TRANSCRIPT 🔗
1
00:00:00,000 --> 00:00:01,375
We see it in digital
2
00:00:01,375 --> 00:00:02,958
transformation all the time.
3
00:00:04,125 --> 00:00:04,750
People make the mistake of
4
00:00:04,750 --> 00:00:05,625
starting with the software.
5
00:00:06,125 --> 00:00:08,458
You need to start with the person behind
6
00:00:08,458 --> 00:00:08,833
You need to start with the person behind
7
00:00:08,833 --> 00:00:11,083
the activity or the process.
8
00:00:11,875 --> 00:00:14,750
Map that process and then make your
9
00:00:14,750 --> 00:00:15,541
decisions from there.
10
00:00:15,958 --> 00:00:16,750
I think people need to
11
00:00:16,750 --> 00:00:18,958
remember that AI is a tool.
12
00:00:20,250 --> 00:00:21,333
It's not a silver bullet.
13
00:00:21,583 --> 00:00:22,875
It can certainly be an assistant or a
14
00:00:22,875 --> 00:00:24,083
It can certainly be an assistant or a
15
00:00:24,083 --> 00:00:25,333
co-pilot to watch you
16
00:00:25,333 --> 00:00:27,583
creating, but solely relying
17
00:00:27,708 --> 00:00:30,708
on AI to create, I think, is a mistake.
18
00:00:32,750 --> 00:00:34,416
And I think that's very evident with Grok
19
00:00:34,416 --> 00:00:36,166
as well, where it's being
20
00:00:36,166 --> 00:00:39,875
retrained to see the world in a
21
00:00:39,875 --> 00:00:41,125
certain way or a certain light in
22
00:00:41,125 --> 00:00:44,750
comparison to the way Gemini and ChatGBT
23
00:00:44,750 --> 00:00:45,625
would operate in relation
24
00:00:45,625 --> 00:00:47,916
to answering questions. I do think
25
00:00:47,916 --> 00:00:50,333
there's a huge responsibility there and I
26
00:00:50,333 --> 00:00:51,708
think in relation to
27
00:00:52,250 --> 00:00:55,458
business owners, certainly I think the
28
00:00:55,458 --> 00:00:56,458
awareness has been
29
00:00:56,458 --> 00:00:57,791
heightened in relation to
30
00:00:58,041 --> 00:00:59,291
the capabilities and
31
00:00:59,291 --> 00:01:00,291
how it needs to be used.
32
00:01:01,833 --> 00:01:03,583
Google has an idea of who you are, what
33
00:01:03,583 --> 00:01:05,000
you're looking for, and
34
00:01:05,000 --> 00:01:07,375
it's moving towards, I think,
35
00:01:08,291 --> 00:01:12,750
tailoring the results exactly to what you
36
00:01:12,750 --> 00:01:14,041
need. They've been trialing
37
00:01:14,041 --> 00:01:16,833
in the US in May last year,
38
00:01:17,541 --> 00:01:20,166
for example, the product feeds, shopping
39
00:01:20,166 --> 00:01:22,541
feeds. For example, if
40
00:01:22,541 --> 00:01:23,416
you're searching for shoes,
41
00:01:24,458 --> 00:01:25,875
Search Engine will already know your shoe
42
00:01:25,875 --> 00:01:27,250
size and it will only show
43
00:01:27,250 --> 00:01:28,041
shoes that would fit you.
44
00:01:28,041 --> 00:01:31,208
So that's where it's going.
45
00:01:38,208 --> 00:01:40,500
So today we welcome CEO of digital
46
00:01:40,500 --> 00:01:43,000
agency, Spéire, Stephen Wilson-Downey.
47
00:01:43,583 --> 00:01:44,958
So, Stephen, I suppose
48
00:01:44,958 --> 00:01:46,208
take us back to the beginning.
49
00:01:46,625 --> 00:01:48,583
What was the specific moment or gap in
50
00:01:48,583 --> 00:01:49,875
the market that convinced you
51
00:01:49,875 --> 00:01:51,916
to launch Spéire here in Cork?
52
00:01:53,750 --> 00:01:55,583
That's a great question. So
53
00:01:55,583 --> 00:01:59,041
Spéire was my second venture.
54
00:02:00,708 --> 00:02:03,625
And initially I came from a background of
55
00:02:03,625 --> 00:02:04,833
food and hospitality.
56
00:02:06,000 --> 00:02:07,166
So my first venture
57
00:02:07,166 --> 00:02:08,750
was Ready Made Meal.
58
00:02:10,250 --> 00:02:12,000
And I was in the early
59
00:02:12,000 --> 00:02:13,791
stages of prototyping it.
60
00:02:13,791 --> 00:02:17,958
I was a student at the time in MTU and I
61
00:02:17,958 --> 00:02:20,958
engaged in the price renovation and did a
62
00:02:20,958 --> 00:02:22,125
pitch and got a little bit of funding.
63
00:02:22,416 --> 00:02:23,333
And then I got on to
64
00:02:23,333 --> 00:02:24,666
the Student Inc. program.
65
00:02:25,583 --> 00:02:28,041
So while I was studying, it was a great
66
00:02:28,041 --> 00:02:31,416
bridge from my professional
67
00:02:31,416 --> 00:02:33,375
career as a chef into business.
68
00:02:34,458 --> 00:02:37,500
And I suppose that student incubation
69
00:02:37,500 --> 00:02:42,000
program over a period of three months,
70
00:02:42,250 --> 00:02:44,000
you know, you're doing your macro
71
00:02:44,000 --> 00:02:45,125
research, your branding,
72
00:02:45,791 --> 00:02:46,708
figuring out your segments.
73
00:02:47,750 --> 00:02:50,333
But the thing that caught me was it
74
00:02:50,333 --> 00:02:52,375
wasn't how to start
75
00:02:52,375 --> 00:02:53,375
off and where to sell.
76
00:02:53,625 --> 00:02:56,791
It was what do you do
77
00:02:56,791 --> 00:02:58,541
when you need to scale?
78
00:02:59,250 --> 00:03:00,458
So when there's a big
79
00:03:00,458 --> 00:03:01,708
demand for you to produce.
80
00:03:02,625 --> 00:03:04,333
So I was researching the next step, which
81
00:03:04,333 --> 00:03:05,625
is essentially outsourcing.
82
00:03:07,500 --> 00:03:10,333
And I very quickly realized that I wasn't
83
00:03:10,333 --> 00:03:13,166
able to get it produced in Ireland.
84
00:03:13,166 --> 00:03:14,708
I would have had to move further afield.
85
00:03:16,000 --> 00:03:19,625
And that would have meant I would have to
86
00:03:19,625 --> 00:03:25,500
hire my own space within a facility, set
87
00:03:25,500 --> 00:03:26,416
up a production line.
88
00:03:26,416 --> 00:03:28,458
And I'm sorry, short after all my
89
00:03:28,458 --> 00:03:30,875
calculations, I needed about one point
90
00:03:30,875 --> 00:03:32,041
two million to get it off the ground.
91
00:03:32,708 --> 00:03:38,458
So naturally, it was an idea that I
92
00:03:38,458 --> 00:03:40,666
decided to put to bed.
93
00:03:41,458 --> 00:03:44,541
But I still had that ambition inside me.
94
00:03:45,041 -->
Load full transcript
00:03:49,708
So going back, it was always technology
95
00:03:49,708 --> 00:03:53,250
and particularly websites and food.
96
00:03:53,875 --> 00:03:55,541
But that's what I had an interest in when
97
00:03:55,541 --> 00:03:56,583
I was in transition year.
98
00:03:57,208 --> 00:03:59,541
And I found that I kind of gravitated
99
00:03:59,541 --> 00:04:03,458
back towards websites when I was doing my
100
00:04:03,458 --> 00:04:04,791
Master of Science in marketing practice.
101
00:04:05,958 --> 00:04:08,541
And that's when I met my my co-founder,
102
00:04:08,541 --> 00:04:11,250
my colleague, Dave, and we put five euro
103
00:04:11,250 --> 00:04:12,708
into the Bank of Ireland on
104
00:04:12,708 --> 00:04:15,541
campus and we just got started.
105
00:04:16,250 --> 00:04:19,583
So we both had an ambition.
106
00:04:20,416 --> 00:04:21,583
But to answer your
107
00:04:21,583 --> 00:04:23,375
question on did we see a gap?
108
00:04:24,166 --> 00:04:27,458
Not entirely. We I suppose we were
109
00:04:27,458 --> 00:04:32,375
studying the the edge of where marketing
110
00:04:32,375 --> 00:04:35,291
is going in combination with what
111
00:04:35,291 --> 00:04:37,041
services were in most demand.
112
00:04:37,541 --> 00:04:41,416
And our very first job, which we got paid
113
00:04:41,416 --> 00:04:44,166
by check, was a roll up banner for a
114
00:04:44,166 --> 00:04:45,333
festival in the city centre.
115
00:04:46,291 --> 00:04:49,208
So, yeah, I suppose it further spaned from
116
00:04:49,208 --> 00:04:53,666
there where we found that we delivered a
117
00:04:53,666 --> 00:04:58,250
lot of value to to SME's and we very much
118
00:04:58,250 --> 00:04:59,916
valued relationship
119
00:04:59,916 --> 00:05:01,166
marketing as our strategy.
120
00:05:01,583 --> 00:05:03,500
So we'd invest our time into customers,
121
00:05:04,291 --> 00:05:05,750
suppliers and even other
122
00:05:05,750 --> 00:05:07,458
operators in our industry as well.
123
00:05:07,875 --> 00:05:10,708
Very much adopting and maintaining the
124
00:05:10,708 --> 00:05:13,041
mantra of it's nice to be nice.
125
00:05:13,750 --> 00:05:18,250
And, yeah, like I'd say around 85 percent
126
00:05:18,250 --> 00:05:19,833
of our inquiries are word of mouth.
127
00:05:20,041 --> 00:05:22,875
The rest are digital. OK. And it sounds
128
00:05:22,875 --> 00:05:23,791
like you've kind of
129
00:05:23,791 --> 00:05:24,875
always had that kind of.
130
00:05:25,916 --> 00:05:27,875
entrepreneurial spirit, would I be
131
00:05:27,875 --> 00:05:28,625
right to say that?
132
00:05:29,750 --> 00:05:33,583
Yeah, it's almost hard to stop yourself
133
00:05:33,583 --> 00:05:35,250
from spinning out
134
00:05:35,250 --> 00:05:36,208
businesses and coming up with
135
00:05:36,208 --> 00:05:38,416
new ideas every now and then when you
136
00:05:38,416 --> 00:05:39,000
come across a new
137
00:05:39,000 --> 00:05:40,500
problem. Like, for example, we
138
00:05:41,291 --> 00:05:44,000
had a spin out project in the
area of
139
00:05:44,000 --> 00:05:44,875
artificial intelligence
140
00:05:44,875 --> 00:05:49,125
back in around 2019, where we
141
00:05:49,125 --> 00:05:50,458
spotted an inefficiency in website
142
00:05:50,458 --> 00:05:51,958
maintenance. And there was
143
00:05:51,958 --> 00:05:53,208
a lot of labor involved in
144
00:05:54,208 --> 00:05:56,166
updating, for example, WordPress websites
145
00:05:56,166 --> 00:05:58,416
when there's plugins or
146
00:05:58,416 --> 00:05:59,791
elements that don't talk to
147
00:05:59,791 --> 00:06:01,375
one another and the website can't
148
00:06:01,375 --> 00:06:03,416
function unless they do so you're trying
149
00:06:03,416 --> 00:06:04,250
to find the best mix of
150
00:06:04,250 --> 00:06:06,541
updates without breaking a website. And
151
00:06:06,541 --> 00:06:08,875
sometimes that job can be
152
00:06:08,875 --> 00:06:10,333
straightforward for a website and
153
00:06:10,333 --> 00:06:14,958
others not so much. So that was the
154
00:06:14,958 --> 00:06:16,333
project there where we
155
00:06:16,333 --> 00:06:17,625
decided to build an AI model
156
00:06:18,333 --> 00:06:21,375
to automate that activity. But again, we
157
00:06:21,375 --> 00:06:22,666
came across that
158
00:06:22,666 --> 00:06:24,541
challenge, trying to make the
159
00:06:24,541 --> 00:06:26,083
commercials work, we could make the cloud
160
00:06:26,083 --> 00:06:28,875
costs affordable for the
161
00:06:28,875 --> 00:06:30,291
customer which would pay on a
162
00:06:30,291 --> 00:06:33,208
per website basis. And at the same time,
163
00:06:33,208 --> 00:06:34,333
Spéire was really, really
164
00:06:34,333 --> 00:06:38,291
growing rapidly during the
165
00:06:38,333 --> 00:06:42,291
pandemic as well. So yeah, we decided to
166
00:06:42,291 --> 00:06:44,166
knuckle down a focus on
167
00:06:44,166 --> 00:06:46,333
what's already growing. And ever
168
00:06:46,375 --> 00:06:49,125
since I've been reigning myself in in
169
00:06:49,125 --> 00:06:52,458
terms of any innovations that will be too
170
00:06:52,458 --> 00:06:53,208
much for distraction.
171
00:06:54,791 --> 00:06:57,125
What advantages are there being based in
172
00:06:57,125 --> 00:06:58,750
a central location like Horgans Key?
173
00:07:00,416 --> 00:07:03,791
Yeah, I suppose you said it there.
174
00:07:04,791 --> 00:07:07,458
Central, like we moved
175
00:07:07,458 --> 00:07:09,916
into spaces, which is
176
00:07:10,208 --> 00:07:14,916
a brand operated by Regis. And we found
177
00:07:14,916 --> 00:07:16,291
that the advantage of
178
00:07:16,291 --> 00:07:17,791
being in there is we're able to
179
00:07:18,333 --> 00:07:20,541
use any of their addresses around the
180
00:07:20,541 --> 00:07:21,958
world. They've around 900
181
00:07:21,958 --> 00:07:23,500
addresses. So for example,
182
00:07:23,500 --> 00:07:26,791
if we fly to Manchester or London, when
183
00:07:26,791 --> 00:07:27,750
we're over and back, we've
184
00:07:27,750 --> 00:07:30,291
got a place to
work and a place
185
00:07:30,291 --> 00:07:33,000
to go on a place to meet clients. And
186
00:07:33,000 --> 00:07:35,333
it's great to have that level of access.
187
00:07:36,250 --> 00:07:36,791
And we were originally
188
00:07:36,833 --> 00:07:40,083
based in a really, really nice unit in
189
00:07:40,083 --> 00:07:41,458
ballincollig but we found that
190
00:07:41,458 --> 00:07:43,666
both staff and customers like
191
00:07:44,291 --> 00:07:46,125
didn't really like coming out there, even
192
00:07:46,125 --> 00:07:46,791
though we have plenty of
193
00:07:46,791 --> 00:07:48,833
parking, and people seem to love
194
00:07:48,833 --> 00:07:51,916
coming into the city centre. So yeah, we
195
00:07:51,916 --> 00:07:53,208
made the difficult decision
196
00:07:53,208 --> 00:07:56,625
to move into the city centre,
197
00:07:56,625 --> 00:07:57,958
even though you know, it was only 10
198
00:07:57,958 --> 00:07:59,375
minutes down the road from where myself
199
00:07:59,375 --> 00:08:00,583
and Dave live. But you
200
00:08:00,750 --> 00:08:04,500
have to go where your customers and staff
201
00:08:04,500 --> 00:08:05,916
feel their happiest.
202
00:08:07,041 --> 00:08:09,708
Fair enough. And talk about your business
203
00:08:09,708 --> 00:08:12,041
now. So you offer various different
204
00:08:12,041 --> 00:08:13,458
services to businesses.
205
00:08:14,208 --> 00:08:16,541
How has the agencies, I suppose, focus
206
00:08:16,541 --> 00:08:18,625
shifted since you first
207
00:08:18,625 --> 00:08:20,500
started? And what drove that
208
00:08:20,500 --> 00:08:22,541
evolution? Is it like basically, like you
209
00:08:22,541 --> 00:08:23,208
said there, when you
210
00:08:23,208 --> 00:08:25,208
change into the city? Is it the
211
00:08:25,250 --> 00:08:28,291
customer expectation or what they needed?
212
00:08:28,291 --> 00:08:30,125
Is that what drove that evolution?
213
00:08:31,958 --> 00:08:36,083
In short, yes. I suppose it's really
214
00:08:36,083 --> 00:08:38,375
important for us to keep the finger on
215
00:08:38,375 --> 00:08:39,750
the pulse in relation to
216
00:08:39,750 --> 00:08:42,333
what challenges are experienced by
217
00:08:42,333 --> 00:08:44,375
customers the most and making
218
00:08:44,375 --> 00:08:46,958
sure that our services align
219
00:08:47,791 --> 00:08:49,833
with them and make sure that we're in the
220
00:08:49,833 --> 00:08:51,125
space that we solve the problem.
221
00:08:51,708 --> 00:08:57,250
So yeah, I mean, we were building a lot
222
00:08:57,250 --> 00:08:58,708
of websites during COVID.
223
00:08:59,083 --> 00:09:00,625
And when people realized
224
00:09:00,625 --> 00:09:03,541
that they needed to get online and
225
00:09:03,541 --> 00:09:06,083
rapidly from there, then
226
00:09:06,083 --> 00:09:08,250
they were figuring out how to
227
00:09:08,250 --> 00:09:12,000
monetize their websites, drive more
228
00:09:12,000 --> 00:09:14,375
traffic through them and
229
00:09:14,375 --> 00:09:15,916
generate a
return on marketing
230
00:09:16,041 --> 00:09:19,250
investment. And again, our services were
231
00:09:19,250 --> 00:09:20,666
operational with that
232
00:09:20,666 --> 00:09:21,500
space. I suppose the
233
00:09:21,500 --> 00:09:24,750
world we're in right now, the disruptor
234
00:09:24,750 --> 00:09:27,916
that is artificial intelligence, like
235
00:09:27,916 --> 00:09:29,166
there's a few facets
236
00:09:29,250 --> 00:09:34,083
to that where people need to figure out
237
00:09:34,083 --> 00:09:35,541
how they remain
238
00:09:35,541 --> 00:09:38,125
discoverable by large language models.
239
00:09:38,625 --> 00:09:42,916
But also, we have done quite a few
240
00:09:42,916 --> 00:09:44,708
projects in the space of
241
00:09:44,708 --> 00:09:46,500
digital transformation and
242
00:09:46,625 --> 00:09:49,500
business process automation. And we're
243
00:09:49,500 --> 00:09:52,000
seeing particularly with SMEs, the
244
00:09:52,000 --> 00:09:53,333
problem that websites
245
00:09:53,541 --> 00:09:56,166
don't talk to other systems that aren't
246
00:09:56,166 --> 00:09:57,458
an integral part of
247
00:09:57,458 --> 00:09:59,125
their operation. So building
248
00:09:59,125 --> 00:10:02,458
that little middleman in between, that's
249
00:10:02,458 --> 00:10:03,500
what we've been doing a
250
00:10:03,500 --> 00:10:05,041
lot of over the past year.
251
00:10:05,583 --> 00:10:08,750
So it is very interesting to see the
252
00:10:08,750 --> 00:10:09,916
skill sets and the
253
00:10:09,916 --> 00:10:10,875
services that you offer
254
00:10:11,083 --> 00:10:12,833
within your business where the demand
255
00:10:12,833 --> 00:10:14,458
tends to fluctuate
256
00:10:14,458 --> 00:10:16,166
over the passage of time.
257
00:10:18,375 --> 00:10:20,625
Okay, and Common Advice, I suppose, tells
258
00:10:20,625 --> 00:10:23,666
agencies to niche down to succeed. But
259
00:10:24,125 --> 00:10:26,416
Spéire has kind of effectively bought this
260
00:10:26,416 --> 00:10:28,250
trend when it comes to your
261
00:10:28,250 --> 00:10:29,708
own campaigns and messaging,
262
00:10:30,208 --> 00:10:32,875
to avoid focusing on specific services in
263
00:10:32,875 --> 00:10:33,791
favour of a broader
264
00:10:33,791 --> 00:10:36,375
approach. And how does that strategy
265
00:10:36,375 --> 00:10:41,291
work for you? Yeah, I suppose, seeing as
266
00:10:41,291 --> 00:10:42,791
we grew organically in
267
00:10:42,791 --> 00:10:45,000
nature and continue to do so,
268
00:10:45,625 --> 00:10:49,541
like in terms of our strategy, it is very
269
00:10:49,541 --> 00:10:50,416
difficult for us to
270
00:10:50,416 --> 00:10:51,583
niche down on a specific
271
00:10:51,875 --> 00:10:54,916
industry or segment where our strategy
272
00:10:54,916 --> 00:10:56,708
tends to take us is where the
273
00:10:56,708 --> 00:11:00,083
demand is. So we would assess
274
00:11:00,125 --> 00:11:01,833
the trends in terms of our incoming
275
00:11:01,833 --> 00:11:02,958
inquiries, what
type of
276
00:11:02,958 --> 00:11:03,916
problems they're experiencing,
277
00:11:04,291 --> 00:11:07,333
and match our campaigns accordingly. You
278
00:11:07,333 --> 00:11:08,583
might notice that we
279
00:11:08,583 --> 00:11:10,833
seldomly post on social media,
280
00:11:11,750 --> 00:11:14,916
and reason being, while it is important
281
00:11:14,916 --> 00:11:17,375
in terms of a channel
282
00:11:17,375 --> 00:11:18,791
for many businesses,
283
00:11:19,958 --> 00:11:23,666
we're in the space of providing solutions
284
00:11:23,666 --> 00:11:25,250
to problems. So we'd be
285
00:11:25,250 --> 00:11:27,791
very active more so in search
286
00:11:27,791 --> 00:11:29,791
engines in terms of our own strategy
287
00:11:29,791 --> 00:11:31,500
targeting really, really
288
00:11:31,500 --> 00:11:35,458
niche key terms and key questions
289
00:11:35,458 --> 00:11:37,500
being asked and making sure that we're
290
00:11:37,500 --> 00:11:41,083
discoverable in the answer. So social
291
00:11:41,083 --> 00:11:42,208
becomes secondary to
292
00:11:42,208 --> 00:11:46,333
that. But also, what's king is our
293
00:11:46,333 --> 00:11:47,458
existing customers making
294
00:11:47,458 --> 00:11:48,333
sure that they're very well
295
00:11:48,333 --> 00:11:51,208
looked after. Word of mouth is huge. And
296
00:11:51,208 --> 00:11:54,333
yeah, networking is still a
297
00:11:54,333 --> 00:11:55,458
huge part of what we do as
298
00:11:55,500 --> 00:11:58,250
well. Business, the very essence of
299
00:11:58,250 --> 00:11:59,625
business is working with people and
300
00:11:59,625 --> 00:12:00,791
working with one another
301
00:12:00,916 --> 00:12:05,000
and collaborating. And I personally
302
00:12:05,000 --> 00:12:08,291
believe that if word of
303
00:12:08,291 --> 00:12:09,250
mouth is your top channel,
304
00:12:09,541 --> 00:12:11,375
you're doing something right. If it's the
305
00:12:11,375 --> 00:12:12,041
other way around,
306
00:12:12,041 --> 00:12:13,000
there might be an issue.
307
00:12:14,583 --> 00:12:16,625
So even though I suppose you're a digital
308
00:12:16,625 --> 00:12:21,375
agency, your strategy is really based on
309
00:12:21,500 --> 00:12:25,541
relationship building. Correct. And even
310
00:12:25,541 --> 00:12:27,041
if you look at the very
311
00:12:27,041 --> 00:12:28,625
bedrock of social media,
312
00:12:30,291 --> 00:12:32,708
it is all about a lot of brands to be
313
00:12:32,708 --> 00:12:35,666
active on social media and similar
314
00:12:35,666 --> 00:12:38,250
channels to establish
315
00:12:38,250 --> 00:12:39,875
and maintain a relationship with their
316
00:12:39,875 --> 00:12:41,291
followers. For example, if
317
00:12:41,291 --> 00:12:43,791
they're a clothing brand or
318
00:12:43,791 --> 00:12:46,791
something like that. Yes, that will be
319
00:12:46,791 --> 00:12:48,041
quite an integral part
320
00:12:48,041 --> 00:12:49,625
of their strategy of their
321
00:12:49,666 --> 00:12:51,500
operating in a business to
consumer
322
00:12:51,500 --> 00:12:53,666
sense. Business to business
323
00:12:53,666 --> 00:12:55,333
might vary slightly, which
324
00:12:56,125 --> 00:12:57,833
I personally believe is always
325
00:12:57,833 --> 00:13:00,708
predominantly based on relationships and
326
00:13:00,708 --> 00:13:02,458
trust building as well.
327
00:13:02,750 --> 00:13:04,833
You know, when somebody is introduced to
328
00:13:04,833 --> 00:13:05,833
you by another customer,
329
00:13:07,583 --> 00:13:09,083
the trust tends to already be
330
00:13:09,083 --> 00:13:14,291
there. Okay. And you clearly have a deep
331
00:13:14,291 --> 00:13:15,750
passion for digital
332
00:13:15,750 --> 00:13:17,291
transformation. Was there a specific
333
00:13:17,458 --> 00:13:19,583
moment or experience early in your career
334
00:13:19,583 --> 00:13:20,791
that helped you on the
335
00:13:20,791 --> 00:13:22,416
world of online advertising and
336
00:13:22,416 --> 00:13:27,791
tech? Yeah, I suppose there are two
337
00:13:27,791 --> 00:13:30,541
facets to that interest. Number one,
338
00:13:32,250 --> 00:13:33,666
advertising has always
339
00:13:33,666 --> 00:13:37,083
interested me in terms of messaging, but
340
00:13:37,083 --> 00:13:39,875
positioning. And I
341
00:13:39,875 --> 00:13:41,125
suppose it's derived from
342
00:13:41,125 --> 00:13:43,291
what I learned in setting up a food
343
00:13:43,291 --> 00:13:44,916
product, you know, where you need to be,
344
00:13:45,250 --> 00:13:46,666
how people view your
345
00:13:46,666 --> 00:13:48,291
product on a shelf and things like that.
346
00:13:49,125 --> 00:13:51,500
And I was always very interested in
347
00:13:51,500 --> 00:13:52,916
positioning. So where
348
00:13:53,291 --> 00:13:55,166
the message and the brand needs to be
349
00:13:55,166 --> 00:13:57,541
placed in relation to the
350
00:13:57,541 --> 00:14:01,208
consumer. And I suppose my
351
00:14:01,291 --> 00:14:05,291
interest is more or less within search
352
00:14:05,291 --> 00:14:06,958
and large language models and
353
00:14:06,958 --> 00:14:08,791
how those algorithms work and
354
00:14:09,541 --> 00:14:14,375
figuring out how you as a business and a
355
00:14:14,375 --> 00:14:16,833
person can make sure that that message
356
00:14:17,791 --> 00:14:19,750
gets filtered through to the right person
357
00:14:19,750 --> 00:14:22,166
on the back of how those
358
00:14:22,166 --> 00:14:23,291
algorithms are structured.
359
00:14:25,000 --> 00:14:28,291
So the other facet then with digital
360
00:14:28,291 --> 00:14:31,541
transformation is no two
361
00:14:31,541 --> 00:14:32,958
businesses are the same and
362
00:14:33,541 --> 00:14:38,333
problem solving is always really, really
363
00:14:38,333 --> 00:14:39,166
interesting. You don't
364
00:14:39,166 --> 00:14:40,166
know what you're going to get
365
00:14:40,583 --> 00:14:44,041
and always finding a really, really
366
00:14:44,041 --> 00:14:47,708
clever value adding efficient solution to
367
00:14:47,708 --> 00:14:48,583
solve a problem that
368
00:14:48,583 --> 00:14:51,833
makes both your customer happy and your
369
00:14:51,833 --> 00:14:54,083
own team happy as well. It's always
370
00:14:54,083 --> 00:14:55,208
brilliant results. You
371
00:14:55,208 --> 00:14:57,208
never know what you're going to get
372
00:14:57,208 --> 00:14:58,416
because no two businesses are
373
00:14:58,416 --> 00:15:01,250
the same. And looking back at
374
00:15:01,291 --> 00:15:03,291
the last few years, what's been the
375
00:15:03,291 --> 00:15:05,416
biggest surprise for you in how the
376
00:15:05,416 --> 00:15:06,541
digital landscape has
377
00:15:06,541 --> 00:15:11,458
evolved? Definitely last year, Q1 in
378
00:15:11,458 --> 00:15:15,750
2025, we saw a big change with how Google
379
00:15:15,750 --> 00:15:17,708
was ranking websites
380
00:15:18,500 --> 00:15:23,166
how advertisements within search on
381
00:15:23,166 --> 00:15:25,291
Google had changed. That
382
00:15:25,291 --> 00:15:27,875
was huge. Like the behaviors
383
00:15:28,333 --> 00:15:30,041
has started to change where people are
384
00:15:30,041 --> 00:15:33,500
asking more questions
385
00:15:33,500 --> 00:15:35,833
within Google, you know, your
386
00:15:36,291 --> 00:15:37,916
searches are a lot longer, but the
387
00:15:37,916 --> 00:15:40,916
picture rates have reduced as well. I'm
388
00:15:40,916 --> 00:15:42,791
sure yourself and many
389
00:15:42,791 --> 00:15:45,041
listeners would have Google something saw
390
00:15:45,041 --> 00:15:47,333
the overview and left the
391
00:15:47,333 --> 00:15:48,208
browsing session. That's
392
00:15:48,208 --> 00:15:51,833
all you needed. So it's almost started to
393
00:15:51,833 --> 00:15:53,291
kind of shift the strategy
394
00:15:53,291 --> 00:15:54,791
and into make sure that you're
395
00:15:54,833 --> 00:15:57,833
getting mentioned as a brand along with
396
00:15:57,833 --> 00:15:59,875
trying to get traffic into
397
00:15:59,875 --> 00:16:03,000
websites and conversion into a
398
00:16:03,000 --> 00:16:07,125
lead or sale. I suppose that was what
399
00:16:07,125 --> 00:16:08,166
prompted me to kind of
400
00:16:08,166 --> 00:16:10,833
understand where the trend was going
401
00:16:11,541 --> 00:16:15,250
and learning a lot about how large
402
00:16:15,250 --> 00:16:16,750
language models would
403
00:16:16,750 --> 00:16:20,750
basically crawl through different
404
00:16:20,791 --> 00:16:22,708
sources of information and rank their
405
00:16:22,708 --> 00:16:24,291
answers, but figuring out, you
406
00:16:24,291 --> 00:16:26,375
know, how are you going to be
407
00:16:26,583 --> 00:16:29,958
the answer to that question? That was I
408
00:16:29,958 --> 00:16:30,791
think that the biggest
409
00:16:30,791 --> 00:16:32,666
surprise I had as part of that was
410
00:16:33,000 --> 00:16:36,083
the importance of PR. You know, I think
411
00:16:36,083 --> 00:16:40,208
previously to that PR as
412
00:16:40,208 --> 00:16:41,666
a
discipline and a service
413
00:16:43,250 --> 00:16:46,000
would have had to compete fiercely with
414
00:16:46,000 --> 00:16:49,208
these digital channels as
415
00:16:49,208 --> 00:16:49,916
a service. And I think that
416
00:16:50,583 --> 00:16:52,250
service has become much more important
417
00:16:52,250 --> 00:16:54,541
again because large language
418
00:16:54,541 --> 00:16:57,000
models look to the likes of
419
00:16:58,041 --> 00:17:00,750
digital news publishers as a really,
420
00:17:00,750 --> 00:17:02,375
really credible source of information.
421
00:17:03,625 --> 00:17:05,375
So that was a nice surprising trend, but
422
00:17:05,375 --> 00:17:08,750
also SEO has become
423
00:17:08,750 --> 00:17:11,708
much more important. Again,
424
00:17:11,791 --> 00:17:13,750
you know, previously would have been if
425
00:17:13,750 --> 00:17:14,291
you have a really
426
00:17:14,291 --> 00:17:15,791
well-aligned machine in terms of
427
00:17:15,791 --> 00:17:19,083
a Google ad campaign and your search
428
00:17:19,083 --> 00:17:19,708
engine optimization
429
00:17:19,708 --> 00:17:22,625
comes secondary to that, that
430
00:17:22,916 --> 00:17:24,875
would have been enough in a lot of cases,
431
00:17:24,875 --> 00:17:26,750
not every case. But
432
00:17:26,750 --> 00:17:28,666
again, it's flipped everything
433
00:17:28,666 --> 00:17:30,333
on its head and it's still made change.
434
00:17:30,750 --> 00:17:31,916
You know, there's no silver
435
00:17:31,916 --> 00:17:34,291
bullet yet and nobody has all
436
00:17:34,291 --> 00:17:37,416
the answers to how this is developing.
437
00:17:38,833 --> 00:17:40,416
But yeah, it's certainly
438
00:17:40,416 --> 00:17:42,250
been the biggest surprise within
439
00:17:42,333 --> 00:17:46,541
past few years. Okay. And like we've all,
440
00:17:46,791 --> 00:17:47,833
we're all used to, I
441
00:17:47,833 --> 00:17:49,875
suppose, hearing the term SEO,
442
00:17:50,291 --> 00:17:52,750
both geo then or generative engine
443
00:17:52,750 --> 00:17:54,583
optimization for our listeners hearing
444
00:17:54,583 --> 00:17:56,375
this for the first time.
445
00:17:56,625 --> 00:17:58,291
How would you simply explain the
446
00:17:58,291 --> 00:17:59,250
difference between the two?
447
00:18:00,166 --> 00:18:03,875
Yeah, it's a new term, but also,
448
00:18:04,083 --> 00:18:05,375
you know, there's another term being used
449
00:18:05,375 --> 00:18:07,291
as well, AEO or
450
00:18:07,291 --> 00:18:08,500
answer engine optimization.
451
00:18:09,375 --> 00:18:11,833
And the goals of both practices are the
452
00:18:11,833 --> 00:18:13,750
exact same. It's literally
453
00:18:13,750 --> 00:18:15,291
getting mentioned by large
454
00:18:15,291 --> 00:18:18,208
language models such as chat, GBT,
455
00:18:19,083 --> 00:18:22,166
Gemini, perplexity, all of these
456
00:18:22,166 --> 00:18:24,208
different models that
457
00:18:24,500 --> 00:18:27,541
somebody would be using. And, you know,
458
00:18:27,541 --> 00:18:30,041
SEO is a good foundation
459
00:18:30,041 --> 00:18:31,708
and a great bedrock to it,
460
00:18:32,000 --> 00:18:34,291
but it is hooking into what you're doing
461
00:18:34,291 --> 00:18:35,333
on social as well that
462
00:18:35,333 --> 00:18:37,041
all needs to start aligning
463
00:18:37,083 --> 00:18:41,333
along with other areas where you need to
464
00:18:41,333 --> 00:18:42,791
be positioning your
465
00:18:42,791 --> 00:18:45,750
brand as the answer to a
466
00:18:45,750 --> 00:18:49,041
question that's being asked. Okay. And so
467
00:18:49,041 --> 00:18:50,500
does the concept of SEO
468
00:18:50,500 --> 00:18:52,416
keywords still exist in the era of
469
00:18:52,416 --> 00:18:59,166
Gen AI? Yes. Like I think keywords are
470
00:18:59,166 --> 00:19:01,666
still an integral part of
471
00:19:01,666 --> 00:19:05,708
any good SEO and GEO strategy,
472
00:19:05,750 --> 00:19:09,375
but I think context and more
473
00:19:09,375 --> 00:19:14,500
conversational type content definitely is
474
00:19:14,500 --> 00:19:15,458
taking more precedence.
475
00:19:16,458 --> 00:19:19,541
But also it's, you know, you're having to
476
00:19:19,541 --> 00:19:21,083
do another type of research as well,
477
00:19:21,083 --> 00:19:22,791
which is question research. So alongside
478
00:19:22,791 --> 00:19:24,208
your keywords, you're
479
00:19:24,208 --> 00:19:25,791
determining what questions are
480
00:19:25,791 --> 00:19:29,916
being asked and the context around them
481
00:19:29,916 --> 00:19:30,833
as well and making sure
482
00:19:30,833 --> 00:19:33,833
that you're relevant and being
483
00:19:33,875 --> 00:19:36,708
positioned as that best possible answer
484
00:19:36,708 --> 00:19:37,541
that that large
485
00:19:37,541 --> 00:19:38,708
language model could give.
486
00:19:41,041 --> 00:19:42,291
What was the hardest part of
487
00:19:42,291 --> 00:19:44,083
transitioning your own mindset from
488
00:19:44,083 --> 00:19:46,166
traditional SEO to this new world
489
00:19:46,166 --> 00:19:53,291
of GEO? It completely derailed my own new
490
00:19:53,291 --> 00:19:55,750
website strategy. So even
491
00:19:55,750 --> 00:19:58,166
though we were ready to go with
492
00:19:58,583 --> 00:20:02,208
a new site map, maybe mid Q1, and then
493
00:20:02,208 --> 00:20:06,375
this trend and change started to come
494
00:20:06,375 --> 00:20:08,000
into effect. And you
495
00:20:08,000 --> 00:20:09,666
kind of have to go back to the drawing
496
00:20:09,666 --> 00:20:12,000
board and start again, which is
497
00:20:12,000 --> 00:20:14,250
effectively what I have done,
498
00:20:15,125 --> 00:20:20,333
like a GEO or AEO strategy, like it's,
499
00:20:20,916 --> 00:20:21,666
you're looking at
500
00:20:21,666 --> 00:20:23,000
about an 18 month timeline.
501
00:20:23,541 --> 00:20:26,958
So the website and your SEO strategy is
502
00:20:26,958 --> 00:20:28,125
the absolute bedrock of
503
00:20:28,125 --> 00:20:29,041
that, making sure that you
504
00:20:29,041 --> 00:20:32,458
get that right and take your time in
505
00:20:32,458 --> 00:20:34,833
doing so. You're better off
506
00:20:34,833 --> 00:20:36,125
starting with the best footing
507
00:20:36,125 --> 00:20:38,833
you possibly can and working from there
508
00:20:38,833 --> 00:20:40,875
and making sure that you
509
00:20:40,875 --> 00:20:42,791
have a website with a content
510
00:20:42,791 --> 00:20:45,208
management system that can keep up with
511
00:20:45,208 --> 00:20:46,833
those trends. So I do
512
00:20:46,833 --> 00:20:48,333
maintain that WordPress is still
513
00:20:48,333 --> 00:20:51,875
the best system to go with. But you know,
514
00:20:51,875 --> 00:20:53,625
the likes of Webflow and
515
00:20:53,625 --> 00:20:55,375
free commerce, Shopify,
516
00:20:55,875 --> 00:20:57,125
you know, there's there's still quite the
517
00:20:57,125 --> 00:20:58,041
discoverable
518
00:20:58,041 --> 00:21:01,000
alongside Squarespace and Wix,
519
00:21:01,000 --> 00:21:03,500
but you certainly have much more control
520
00:21:03,500 --> 00:21:07,291
with a WordPress website in terms of
521
00:21:07,291 --> 00:21:09,416
schema. I'm not sure
522
00:21:09,416 --> 00:21:12,166
of many of your listeners have heard the
523
00:21:12,166 --> 00:21:14,041
term schema, but in essence,
524
00:21:15,500 --> 00:21:16,541
to simplify the explanation,
525
00:21:16,916 --> 00:21:18,666
it's kind of like applying a label to a
526
00:21:18,666 --> 00:21:21,083
web page to tell the crawler
527
00:21:21,083 --> 00:21:22,250
what it is. So for example,
528
00:21:22,250 --> 00:21:24,041
if your page is a recipe, you're telling
529
00:21:24,041 --> 00:21:26,375
the crawler that it is a
530
00:21:26,375 --> 00:21:27,708
recipe to make it easier
531
00:21:28,041 --> 00:21:30,750
to measure the relevance and to help it
532
00:21:30,750 --> 00:21:33,458
index it properly. So you
533
00:21:33,458 --> 00:21:34,500
mentioned there about kind of
534
00:21:34,500 --> 00:21:36,625
some of the fundamentals, but like what
535
00:21:36,625 --> 00:21:38,500
are the fundamental things that a
536
00:21:38,500 --> 00:21:39,541
business needs to do
537
00:21:39,541 --> 00:21:41,875
to get mentioned in answers by a giant of
538
00:21:41,875 --> 00:21:44,375
engines like Chat gpt
539
00:21:44,375 --> 00:21:48,083
Yeah, so it's a term
540
00:21:49,958 --> 00:21:52,791
that I've come across called semantic
541
00:21:52,791 --> 00:21:54,791
seeding. So essentially, there's
542
00:21:54,791 --> 00:21:57,541
different facets of that. So
543
00:21:58,208 --> 00:22:01,750
your website is the best one to begin
544
00:22:01,750 --> 00:22:03,791
with. So you do your your
545
00:22:03,791 --> 00:22:07,000
answer or question research,
546
00:22:07,083 --> 00:22:09,583
should I say. Answer the public is a
547
00:22:09,583 --> 00:22:13,000
great tool. That's been bought by Neil
548
00:22:13,000 --> 00:22:16,291
Patel. And his tool,
549
00:22:16,291 --> 00:22:18,500
Uber suggest, but essentially, Answer the
550
00:22:18,500 --> 00:22:19,291
public has been around a
551
00:22:19,291 --> 00:22:20,583
long time. It's a listening post
552
00:22:21,166 --> 00:22:24,125
in relation to what questions are being
553
00:22:24,125 --> 00:22:25,291
imported into Google as a
554
00:22:25,291 --> 00:22:27,458
browser, that trend is continuing
555
00:22:27,458 --> 00:22:30,208
to grow, thanks to Google AI overview. So
556
00:22:30,208 --> 00:22:31,291
that's a great place to
557
00:22:31,291 --> 00:22:35,208
start. And then if you open up
558
00:22:35,625 --> 00:22:38,791
the large language models, one by one,
559
00:22:39,250 --> 00:22:40,541
and but make sure that
560
00:22:40,541 --> 00:22:42,875
it's either in a an incognito
561
00:22:43,166 --> 00:22:44,958
session, or making sure that
562
00:22:44,958 --> 00:22:46,708
personalization settings are turned off
563
00:22:46,708 --> 00:22:49,583
and simply putting in
564
00:22:49,583 --> 00:22:52,333
a prompt and asking it, you know, what
565
00:22:52,333 --> 00:22:53,583
are the typical questions
566
00:22:53,583 --> 00:22:56,291
you are asked if x is looking
567
00:22:56,291 --> 00:22:59,375
for x, and then you start to see what
568
00:22:59,375 --> 00:23:01,583
answers are coming up in terms of how
569
00:23:01,583 --> 00:23:02,333
people are framing their
570
00:23:02,333 --> 00:23:04,666
questions. And that gives you clues then
571
00:23:04,666 --> 00:23:05,958
on the sources. So you
572
00:23:05,958 --> 00:23:06,958
know, you can click into the
573
00:23:06,958 --> 00:23:09,291
sources of where this content is being
574
00:23:09,291 --> 00:23:10,291
positioned. So that's the
575
00:23:10,291 --> 00:23:12,833
first clue, and alongside with what
576
00:23:12,833 --> 00:23:14,916
you need to be at the writing on your
577
00:23:14,916 --> 00:23:18,250
website. So the way I'd approach it is,
578
00:23:18,625 --> 00:23:19,541
alongside your keyword
579
00:23:19,541 --> 00:23:23,083
research, assign a keyword and a question
580
00:23:23,083 --> 00:23:25,250
to your page. So it needs
581
00:23:25,250 --> 00:23:26,333
to rank for the keyword and
582
00:23:26,333 --> 00:23:28,500
needs to answer a question. That's number
583
00:23:28,500 --> 00:23:31,291
one. And number two, then
584
00:23:31,291 --> 00:23:33,583
is you align your social
585
00:23:34,666 --> 00:23:36,333
strategy alongside that your social
586
00:23:36,333 --> 00:23:38,791
media. So any post on
587
00:23:38,791 --> 00:23:41,791
your your channels video is
588
00:23:41,791 --> 00:23:43,458
getting at the moment in relation to
589
00:23:43,458 --> 00:23:48,500
reach and engagement. If you are
590
00:23:48,500 --> 00:23:49,791
uploading captions and
591
00:23:49,791 --> 00:23:52,666
transcripts along with your videos that
592
00:23:52,666 --> 00:23:53,958
enables you to be more
593
00:23:53,958 --> 00:23:55,291
discoverable, that's what the
594
00:23:56,750 --> 00:23:59,791
LLMs are picking up is the transcripts
595
00:23:59,791
--> 00:24:01,708
going along with the videos, and
596
00:24:01,708 --> 00:24:02,625
particularly on YouTube
597
00:24:02,625 --> 00:24:05,541
as well, perplexities, citations, a large
598
00:24:05,541 --> 00:24:06,708
volume of perplexity
599
00:24:06,708 --> 00:24:08,833
citations come from YouTube videos.
600
00:24:09,583 --> 00:24:11,041
There is your site map submission as
601
00:24:11,041 --> 00:24:13,500
well, making sure that you're
602
00:24:13,500 --> 00:24:15,208
submitting your site map into
603
00:24:15,291 --> 00:24:17,750
every search engine you possibly could.
604
00:24:18,250 --> 00:24:19,416
And this includes the likes
605
00:24:19,416 --> 00:24:20,666
of Wayback Machine as well.
606
00:24:21,250 --> 00:24:23,541
Then you're looking at your credibility
607
00:24:23,541 --> 00:24:25,958
channels. So Trustpilot is huge.
608
00:24:26,291 --> 00:24:28,125
Trustpilot gets hooked in
609
00:24:28,291 --> 00:24:32,166
by chat GPT quite a lot and your maps as
610
00:24:32,166 --> 00:24:35,375
well. But also, you know, there's some
611
00:24:35,375 --> 00:24:36,250
good low hang fruits
612
00:24:36,291 --> 00:24:38,416
there apart from Google My Business,
613
00:24:38,416 --> 00:24:40,375
which be Apple Business Connect. So
614
00:24:40,375 --> 00:24:41,291
making sure that you're
615
00:24:41,291 --> 00:24:43,166
discoverable within the Apple ecosystem.
616
00:24:44,583 --> 00:24:46,833
Finally, the PR element of
617
00:24:46,833 --> 00:24:49,208
it as well, seeing how you can
618
00:24:50,041 --> 00:24:53,708
position yourself to be mentioned in PR
619
00:24:53,708 --> 00:24:55,375
articles that would be
620
00:24:55,375 --> 00:24:59,458
pulled in via the large language
621
00:24:59,458 --> 00:25:01,833
models, and also your directories as
622
00:25:01,833 --> 00:25:04,166
well. Whatever directory website that you
623
00:25:04,166 --> 00:25:05,458
can get yourself into
624
00:25:05,458 --> 00:25:08,625
that is credible. That's all information
625
00:25:08,625 --> 00:25:09,333
that's getting picked
626
00:25:09,333 --> 00:25:11,541
up. So typically, you know,
627
00:25:11,541 --> 00:25:13,791
defining, you know, your product or
628
00:25:13,791 --> 00:25:15,250
service if it's answering a
629
00:25:15,250 --> 00:25:17,083
question or solving a problem,
630
00:25:17,250 --> 00:25:19,583
that's the best place to start and fan
631
00:25:19,583 --> 00:25:20,375
your content across
632
00:25:20,375 --> 00:25:22,625
then. The thing that's really
633
00:25:22,625 --> 00:25:24,791
important is consistency. Make sure that
634
00:25:24,791 --> 00:25:25,791
the message is
635
00:25:25,791 --> 00:25:29,666
consistent. And it's based on
636
00:25:30,416 --> 00:25:32,416
experience and authority. So make sure
637
00:25:32,416 --> 00:25:33,291
that what you're doing is
638
00:25:33,291 --> 00:25:37,916
original. And that there's a bit
639
00:25:37,916 --> 00:25:40,583
of evidence behind it, particularly if
640
00:25:40,583 --> 00:25:41,625
it's your own
primary
641
00:25:41,625 --> 00:25:43,833
research as well. So if there's any
642
00:25:43,875 --> 00:25:45,375
studies that you've conducted or white
643
00:25:45,375 --> 00:25:46,916
papers, they're all huge
644
00:25:46,916 --> 00:25:49,625
signals of trust for large
645
00:25:49,625 --> 00:25:52,916
language models. Okay, lots of tools
646
00:25:52,916 --> 00:25:54,083
there and tips. Thanks for
647
00:25:54,083 --> 00:25:57,083
that, Steven. Huge. And can you
648
00:25:57,083 --> 00:25:59,208
give an example, I suppose, of in a time
649
00:25:59,208 --> 00:26:00,625
you successfully planted
650
00:26:00,625 --> 00:26:01,958
a concept for a client so
651
00:26:01,958 --> 00:26:05,791
that an AI picked it up? Yeah, so I can
652
00:26:05,791 --> 00:26:07,458
name the type of business,
653
00:26:08,583 --> 00:26:10,791
but I can't name the brand, but
654
00:26:10,833 --> 00:26:14,375
it is a court based construction business
655
00:26:14,375 --> 00:26:15,791
that does a lot of
656
00:26:15,791 --> 00:26:18,500
renovations. And it was quite simple.
657
00:26:18,500 --> 00:26:20,666
They were already quite active on their
658
00:26:20,666 --> 00:26:22,750
social media channels. And
659
00:26:22,750 --> 00:26:25,291
it basically all it required
660
00:26:25,541 --> 00:26:28,708
was aligning the messaging from the
661
00:26:28,708 --> 00:26:30,916
website right across to all the channels
662
00:26:30,916 --> 00:26:31,791
that they're active on
663
00:26:32,333 --> 00:26:36,541
and making sure that they're on
664
00:26:36,541 --> 00:26:38,625
particular directories as
665
00:26:38,625 --> 00:26:40,916
well. That will be hooked in.
666
00:26:41,666 --> 00:26:47,666
And yeah, they'd be always in the five
667
00:26:47,666 --> 00:26:49,166
answers that you'd get
668
00:26:49,166 --> 00:26:50,958
when you're answering certain
669
00:26:51,000 --> 00:26:53,000
questions, especially on chat GPT.
670
00:26:54,000 --> 00:26:56,000
Google AI overviews is
671
00:26:56,000 --> 00:26:59,833
still, we haven't cracked how
672
00:26:59,875 --> 00:27:03,791
Google AI overviews work 100% yet. I
673
00:27:03,791 --> 00:27:05,250
haven't seen anybody else
674
00:27:05,250 --> 00:27:08,583
crack it yet, but it is quite
675
00:27:08,583 --> 00:27:10,791
volatile and inconsistent. And the only
676
00:27:10,791 --> 00:27:11,583
thing I can really
677
00:27:11,583 --> 00:27:15,041
attribute that to is possibly the
678
00:27:15,166 --> 00:27:19,166
personalization of the search experience.
679
00:27:19,833 --> 00:27:21,833
So Google has an idea of who you are,
680
00:27:21,833 --> 00:27:22,791
what you're looking for,
681
00:27:23,083 --> 00:27:25,750
and it's moving towards, I think,
682
00:27:26,625 --> 00:27:28,791
tailoring the results
683
00:27:28,791 --> 00:27:31,500
exactly to what you need.
684
00:27:33,708 --> 00:27:36,666
Certainly, I know which they've been
685
00:27:36,666 --> 00:27:38,125
trialing in the US in May
686
00:27:38,125 --> 00:27:41,541
last year was,
for example,
687
00:27:41,916 --> 00:27:45,041
the product feeds, shopping feeds. So for
688
00:27:45,041 --> 00:27:46,333
example, if you're searching for shoes,
689
00:27:47,333 --> 00:27:49,083
search engine already know your shoe size
690
00:27:49,083 --> 00:27:50,250
and it will only show
691
00:27:50,250 --> 00:27:52,250
shoes that would fit you.
692
00:27:53,958 --> 00:27:58,666
That's where it's going. That sounds good
693
00:27:58,666 --> 00:28:01,041
to me. There's nothing
694
00:28:01,041 --> 00:28:02,875
worse than picking on that shoe
695
00:28:02,875 --> 00:28:04,708
and realizing it's not in your size. I'm
696
00:28:04,708 --> 00:28:06,083
telling you. Yeah, yeah,
697
00:28:06,083 --> 00:28:08,416
100%. But everyone can argue,
698
00:28:08,416 --> 00:28:12,916
limits the element of discoverability,
699
00:28:13,458 --> 00:28:15,916
going outside your usual tastes,
700
00:28:15,958 --> 00:28:19,041
are we parting with that potentially as
701
00:28:19,041 --> 00:28:21,166
well? So that remains to be seen.
702
00:28:22,125 --> 00:28:22,916
Positive and negative
703
00:28:22,916 --> 00:28:23,875
to everything, I suppose.
704
00:28:24,416 --> 00:28:24,625
100%.
705
00:28:25,041 --> 00:28:25,250
Yeah.
706
00:28:25,583 --> 00:28:25,791
Yeah.
707
00:28:26,166 --> 00:28:28,958
AI moves fast and often breaks things.
708
00:28:29,583 --> 00:28:31,458
What has been the biggest challenge for
709
00:28:31,458 --> 00:28:32,375
you personally in
710
00:28:32,375 --> 00:28:33,625
keeping up with these changes
711
00:28:33,750 --> 00:28:34,791
while running a business?
712
00:28:37,291 --> 00:28:37,708
Great question.
713
00:28:38,875 --> 00:28:41,541
I think it's very tempting to
714
00:28:41,541 --> 00:28:43,666
rush into adopting something.
715
00:28:45,875 --> 00:28:48,541
And there's a lot of hype and trend.
716
00:28:51,666 --> 00:28:54,625
I suppose trend movement, for lack of a
717
00:28:54,625 --> 00:28:55,875
better term, that will come into my head.
718
00:28:56,208 --> 00:29:00,166
But I think, you know, I do certainly
719
00:29:00,166 --> 00:29:01,833
believe we're in a bubble at the moment.
720
00:29:02,791 --> 00:29:04,375
And it's making sure that we're making
721
00:29:04,375 --> 00:29:05,458
the right decisions,
722
00:29:05,458 --> 00:29:06,833
choosing the right models
723
00:29:06,833 --> 00:29:07,958
that would add value
724
00:29:07,958 --> 00:29:09,333
for us and the customer.
725
00:29:09,750 --> 00:29:11,875
The thing you don't want to compromise is
726
00:29:11,875 --> 00:29:13,458
the quality of your output.
727
00:29:15,375 --> 00:29:17,208
And we see it in digital
728
00:29:17,208 --> 00:29:18,791
transformation all the time.
729
00:29:19,958 --> 00:29:20,583
Don't make the mistake of
730
00:29:20,583 --> 00:29:21,458
starting with the software.
731
00:29:21,958 --> 00:29:25,375
You need to start with the person behind
732
00:29:25,375 --> 00:29:27,625
the activity or the process.
733
00:29:28,416 --> 00:29:31,291
Map that process and then make your
734
00:29:31,291 --> 00:29:32,083
decisions from there.
735
00:29:33,500 --> 00:29:34,291
I think people need to
736
00:29:34,291 --> 00:29:36,500
remember that AI is a tool.
737
00:29:37,791 --> 00:29:38,875
It's not a silver bullet.
738
00:29:39,666 --> 00:29:44,083
It can certainly be an assistant or a
739
00:29:44,083 --> 00:29:45,333
co-pilot to watch you
740
00:29:45,333 --> 00:29:47,583
creating, but solely relying
741
00:29:47,708 --> 00:29:50,708
on AI to create, I think, is a mistake.
742
00:29:51,541 --> 00:29:53,708
I think the general public and your
743
00:29:53,708 --> 00:29:55,458
audience can tell the
744
00:29:55,458 --> 00:29:57,041
difference between AI generated
745
00:29:57,125 --> 00:29:59,666
content versus something that has been
746
00:29:59,666 --> 00:30:02,625
made and shared by a person.
747
00:30:03,375 --> 00:30:05,375
And I think there's particularly in
748
00:30:05,375 --> 00:30:06,500
marketing, I think a lot of
749
00:30:06,500 --> 00:30:07,666
marketers are learning that
750
00:30:07,666 --> 00:30:08,166
at the moment.
751
00:30:09,125 --> 00:30:11,000
And I think there will be other
752
00:30:11,000 --> 00:30:13,875
businesses and people that
753
00:30:13,875 --> 00:30:16,125
will learn that over the next
754
00:30:16,208 --> 00:30:17,291
couple of years, I think.
755
00:30:17,708 --> 00:30:19,583
But certainly, I think the change that it
756
00:30:19,583 --> 00:30:20,416
has made, it has
757
00:30:20,416 --> 00:30:24,416
allowed us to do more volume.
758
00:30:25,416 --> 00:30:27,250
I think that's an important thing to
759
00:30:27,250 --> 00:30:28,500
recognize that if it's
760
00:30:28,500 --> 00:30:29,875
simplifying work, you're probably
761
00:30:29,875 --> 00:30:33,166
going to have to do a higher volume of
762
00:30:33,166 --> 00:30:34,125
what you do already to
763
00:30:34,125 --> 00:30:35,958
maintain the same level
764
00:30:36,125 --> 00:30:38,000
of revenue or sales.
765
00:30:39,458 --> 00:30:41,291
So you mentioned there kind of about a
766
00:30:41,291 --> 00:30:42,708
common business mistake.
767
00:30:43,208 --> 00:30:45,625
When you see businesses trying to adapt
768
00:30:45,625 --> 00:30:48,166
to this new era of AI,
769
00:30:48,875 --> 00:30:49,458
do you think that's the
770
00:30:49,458 --> 00:30:51,000
most common mistake that you see them
771
00:30:51,000 --> 00:30:52,791
making or what, in your
772
00:30:52,791 --> 00:30:55,000
opinion, is the most prevalent
773
00:30:55,041 --> 00:30:55,500
mistake?
774
00:30:56,166 --> 00:30:58,250
I think the most prevalent mistake is
775
00:30:58,250 --> 00:30:59,958
they rush into it and
776
00:30:59,958 --> 00:31:01,708
start using it straight off
777
00:31:01,708 --> 00:31:02,041
the bat.
778
00:31:03,041 --> 00:31:07,500
And then it's not necessarily aligning
779
00:31:07,500 --> 00:31:10,958
with how a person or a
780
00:31:10,958 --> 00:31:12,291
process is operating within
781
00:31:12,291 -->
00:31:12,666
a business.
782
00:31:13,541 --> 00:31:17,875
So it's never a mistake to look at
783
00:31:17,875 --> 00:31:19,708
automating or making
784
00:31:19,708 --> 00:31:20,583
something more efficient.
785
00:31:21,916 --> 00:31:23,875
But it is really important to begin with
786
00:31:23,875 --> 00:31:24,666
the person that's doing
787
00:31:24,666 --> 00:31:26,625
the job first and determining
788
00:31:26,625 --> 00:31:29,958
what is the highest quality output from
789
00:31:29,958 --> 00:31:31,208
that activity and making
790
00:31:31,208 --> 00:31:32,333
sure that it's not compromised.
791
00:31:32,708 --> 00:31:34,833
You know, is automating something for the
792
00:31:34,833 --> 00:31:35,541
sake of doing so
793
00:31:35,541 --> 00:31:38,041
going to make the task 10
794
00:31:38,041 --> 00:31:39,416
times harder because you're correcting?
795
00:31:40,666 --> 00:31:43,375
You know, that's just one symptom of
796
00:31:43,375 --> 00:31:46,458
rushing into using a piece
797
00:31:46,458 --> 00:31:48,000
of technology straight away.
798
00:31:49,041 --> 00:31:49,166
OK.
799
00:31:50,041 --> 00:31:52,541
And there's also that kind of fear that
800
00:31:52,541 --> 00:31:54,458
widespread use of Gen AI
801
00:31:54,458 --> 00:31:57,250
dehumanizes us all and devalues,
802
00:31:57,250 --> 00:31:58,333
I suppose, human interaction.
803
00:31:59,083 --> 00:32:01,000
How do you balance using these high tech
804
00:32:01,000 --> 00:32:02,375
tools while maintaining
805
00:32:02,375 --> 00:32:03,666
close client relationships?
806
00:32:09,791 --> 00:32:10,500
That's a good question.
807
00:32:11,041 --> 00:32:17,041
Like, I suppose I wouldn't let AI become
808
00:32:17,041 --> 00:32:20,666
the barrier of customer interaction.
809
00:32:20,958 --> 00:32:24,000
You know, for example, yeah, if you have
810
00:32:24,000 --> 00:32:25,041
an AI agent on your
811
00:32:25,041 --> 00:32:26,916
website that's triaging certain
812
00:32:26,916 --> 00:32:28,791
things where somebody might be looking
813
00:32:28,791 --> 00:32:31,750
for a solution to a
814
00:32:31,750 --> 00:32:32,833
really common problem that
815
00:32:32,833 --> 00:32:33,666
they don't really need
816
00:32:33,666 --> 00:32:34,958
to talk to someone about.
817
00:32:35,708 --> 00:32:36,250
Yeah, that's that's
818
00:32:36,250 --> 00:32:37,208
something that's useful.
819
00:32:37,208 --> 00:32:40,416
But if if you're actively looking at
820
00:32:40,416 --> 00:32:44,416
utilizing AI to, I suppose,
821
00:32:44,416 --> 00:32:45,375
create a bit of a distance
822
00:32:45,375 --> 00:32:48,166
between customer and service delivery, I
823
00:32:48,166 --> 00:32:49,291
think that's that's a mistake.
824
00:32:51,166 --> 00:32:54,416
I also think that from a content
825
00:32:54,416 --> 00:32:57,375
perspective and even the use
826
00:32:57,375 --> 00:33:02,208
of AI for malicious purposes,
827
00:33:02,208 --> 00:33:03,583
you know, you see a lot of the bot
828
00:33:03,583 --> 00:33:04,875
activity, things like that.
829
00:33:05,583 --> 00:33:07,958
I think that, yeah,
830
00:33:08,291 --> 00:33:09,375
waters will get muddied.
831
00:33:09,916 --> 00:33:11,416
Messaging will get skewed a bit.
832
00:33:12,000 --> 00:33:14,333
And I think we're experiencing it right
833
00:33:14,333 --> 00:33:15,833
now, right before, you
834
00:33:15,833 --> 00:33:17,500
know, AI was was coming to a
835
00:33:17,500 --> 00:33:22,541
fore as a trend that information and
836
00:33:22,541 --> 00:33:25,541
accuracy and facts have
837
00:33:25,541 --> 00:33:26,541
become a bit distorted.
838
00:33:26,958 --> 00:33:29,458
And I think, you know, use of large
839
00:33:29,458 --> 00:33:31,041
language models and
840
00:33:31,041 --> 00:33:34,541
inaccuracies, I think the volume of those
841
00:33:34,583 --> 00:33:35,458
will only increase.
842
00:33:35,708 --> 00:33:39,250
So I think having a human in the loop is
843
00:33:39,250 --> 00:33:41,416
essential for any automated
844
00:33:41,416 --> 00:33:42,958
process to be that last line
845
00:33:42,958 --> 00:33:43,333
of defense.
846
00:33:44,166 --> 00:33:45,875
You know, you don't want to be on the
847
00:33:45,875 --> 00:33:49,750
wrong side of an inaccuracy that's that's
848
00:33:49,750 --> 00:33:51,000
going to really backfire.
849
00:33:52,250 --> 00:33:54,083
And talking about, say, false information
850
00:33:54,083 --> 00:33:55,583
like Gen AI is
851
00:33:55,583 --> 00:33:57,625
essentially a very good guesser.
852
00:33:57,625 --> 00:33:59,291
And of course, it gets things wrong.
853
00:33:59,875 --> 00:34:01,500
What challenges do you face when helping
854
00:34:01,500 --> 00:34:02,833
clients correct false
855
00:34:02,833 --> 00:34:04,833
information that AI models have
856
00:34:04,833 --> 00:34:06,208
generated about their business?
857
00:34:07,125 --> 00:34:10,125
Yeah, luckily, we haven't come across
858
00:34:10,125 --> 00:34:11,375
that problem before.
859
00:34:12,041 --> 00:34:13,208
But I think it's
860
00:34:13,208 --> 00:34:14,625
inevitable that it will happen.
861
00:34:15,083 --> 00:34:18,416
And so I suppose the way I deal with it
862
00:34:18,416 --> 00:34:21,875
and advise our clients if that situation
863
00:34:21,875 --> 00:34:23,250
arises is, you know,
864
00:34:23,250 --> 00:34:26,375
identify the source, you know, where are
865
00:34:26,375 --> 00:34:28,083
they seeing that information, where their
866
00:34:28,083 --> 00:34:28,875
customers seeing that
867
00:34:28,875 --> 00:34:31,041
information and where's it coming from?
868
00:34:32,083 --> 00:34:35,375
And then look at your own channels versus
869
00:34:35,375 --> 00:34:38,291
what is not owned by you.
870
00:34:38,333 --> 00:34:40,666
So, for example, looking at your own
871
00:34:40,666 --> 00:34:45,625
website, FAQs to directly address things.
872
00:34:47,041 --> 00:34:49,625
Also, you're even having schema markup on
873
00:34:49,625 --> 00:34:52,041
pages as well to add context as to what a
874
00:34:52,041 --> 00:34:52,875
certain page is about
875
00:34:53,458 --> 00:34:57,416
would help correct to an extent where
876
00:34:57,416 --> 00:34:59,041
that page is being discoverable.
877
00:34:59,833 --> 00:35:01,666
Your own channels contact the publisher.
878
00:35:02,041 --> 00:35:04,708
Every publisher, I think, in good faith
879
00:35:04,708 --> 00:35:05,666
wants to make sure that
880
00:35:05,666 --> 00:35:07,958
their information is accurate.
881
00:35:08,916 --> 00:35:10,250
So I'd be very surprised if a publisher
882
00:35:10,250 --> 00:35:11,500
says no, if there's
883
00:35:11,500 --> 00:35:13,375
something incorrect being cited.
884
00:35:14,416 --> 00:35:16,958
Third party, your Google My Business
885
00:35:16,958 --> 00:35:19,666
socials, all that kind of stuff, seeing
886
00:35:19,666 --> 00:35:21,500
how you can address the
887
00:35:21,500 --> 00:35:24,666
inaccuracy from that angle.
888
00:35:24,916 --> 00:35:27,958
So if you're taking stock of where your
889
00:35:27,958 --> 00:35:30,416
sources are, where that information is
890
00:35:30,416 --> 00:35:31,833
being surfaced and how you
891
00:35:32,041 --> 00:35:33,000
tend to address it.
892
00:35:33,833 --> 00:35:36,083
And the thing I would say is positive
893
00:35:36,083 --> 00:35:37,625
reinforcement copy, so
894
00:35:37,625 --> 00:35:39,041
not defensive kind of copy.
895
00:35:39,833 --> 00:35:42,916
So you're confidently, you know, I
896
00:35:42,916 --> 00:35:46,916
suppose fixing the narrative where you're
897
00:35:46,916 --> 00:35:48,291
integrating FAQs into your
898
00:35:48,291 --> 00:35:53,375
website, you know, what what my company
899
00:35:53,375 --> 00:35:55,375
does versus what it doesn't do a blog
900
00:35:55,375 --> 00:35:56,833
post to that effect as well.
901
00:35:57,291 --> 00:36:00,208
Or even framing contents like common
902
00:36:00,208 --> 00:36:03,250
questions about our service X, Y and Z.
903
00:36:03,708 --> 00:36:06,458
So if you're if you're thinking more to
904
00:36:06,458 --> 00:36:08,791
do with producing explainers
905
00:36:08,791 --> 00:36:11,250
in line with aiming to get in
906
00:36:11,250 --> 00:36:13,791
contact with the channels that you don't
907
00:36:13,791 --> 00:36:16,541
own and to try and correct that
908
00:36:16,541 --> 00:36:19,083
information and be patient.
909
00:36:19,333 --> 00:36:20,916
You know, it is a bit of a timeline
910
00:36:20,916 --> 00:36:23,291
behind it and resubmit your site maps as
911
00:36:23,291 --> 00:36:24,541
well quite frequently once
912
00:36:24,666 --> 00:36:25,583
you're making these changes.
913
00:36:27,708 --> 00:36:28,541
In a nutshell, that's
914
00:36:28,541 --> 00:36:30,083
that's how I'd go about it.
915
00:36:30,333 --> 00:36:33,541
But certainly starting with where it's
916
00:36:33,541 --> 00:36:35,750
coming from and then seeing how you can
917
00:36:35,750 --> 00:36:38,625
tailor your own content and
918
00:36:39,333 --> 00:36:41,041
changing the conversation,
essentially.
919
00:36:41,333 --> 00:36:43,291
OK, give them, I suppose, the specific
920
00:36:43,291 --> 00:36:45,750
correct information that makes it hard
921
00:36:45,750 --> 00:36:47,416
for them to see the incorrect
922
00:36:47,458 --> 00:36:47,875
information.
923
00:36:48,541 --> 00:36:51,875
Yes. And, you know, the more sources that
924
00:36:51,875 --> 00:36:53,833
you could that you have at
925
00:36:53,833 --> 00:36:55,416
your disposal that are highly
926
00:36:55,458 --> 00:36:58,583
credible, for example, if you are
927
00:36:58,583 --> 00:37:01,166
encouraging your customers to post into
928
00:37:01,166 --> 00:37:03,833
Trustpilot to support the more
929
00:37:03,875 --> 00:37:05,583
accurate information about your business,
930
00:37:07,083 --> 00:37:09,500
Trustpilot is getting hooked into chat
931
00:37:09,500 --> 00:37:13,166
GPT quite regularly to ensure
932
00:37:13,791 --> 00:37:16,333
that it has that credible seal of
933
00:37:16,333 --> 00:37:18,250
approval when it's providing an answer.
934
00:37:19,375 --> 00:37:21,958
OK. And do you think that the tech
935
00:37:21,958 --> 00:37:24,375
industry is guilty of overcomplicating
936
00:37:24,375 --> 00:37:26,333
AI for the average business owner?
937
00:37:26,666 --> 00:37:28,458
And how do we fix that?
938
00:37:31,583 --> 00:37:32,958
I think in some instances, businesses
939
00:37:32,958 --> 00:37:37,375
like mine could overcomplicate in terms
940
00:37:37,375 --> 00:37:39,458
of use of jargon and.
941
00:37:41,208 --> 00:37:44,500
Probably not helping the situation with
942
00:37:44,500 --> 00:37:47,041
making it easier for
943
00:37:47,041 --> 00:37:47,958
customers to understand.
944
00:37:48,708 --> 00:37:52,541
I think what I see in general is the
945
00:37:52,541 --> 00:37:57,541
opposite where they are oversimplifying
946
00:37:57,541 --> 00:37:59,916
and pitching as this silver
947
00:37:59,958 --> 00:38:02,000
bullet that's going to solve everything.
948
00:38:02,708 --> 00:38:03,958
Like there's a lot of things out there
949
00:38:03,958 --> 00:38:06,708
where it doesn't need an
950
00:38:06,708 --> 00:38:07,958
AI model to solve a problem.
951
00:38:08,250 --> 00:38:10,916
It could be a simple API or a web hook,
952
00:38:11,750 --> 00:38:13,625
you know, just because it automates
953
00:38:13,625 --> 00:38:15,541
something, it doesn't have to be
954
00:38:15,708 --> 00:38:16,958
AI. And I think people can
955
00:38:16,958 --> 00:38:18,416
fall into that trap as well.
956
00:38:18,916 --> 00:38:20,208
You see it on nearly everything you're
957
00:38:20,208 --> 00:38:24,333
using. AI this, AI that nobody really
958
00:38:24,333 --> 00:38:25,250
knows what are doctors.
959
00:38:25,458 --> 00:38:26,166
There's need to have it
960
00:38:26,166 --> 00:38:27,083
there in the first place.
961
00:38:27,875 --> 00:38:30,000
So I think that's kind of
962
00:38:30,000 --> 00:38:30,791
where we're at at the moment.
963
00:38:31,083 --> 00:38:32,666
You know, I do believe we're in bubble
964
00:38:32,666 --> 00:38:36,083
territory where the use and the value and
965
00:38:36,083 --> 00:38:38,166
the hype are all getting the
966
00:38:38,333 --> 00:38:39,166
same airtime.
967
00:38:40,375 --> 00:38:41,833
And I do believe a
968
00:38:41,833 --> 00:38:43,166
correction is due on that front.
969
00:38:43,666 --> 00:38:45,416
Certainly. I mean, if you compare Google
970
00:38:45,416 --> 00:38:48,958
with open AI, Google has been around a
971
00:38:48,958 --> 00:38:49,916
long time to have their own
972
00:38:49,916 --> 00:38:53,333
data centers open AI while they've gotten
973
00:38:53,333 --> 00:38:56,250
the jump on chat, you know, being first
974
00:38:56,250 --> 00:38:57,666
to mark in that sense, they have
975
00:38:57,666 --> 00:38:58,791
to build their own data centers.
976
00:38:59,625 --> 00:39:00,875
And there's still loss making.
977
00:39:01,083 --> 00:39:01,750
You know, there's a lot of
978
00:39:01,750 --> 00:39:03,708
users that are using it for free.
979
00:39:04,000 --> 00:39:06,291
So when they start charging and charging
980
00:39:06,291 --> 00:39:08,333
what they're meant to be charging, are
981
00:39:08,333 --> 00:39:09,166
they going to be able to hold
982
00:39:09,166 --> 00:39:10,041
onto that market share
983
00:39:10,041 --> 00:39:11,125
that remains to be seen?
984
00:39:12,166 --> 00:39:14,750
OK, and are most owners still overwhelmed
985
00:39:14,750 --> 00:39:16,791
by the jargon or are they starting to
986
00:39:16,791 --> 00:39:19,125
genuinely understand how tools like
987
00:39:19,166 --> 00:39:21,416
chat GPT fit into their daily operations?
988
00:39:23,083 --> 00:39:25,458
It's a good question. I think business
989
00:39:25,458 --> 00:39:26,541
owners are starting to
990
00:39:26,541 --> 00:39:29,541
realize the actual value of its use.
991
00:39:30,000 --> 00:39:32,208
So, you know, they're starting to
992
00:39:32,208 --> 00:39:34,541
understand the concept of garbage in
993
00:39:34,541 --> 00:39:35,583
garbage out, you
994
00:39:35,583 --> 00:39:37,375
know, having to make sure
995
00:39:37,458 --> 00:39:39,791
that the information that you're feeding,
996
00:39:39,791 --> 00:39:40,833
whatever model you're using
997
00:39:40,833 --> 00:39:42,583
is really clean and accurate.
998
00:39:43,875 --> 00:39:47,208
But also using it for everything, they're
999
00:39:47,208 --> 00:39:49,625
starting to realize that the quality is
1000
00:39:49,625 --> 00:39:53,000
far less compared to something
1001
00:39:53,000 --> 00:39:54,083
that they would have done themselves.
1002
00:39:54,416 --> 00:39:56,458
And they probably would have ended up
1003
00:39:56,458 --> 00:39:58,208
spending the same amount of time, if not
1004
00:39:58,208 --> 00:40:01,791
more, having AI in
1005
00:40:01,791 --> 00:40:02,708
the picture of them out.
1006
00:40:04,000 --> 00:40:06,208
So, yeah, it does remain to be seen, I
1007
00:40:06,208 --> 00:40:09,625
think, certainly in terms of learning
1008
00:40:09,625 --> 00:40:11,625
experience, I think they're very much
1009
00:40:12,500 --> 00:40:15,666
made that journey. I think last year and
1010
00:40:15,666 --> 00:40:16,375
the year before was a
1011
00:40:16,375 --> 00:40:17,500
lot about the discovery.
1012
00:40:18,250 --> 00:40:18,958
How is it going to
1013
00:40:18,958 --> 00:40:19,875
provide value to my business?
1014
00:40:20,083 --> 00:40:22,041
I think that there is 2026, I think, will
1015
00:40:22,041 --> 00:40:23,291
be all understanding.
1016
00:40:24,458 --> 00:40:27,250
OK, and we often see the more
1017
00:40:27,250 --> 00:40:29,291
controversial side of the AI in the
1018
00:40:29,291 --> 00:40:30,583
headlines, for example,
1019
00:40:30,583 --> 00:40:31,583
like the recent scandal
1020
00:40:31,666 --> 00:40:34,291
involving Grok or fears around bias and
1021
00:40:34,291 --> 00:40:36,625
safety. When you're advising a business,
1022
00:40:36,833 --> 00:40:37,666
do these controversies
1023
00:40:37,666 --> 00:40:39,250
make them hesitant to engage?
1024
00:40:41,000 --> 00:40:44,791
No, I don't think it contributes to a
1025
00:40:44,791 --> 00:40:47,666
hesitance of use. Definitely a vigilance.
1026
00:40:49,500 --> 00:40:50,833
I think there's probably no harm anyway.
1027
00:40:51,750 --> 00:40:52,666
Agreed. Yeah, you took
1028
00:40:52,666 --> 00:40:53,416
the words out of my mouth.
1029
00:40:53,666 --> 00:40:56,958
I mean, I think it's all be it that the
1030
00:40:56,958 --> 00:40:59,250
grok scandal was horrendous and awful,
1031
00:41:00,041 --> 00:41:01,708
but it certainly
1032
00:41:01,708 --> 00:41:03,041
ripped off the bandaid in
1033
00:41:03,041 --> 00:41:06,583
relation to the capabilities of AI for
1034
00:41:06,583 --> 00:41:09,041
abuse rather than use.
1035
00:41:09,458 --> 00:41:11,625
And I think it's opened up a lot of
1036
00:41:11,625 --> 00:41:14,916
people's eyes in relation to, yeah, the
1037
00:41:14,916 --> 00:41:16,875
ethics behind it being
1038
00:41:16,875 --> 00:41:18,750
the operator of the large
1039
00:41:18,791 --> 00:41:21,416
language model and particularly people
1040
00:41:21,416 --> 00:41:22,250
building technology
1041
00:41:22,250 --> 00:41:23,208
in that space as well.
1042
00:41:23,541 --> 00:41:26,291
They have to be very, very careful and
1043
00:41:26,291 --> 00:41:29,458
examine all aspects of use, privacy by
1044
00:41:29,458 --> 00:41:33,666
design, the whole lot like AI.
1045
00:41:34,166 --> 00:41:37,583
As much as everybody makes an effort to
1046
00:41:37,583 --> 00:41:40,166
make sure that it isn't biased, it always
1047
00:41:40,166 --> 00:41:42,291
will be if it's being
1048
00:41:42,291 --> 00:41:44,291
told what the truth is.
1049
00:41:44,958 --> 00:41:47,375
And I think that's very evident with work
1050
00:41:47,375 --> 00:41:51,458
as well, where it's being retrained to
1051
00:41:51,458 --> 00:41:54,250
see the world in a certain way or a
1052
00:41:54,250 --> 00:41:54,958
certain relation in
1053
00:41:54,958 --> 00:41:58,208
comparison to the way Gemini and Chat
1054
00:41:58,208 --> 00:41:59,583
GPT would operate in relation to
1055
00:41:59,583 --> 00:42:03,125
answering questions. So I do think there
1056
00:42:03,125 --> 00:42:05,333
is a huge responsibility there.
1057
00:42:05,375 --> 00:42:07,916
And I think in relation to business
1058
00:42:07,916 --> 00:42:12,041
owners, certainly, I think the awareness
1059
00:42:12,041 --> 00:42:15,500
has been heightened in relation to the
1060
00:42:15,500 --> 00:42:17,541
capabilities and how it needs to be used.
1061
00:42:18,875 --> 00:42:22,625
OK, and what is one major opportunity you
1062
00:42:22,625 --> 00:42:23,833
think local businesses
1063
00:42:23,833 --> 00:42:25,000
are currently overlooking?
1064
00:42:27,583 --> 00:42:34,750
There are low hanging fruits there in
1065
00:42:34,750 --> 00:42:36,666
relation to your own rankings.
1066
00:42:37,208 --> 00:42:40,708
I think Reddit, like Reddit is a really
1067
00:42:40,708 --> 00:42:43,375
good channel because it's self-moderated,
1068
00:42:43,375 --> 00:42:45,958
but Google is hooking it into its search
1069
00:42:45,958 --> 00:42:48,875
engine results and so is Open AI.
1070
00:42:49,666 --> 00:42:53,250
So if you're respecting the community
1071
00:42:53,250 --> 00:42:56,416
that's on the platform and you're using
1072
00:42:56,416 --> 00:42:58,291
it to bolster up your own website
1073
00:42:58,291 --> 00:43:00,458
presence, it can become a really, really
1074
00:43:00,458 --> 00:43:03,708
good asset to developing your own
1075
00:43:03,708 --> 00:43:07,000
awareness in relation to people being
1076
00:43:07,000 --> 00:43:08,208
aware of your brand.
1077
00:43:09,833 --> 00:43:14,375
Trust pilot is huge as well. I think if you
1078
00:43:14,375 --> 00:43:16,458
haven't already done so, encourage your
1079
00:43:16,458 --> 00:43:18,541
customers to leave reviews.
1080
00:43:18,541 --> 00:43:21,791
I mean, we're going to be reaching out to
1081
00:43:21,791 --> 00:43:23,750
our customers as well as part of our
1082
00:43:23,750 --> 00:43:29,083
website launch to encourage them to leave
1083
00:43:29,083 --> 00:43:31,125
those signals as well of
1084
00:43:31,125 --> 00:43:32,750
their experience of our service.
1085
00:43:33,708 --> 00:43:35,791
I think every agency is different. So
1086
00:43:35,791 --> 00:43:38,416
some customers fit one agency, other
1087
00:43:38,416 --> 00:43:39,416
customers fit another.
1088
00:43:40,833 --> 00:43:42,458
And that's totally fine. So I think the
1089
00:43:42,458 --> 00:43:44,458
low hanging fruit there is
1090
00:43:44,458 --> 00:43:46,333
Reddit and Trust pilot for sure.
1091
00:43:47,041 --> 00:43:49,333
And add captions and transcripts to your
1092
00:43:49,333 --> 00:43:50,333
video content as well.
1093
00:43:51,375 --> 00:43:54,333
OK, and aside from reading up on GEO,
1094
00:43:54,541 --> 00:43:56,291
what is the best habit a business owner
1095
00:43:56,291 --> 00:43:58,791
can adapt to stay resilient in such a
1096
00:43:58,791 --> 00:44:00,458
fast changing digital world?
1097
00:44:03,000 --> 00:44:07,666
Certainly, unfortunately, people might
1098
00:44:07,666 --> 00:44:09,541
not like getting in front of the camera.
1099
00:44:10,458 --> 00:44:12,125
I think that's becoming more and more
1100
00:44:12,125 --> 00:44:17,083
important. And yeah, leveraging your
1101
00:44:17,083 --> 00:44:19,041
website with fullest potential as well.
1102
00:44:20,625 --> 00:44:22,458
I mean, your website is
1103
00:44:22,458 --> 00:44:24,083
your 24 seven shop front.
1104
00:44:24,750 --> 00:44:30,666
So thinking about how the search engines
1105
00:44:30,666 --> 00:44:33,458
and our English models view your site and
1106
00:44:33,458 --> 00:44:35,166
just chipping away page by page.
1107
00:44:36,291 --> 00:44:38,458
And you should see that evergreen
1108
00:44:38,458 --> 00:44:40,500
strategy come into effect over time.
1109
00:44:41,791 --> 00:44:44,208
But yeah, it's certainly if I can boil it
1110
00:44:44,208 --> 00:44:47,833
down, your website plus video, there's a
1111
00:44:47,833 --> 00:44:48,708
lot that could be done there.
1112
00:44:50,291 --> 00:44:52,041
Very good. Well, thanks so much, Stephen,
1113
00:44:52,041 --> 00:44:53,958
for your insights and for chatting to me
1114
00:44:53,958 --> 00:44:55,333
today on the Core Creative Podcast.
1115
00:44:56,000 --> 00:44:57,500
If you'd like to learn more about Stephen
1116
00:44:57,500 --> 00:44:59,916
and Spéire, you can find links to socials
1117
00:44:59,916 --> 00:45:01,750
website on corkcreative.ie.
1118
00:45:02,958 --> 00:45:03,583
Thanks very much.