
32:50
It's usually me.
32:51
Um,
32:53
love that.
32:55
before we get into your, I want to talk to you about your, your brand, if I may.
32:58
But, um, before we do that, let me do this thing while I remember where I ask
33:03
my guest, which is you, Mr. Anderson,
33:05
Yes.
33:05
uh, for a question.
33:07
This is where I take the question.
33:08
At some point in the future, I'll answer this question on social media.
33:12
So, Matt, what's your question for me?
33:15
Um, request for you.
33:17
Um, I think in the, in, in, in, in the world of web and, and.
33:23
you know, What, what does, what does a AI mean for sort of the world of web design?
33:27
Um, uh, really I think it's, um, it's one area that I, I've been looking
33:31
at, um, Shopify for example, is, um, as a, as a tool that I'm, I'm
33:37
developing and I know I had a good rudimentary understanding of it.
33:41
A lot of my deep expertise within
33:43
is
33:44
within Amazon, and I think.
33:47
where, where do you think that the world of AI is gonna
33:50
take the world of web design?
33:52
Because I think, um.
33:53
I think there's a, a really, there's a really good opportunity there.
33:56
There's already tools out there for designing websites,
33:59
using ai.
34:00
but you know, going back to your web copy and your, you know, and you have
34:06
AI images, there's always got that dystopian feel about them, haven't they?
34:10
They're always
34:11
like, they've got this slightly creepy feel as
34:13
these people.
34:14
ones are ai.
34:16
So, you But where, but you know where, where, where, because you've got SEO
34:22
O key Web research, um, copywriting,
34:24
coding could all be done by ai, ai Eventually I saw a podcast, I think
34:29
it was the Diary and the CEO and I can't remember the guy's name on it.
34:32
He was talking about agen I be able to launch a website, uh, and go,
34:38
actually, what if you're launching this website in a beauty brand
34:41
within
34:41
this market?
34:42
That's great.
34:43
but have you thought about launching it in Hindu?
34:45
And would you like me to translate it into that language as well,
34:48
you know, and, uh, you know,
34:49
will, will it go in that area or would it, is it just, just too,
34:52
quick, too complex really?
34:53
So yeah.
34:54
What's it mean for the world of ai, I think is, um, definitely something that,
34:59
um, I'm trying to get my head around and I'm not really, it, it's, yeah,
35:03
like I say, the Pandora's box is open
35:06
Yeah.
35:06
Yeah.
35:07
in theory.
35:08
A lot of these things could be or should be very or very soon.
35:12
possible.
35:13
but where?
35:14
Where's it going?
35:14
Really?
35:15
Very good question.
35:17
Yeah.
35:18
Yeah.
35:18
have a little ponder about that, if you wanna know how
35:20
I'm gonna answer that question.
35:21
Can follow me on LinkedIn, Matt Edmundson?
35:23
Just find me there and I
35:24
Yeah.
35:24
answering that question.
35:26
Uh, Matt, listen, tell me, uh, obviously if you are launching a brand, there's a
35:30
limit to what you do want to say, but um.
35:33
What are you launching?
35:35
Uh, as much as you can say, uh, towards the end of the year.
35:39
So
35:39
I'm watching, um, a brand in the pet space.
35:42
Um, I've been work, working.
35:44
in the world of sort of digital, I worked with a number of pet brands previously,
35:49
um, like within Mars, Mars Pet Care, Spillers and Trill and you Move and, and
35:55
JSON
35:55
and Paige and brands like that.
35:57
So I've, I've, I am passionate about the pet industry.
35:59
It's, it's growing significantly.
36:01
Um, and, um, Um, I'm very interested in, um, in, in health around, uh,
36:08
animals like, within Cat and Dog.
36:11
Um, and around anxiety and the like, and how, how, what, what fascinates me is how
36:19
humans have, um, adopted their own habits.
36:23
Whether it be like turmeric for, you know, for, for, for recovery, for example.
36:28
And then actually taking that theory and going actually, well
36:31
my dog could use some term Rick.
36:32
'cause it's
36:33
getting old.
36:33
And you know,
36:34
the, so it's this whole sort of pattern of.
36:37
uh, human trends into, into the world of, uh, in pet.
36:40
So that's what I'm interested in really.
36:42
Um,
36:43
and
36:43
it's, it's a, and um, just a whole sort of, um, health, health
36:48
and health and beauty area.
36:50
But pet pet is, is fun.
36:52
Pets.
36:52
Pet's a
36:53
fun cat.
36:53
Yeah.
36:54
it, fascinates me.
36:55
We have this brand,
36:57
Yeah.
36:57
Um, the Vegetology brand.
36:59
Yes.
36:59
if you can, uh, there you go.
37:01
It's on the screen now.
37:02
Vegetology make, um, supplements, right?
37:05
I'm a, I, I am the MD I suppose, of this company and I, I get the,
37:09
the beautiful, chore of making sure that the sales are growing.
37:13
Et cetera, et cetera.
37:14
One of the things that fascinates me about this brand, right?
37:17
So the Omega-3, it is a vegan brand.
37:20
In other words, every product is vegan certified.
37:22
A lot of, um, vegetarians, uh, it's also vegetarian certified, but our
37:26
market is obviously any, anybody, because anybody can take them.
37:29
Even if, if, if you eat meat, you can still eat.
37:32
Uh, take these supplements, right?
37:33
Yes.
37:35
One of the things that's fascinated me, about this is obviously you're
37:39
drawing distinctions, um, between, with vegan brands, you are very
37:43
concerned about the animal kingdom.
37:45
Uh, in terms of how these products were made, what we've noticed is is, and
37:51
what I could never have predicted, uh, with, with all the best hindsight in the
37:56
world, is for example, we have, there are veterinarians in the UK who will order.
38:04
Omega-3 as a liquid format to give to dogs.
38:09
Um, and so they, you know, I'm not, I, I'm not saying one way
38:12
or the other, you can do this.
38:13
I'm just saying this is what somebody does.
38:16
Um, and they're starting to give animals human supplements.
38:20
Um, and so I said to uh, um.
38:24
Uh, the, the ch the, the guy that sort of, Chris, he's the founder of Vegetology.
38:29
There's a brain the size of a planet, real science guy, right?
38:32
And, um, I'm like, well, why don't we do pet supplements?
38:36
I mean, it seems to be like a bit of a no brainer that people want them.
38:40
And he said to me, Matt, the legislation for human supplements
38:44
is about an inch thick.
38:47
Yes,
38:47
for patch supplements is about 10 inches thick.
38:50
Yes.
38:51
It's a really, it's a really interesting phenomenon that we're in.
38:55
So you are getting into that space.
38:57
Good on you, man.
38:58
That, that sounds like it could be, could be quite good fun.
39:01
Yeah, it is good.
39:01
It is good,
39:02
good fun.
39:03
But it it, it is interesting about this sort of human.
39:06
pet sort of crossover.
39:07
I was talking to a a, a vinegar brand, Apple side, a vinegar brand.
39:12
And, uh, they were noticing that we were going with some
39:15
large bulk colds from customers.
39:16
And basically it was, it was, a horse owners buying it
39:20
and putting it in the feed, uh, to help with gut health, but they were just like
39:24
taking the human products and bang it
39:27
in the hu in the horse speed.
39:28
So I think, yeah, it's definitely, there's a, the crossover each way.
39:31
And you do get, like, in the world of, I, I sat in a. In a, in a in a client
39:36
meeting in the world of dog and they just released this new, new chew Supplement.
39:41
And, uh, of course
39:42
they,
39:42
they, they, all hand a pack around and go, just kind, let's
39:45
try one sort of thing so that
39:48
they all, they all eat, all eat the dog food and, and, uh, and so yeah, it's
39:52
uh, it's a, bit, yeah, it gets, can get a little bit strange at times, I
39:55
guess when you're sort of sitting there eating doggy snacks, but you know, I,
40:00
it's
40:00
not
40:00
I don't mind.
40:01
I, I, genuinely don't mind.
40:04
legislation you mentioned, it's probably actually, uh
40:07
uh.
40:08
is.
40:08
It reminds me of the film, uh, you know, lethal Weapon with
40:12
Mel Gibson and his eating dog snacks to curb his smoking habit.
40:16
Yes,
40:16
yeah, going back a long time now, ladies,
40:19
yes, yes.
40:20
Lethal Weapon movies.
40:22
Now, before we wrap up, uh, lemme just say if you have found this whole
40:26
conversation about ai interesting, I would love to invite you to
40:30
join us in our weekly newsletter.
40:33
Uh, in the newsletter we go through all the stuff that Matt talks about.
40:36
We're go and find some extra research, some extra stuff, gonna be expanding on
40:40
this whole topic around AI and eCommerce.
40:42
Um, because you know what?
40:44
It's brilliant.
40:44
I found that the podcast format is brilliant for these big ideas.
40:47
Love these conversations.
40:49
And the newsletter is where we get to break them down a little bit.
40:51
So, uh, do come sign up for that.
40:53
The link for that is in the show notes.
40:56
But Matt, listen, I've really enjoyed the conversation, man.
41:00
It's always good to think about ai.
41:02
I'm especially enjoying the skateboard on the wall behind you.
41:05
Uh,
41:06
the bright yellow skateboard.
41:10
That's, uh, that's, that's, that's quite, you've got some
41:12
interesting art on your walls.
41:13
It's very, very,
41:15
quite eclectic.
41:17
Yeah, it's eclectic mix in here.
41:18
Some, um, it's my, my, my, um, office and um, I always joke, it's sort
41:23
of like the stuff my wife doesn't want on the walls in the house.
41:26
This
41:27
is
41:27
I get, I get
41:28
what I want.
41:29
That's so true for every man cave, right?
41:31
Everything in this, everything in this office.
41:34
I'm in my man cave at the house at the moment, so it's in effect,
41:38
Yeah.
41:38
you know, it's my, my home office, but in, in here is all the
41:41
stuff my wife doesn't want out.
41:42
So, uh, like my little Lego, Indiana Jones figures, you know,
41:46
Yes.
41:46
Yeah, I'm glad.
41:47
I'm glad I'm not alone.
41:48
I've, I've got lots of things, you know, all, all your little
41:52
mantros you put on your desk and
41:54
things,
41:54
Yeah, man.
41:55
Trophies.
41:55
I love that phrase.
41:58
That's what it is.
41:58
I've got these man trophies.
42:01
um.
42:02
So, yes, I've got a, this, little chrome grenade anyway, but, um,
42:11
yeah.
42:12
Fun, fun, fun, fun.
42:13
Uh, listen, Matt, uh, tell us how people can reach you.
42:16
Tell us how people can connect with you if they wanna do that.
42:18
What's the best way?
42:21
Uh, yeah.
42:21
So, um, I'm on LinkedIn, so Matt, Matt Anderson.
42:24
Um, and, uh, if you'd like to check out more about Dragonfly dragonflyai.co is
42:31
the website and so do, do check us out and, um, yeah, do feel free to, um.
42:37
Drop me a line on LinkedIn and a dm. Um, if you find anything that's
42:40
interesting we've been talking about, then do, um, and I'll be, um, you
42:44
know, keeping an eye out for the posts about, um, AI and website design.
42:49
'cause I think that's, that's something that, you know, I
42:51
haven't, I, I'm not the expert.
42:53
I haven't cracked it yet.
42:54
but I'd say Pandora's Box.
42:55
is open, so, uh, I'll be there, be there commenting as
42:58
well.
42:59
boxes.
43:00
Yeah.
43:01
open.
43:01
No.
43:01
Fantastic.
43:02
Well, listen, Matt, I've thoroughly enjoyed the conversation.
43:05
Thank you so much for joining me.
43:06
Uh, let me do the usual.
43:08
Yes, there we go.
43:11
that's been awesome.
43:14
Well, that's it from, yeah.
43:19
Brilliant, brilliant, brilliant.
43:21
So that's it from me.
43:22
That's it from Matt.
43:23
Thank you so much for joining us.
43:24
Have a fantastic week wherever you are in the world.
43:27
I will see you next time.
43:28
Bye for now.