From Christchurch to 100 million global users

This week we speak to Max Ferguson who started Lumin in 2010 to address a growing need for a robust, cloud-based way to edit and share portable document format files, better known as PDFs.

15 years on and Lumin has over 100 million users worldwide and is successfully rolling out more document collaboration products and features, managing to find success with the often-tricky freemium model.

Plus, what's the latest on Deepseek, and the unpacking the NZ government's plans for using generative AI.

Reading list

See omnystudio.com/listener for privacy information.

Episode Transcript

Speaker 1 (00:00):
We've all heard of Adobe's PDF format, the file format
that lets you share and present documents without worrying about
the software, hardware or operating system being used.

Speaker 2 (00:11):
But have you heard about the Kiwi company that has
racked up one hundred million users worldwide with its PDF
editing tools. It's an under the radio success story that's
leveraging big platforms like Microsoft three sixty five and Google
Drive to take software to the world.

Speaker 1 (00:28):
It's also a pretty incredible tale of a company that
basically threw out its existing product and replaced it from
the ground up just in time for the pandemic to arrive.

Speaker 3 (00:38):
We kind of got rid of like all the code
were written beforehand and we just started from scratch. That
kicked off a year and a half journey where we
rebuilt everything that was woman and then release that in
twenty nineteen. I can say it was a huge amount
of stress kind of leading up to that, especially those
last few weeks where you realize you're effectively going to
turn off the entire service, which at the time probably

(01:00):
thirty forty more than people using it.

Speaker 1 (01:02):
I'm Peter Griffin, I'm Ben Moore, and on the Business
of Tech Powered by a two degrees. Business this week
is Max Ferguson, founder of christ Church based software as
a service company Lumen.

Speaker 2 (01:14):
It's just over ten years since Max started Lumen. As
so many startups, begin in a dorm room at Stanford University.
It's now probably one of the most widely used New
Zealand made software products around.

Speaker 1 (01:26):
You've probably used it yourself at some point as an
alternative to Adobe acrobats product to edit a PDF. But
as Max explains, Luman has evolved well beyond a pdf editor.

Speaker 2 (01:38):
Yeah, it's very much becoming a platform for signing documents,
collaborating with documents, verifying your identity as contracts and forms
get easier to deal with online. So the interview with
Max coming up. But first Ben, we need to gauge
the lay of the land on deep Seek and look
at some AI developments out of our own government. Here.

Speaker 1 (02:00):
Well, the AI conversation continues, and I think the world
has definitely shifted this year in that regard with Deep Seek,
and I think just the attitudes to China more broadly,
with everything that's going on with TikTok and some of
the stuff coming out of the US that has ground

(02:21):
people's gears a little bit. Obviously, when deep Seek first
exploded onto the scene, there was a flurry of interest
and excitement and discussion, and then slowly, over the next
couple of weeks there's been a real questioning of some
of the narrative and some kind of people trying to
figure out what it actually means. So, Peter, what are

(02:43):
you hearing or reading about with the actual application of
some of what deep seek has done.

Speaker 2 (02:49):
Yeah, well, you know, people have been pouring over deep Seek.
Those who really are close to the AI large language
model world been pouring over it since late December. Anyway.
That was the V three model, then the R one
model that have come out, and really, you know, it
stacks up. So a lot of the experts are saying

(03:11):
that it really is quite ingenious. There's a lot of creativity.
There's a lot of innovation that has happened. Because it's
open source, they can literally get down into the weeds
and look at it, how it works. It's got a
very generous licensing arrangement around it, the MIT licensing schemes.
So people are running it. I've been running it on

(03:31):
my laptop. I know lots of people in New Zealand
who are experimenting with it. It's already built into Perplexity pro,
so I'm using it on a day to day basis.
That's how quickly it's been embraced. And this is obviously
the not the censored version, the web based version that
is available as an app and went to the top

(03:52):
of the app store last week. This is the model
that you can take away, build on top of, adapt
the code base, and have the pure sort of R
one experience. So I think the first thing is the
technical experts are sort of an agreement that there's some
really cool stuff going on here. There's maybe a question

(04:13):
mark over the actual costs of the training, how much
computing capacity was applied to this, to what extent it
relied on other models, maybe open AIS models to help
train R one and V three. So still some question marks,
but you know, a big win for China. And I
think the second thing is the just how it's sort

(04:37):
of turbocharging this sort of anxiety about the race between
China and the US for AI supremacy. And there's been
lots of people sort of dialing back those concerns or
trying to because we're in sort of result seasons for
some of these big AI companies like Meta and Mark
Zuckerberg really doubled down, tried to cam things down, said

(05:00):
we're forging ahead with our sixty five billion dollar investment
and data center infrastructure. You Microsoft is saying the same thing,
So they're basically trying to tell the market this doesn't
undermine our business model. We think all of this investment
is still warranted. And I think they probably have a
good argument there. But in the back of their minds

(05:20):
and in those companies, you know, apparently Meta has set
up a war room to look into deep seek and
what the implications of it are. They will adopt some
of those approaches and some of that technology, but fundamentally,
how is it going to change the economics the tokenomics
of the AI world. That's the big question. And surely

(05:40):
it's got to push down prices over time.

Speaker 1 (05:43):
Well you'd think so. But on the other hand, Sam
Altman said that you know that even at two hundred
dollars a month, they're not making any money from a
single user based on usage. So whether or not I
can push it down enough for the like of open
ai to reclaim the what close to coming up to

(06:03):
trillion dollars or whatever, it is that they're they're planning
on spending on developing their models. It seems difficult to envisage.
I truly really do see open ai and as more
of a research institution that's being funded by private capital.

(06:25):
I do think that anybody who is expecting to get
a return on investment might be shocked down the line
when it comes to open ai specifically. Although it's got
the backing of Microsoft, you know, it's got all that
Microsoft money in it, so maybe it will just get
folded into there at some point and some of those

(06:47):
investors will see a little bit of return, but yeah,
it's just a huge amount invested for what that has
been produced. If they are still struggling to make any
money at two hundred dollars an hour, as two hundred
dollars a month.

Speaker 2 (07:00):
So yeah, and that's for that pro chat GPTs pro
for basically enterprise licenses two hundred US a month. And
that plan allows you access at the moment to open
ayes new agent, which is called Operator. It's just been

(07:21):
put out sort of as a beta release, and so
you know, this really I think is the is the
beginning of trying to accelerate development to try and claw
back some of that money and be able to charge
more of this. We've talked about agents before, where you
get this AI on your behalf to undertake tasks. So

(07:42):
Operator we'll look at a web browser, will look at
what is going on in that web browser, and potentially
just by looking at it, rather than needing an API
or anything like that, it can read what's on the
screen and it will suggest things and make decisions on that.
So they're working with a bunch of sort of big
US companies to test that out there, likes Uber and

(08:03):
those sorts of companies, so trying to do day to
day mundane tasks and automate them. So Operator is really
their big push. We've talked about Salesforce and others Microsoft
who are rolling up agents, so we've yet to see
what the cost implications of that. And of course they
launched last week chat gpt gov as well, so a

(08:25):
lot of American government bureaucrats have been using chat GPT.
They'd probably be less of them soon with the DOGE initiative,
but they're well into it. Ninety thousand of them have
been experimenting with it. So chat GPT Gov is basically
chat GPT, but in a secure environment in the Azure

(08:47):
cloud and all of those policies and that that federal
workers need to adhere to. I sort of built into
it in terms of how you access and share and
use information, so that I think that the answer for
open AI to competitors like deepseek is that they're a
big trusted American company. They need to forge your head

(09:08):
now with monetizing their large language models in ways that
are going to see big companies and governments and consumers
using their agents and their AI assistance.

Speaker 1 (09:21):
Yeah, I going back to operator, I have to say
I don't quite get it, and like I kind of
understand that it's a cool idea for a robot that
can navigate a website, right, Like that's kind of interesting theoretically,
but it's always going to be significantly worse than apish

(09:47):
And if you have, like if I would rather have
a chat GPT integration with Uber API or like door
Dash API or whatever else, if we can figure out
a way to improve that system, make that more accessible,
because these are websites designed for human navigation and for

(10:08):
human eyes and human brains, and I think maybe as
a test of being able to emulate a human like capability,
it's a good experiment, but you know, looking at some
of the reporting around the use of operator, it's been okay,
it can do some things. It needs some intervention to

(10:29):
help it along, but for the most part it be
more efficient just to do it yourself from the start.
Whether that can ever become better than that, whether it
can reach the point where it's fully useful, is a
big question mark for me, because especially when you've got

(10:50):
the possibility of creating something that is native to computers
to be able to integrate with. So yeah, as an
interesting one, cool, but I'm not sure of the impact
that it's going to actually have in the long term.

Speaker 2 (11:05):
We'll see it's powered by a computer using agent model,
which basically uses the buttons, the navigation menus, the forms
on a web page, and I guess part of the
reason why they're doing that is if you want to
interact with several different services, I don't know how it
works to have several APIs maybe operating at the same time.

(11:27):
So maybe there's some limitations that they're trying to overcome there.
But the sorts of companies that they're working with Door Dash, eBay, Instacart,
price Line, stub Hub, I think it will be pretty
simplistic sort of stuff. You know, do you want to
buy something on eBay? Do you want to search for

(11:47):
tickets on stub Hub? Can you get an AI agent
to do this? And look, I don't have huge expectations
for agents in maybe on the likes the Salesforce platform,
where they're controlling the whole platform, they've got access to

(12:07):
all of your data. It's probably a bit easier to do.
But this sort of stuff where you roam across the
web is going to take a lot longer to get right.

Speaker 1 (12:15):
Yeah, but maybe maybe I'm just being a luddyite, you know,
maybe I'm just not Maybe I'm not the visionary, the
Silicon Valley person who can see the future.

Speaker 2 (12:25):
Yeah, i'd see the demos hopefully in the next few months.
Some people do innovative things with it. But meanwhile, back
here in New Zealand, I mean, it's it's been really good.
I've seen some really thoughtful commentary on deep seek from
AI community here and startups and that. But looking at governments,
some developments being last week to give public servants a

(12:48):
bit more guidance on how to use AI. This is
the framework which we thought was going to sort of
come out last September, so quite a big delay on that.
Do you make of it? This plan on a page.

Speaker 1 (13:02):
Yeah, it's pretty basic, isn't it. I Mean it's a
busy page, but kind of really just comes down to
figure out where it might be useful and then use
it like that seems to be the kind of the
kind of distillation excuse the term of what they've got
on the page there. I think more interesting is kind

(13:27):
of the responsible guidance that they there's a responsible use
guidance that they released. I think that kind of might
help people to think through that that thing of well,
where I'm using it, is it going to be useful?
And how am I going to use it? It gives quite
a good range of kind of guidelines for public service

(13:50):
about what are the appropriate ways to use it, what
to be cautious of, how it can be useful. And
I talked to Freth Tweety from Simply Privacy, who's been
a guest before, and she said that it avoided using
risk language, which is her way of saying, you know,
it's more like gentle kind of gentle boundaries for the

(14:11):
users of the public service. So overall, I think that
guidance is probably a bit more interesting and useful to
look at if you want to understand how the public
service is planning to use it, whereas that one pager
is kind of like, oh, yeah, we're going to use
it all over the place and it's going to be great.

Speaker 2 (14:28):
Yeah, lots of AI buzzwords. There's lots of you know,
the OECDS principles and responsible use of AI, inclusive growth,
human rights, transparency, accountability, all of that sort of stuff,
which is great, you know, before we race off and
use this. But I think, as people like Frith have
pointed out, there's really not much detail about, you know,

(14:50):
how you implement this and the capability, what is being
done to build the capability. And that is a real
problem because a survey by DA of government departments last
year they asked them what is the big barrier to
the uptake off AI, and they said skills and capability.
That was by far that the big barrier. We just

(15:12):
don't have enough people, particularly with what is going on
with the restructuring off the public service. They're not really
empowered to spend time thinking about this. So this is
all well and good to have a framework and to
be referring to that as you develop AI. But we've
got to get going, we've got to start building stuff,
and at the moment it just seems like a massive

(15:34):
lull because of the disruption, the lack of funding. So
what is the plan to actually get a workforce in
the public sector, you know, sandboxing things, experimenting, sharing ideas,
launching things. That's what Judith Collins, as Digitizing Government Minister
wants two years on from chat GPT's debut, really not

(15:55):
seeing anything substantive other than what Callahan Innovation did with
its rather small experiment gove GPT and that team. What's
the future of them once Callahan's disestablished.

Speaker 1 (16:07):
Yeah, yeah, And I think it kind of speaks a
little bit to the fact that it does take more
time then maybe Judith Collins might have expected to actually
get these projects going and figure out where it is
effective and time saving and products of it enhancing and
where it's just doing the same doing a similar thing

(16:28):
at the similar pace. And that's probably a general issue
across the technology, is that there's a lot of use it.
It will enhance productivity. Just get in there, get in
there and use it and use it, and it's like, well, Okay,
that's a good first step, but then that experimentation needs
to be a little bit more focused to say, here's
an idea of what I can do. Let's test it

(16:50):
to see if it works, and let's look at the output,
let's assess it.

Speaker 2 (16:55):
Let's look at.

Speaker 1 (16:55):
How long that actually takes from beginning to end versus
a human doing it and balancing all of these things
and then putting them into production workflows. Actually a little
bit more difficult, not impossible, but it does take dedicated
time and effort and planning and thought.

Speaker 2 (17:12):
Yeah, oh well, at least they've got the they've laid
the groundwork now, so everyone should hopefully be on the
same PowerPoint slide.

Speaker 1 (17:22):
Well there's only one so now, christ Church Company lumin
Like many New Zealand startups, it grew out of a
frustrating problem.

Speaker 2 (17:38):
Yeah, think back over ten to fifteen years before cloud
platforms became entrenched about what a pain in the BUTTT
was editing and sharing collaborating on documents.

Speaker 1 (17:49):
Adobe ranged supreme with its Adobe Acrobat software and the
PDF format.

Speaker 2 (17:54):
Great concept of universal file format you could open anywhere,
but Adobe wanted to charge you a fair chunk of
cash to use it, and the free PDF reader really
wasn't that great.

Speaker 1 (18:05):
So into Max Ferguson, who had studied civil engineering at
both Canterbury University and Stanford before getting into computer science
and software, and he's seen the nightmare of swapping documents
in the construction industry where files were edited and emailed
around or swapped on USB sticks, which was pretty much
the norm everywhere at the time.

Speaker 2 (18:26):
So he set out to change that and rode the
SaaS wave to global success.

Speaker 1 (18:33):
And so, Peter, you sat down with Max Ferguson from
Lumen to have a discussion about the company, its history
and where it's headed. So let's have a listen to
that interview now.

Speaker 2 (18:48):
Max, Welcome to the business of tech. Congratulations one hundred
million customers for LUMAN, New Zealand founded software company. That
is staggering. You know, I'm trying to think of companies
founded in New Zealand that have that sort of level
of reach. I mean, zero has something like four point
two million customers, have done incredibly well. Can you think

(19:10):
of any other software company that has that sort of
usage around the world founded in New Zealand?

Speaker 3 (19:16):
Not too many, Peter, thanks for having me, but yeah,
not too many. It is quite a large scale achievement
for Lumen, and you know, we're all about scale, but
I can't think of it too many others that have
sort of reached that scale coming out of New Zealand.

Speaker 2 (19:34):
So break that down for US one hundred million customers.
I'm a big Google Workspace user, so I've used Lumen
as a free user because it's just rarely that I
need to edit a PDF. But this is a bit
of software that sits on in ecosystems like Google Workspace.

(19:54):
I had to look at You've had something like twenty
three million downloads off Lumen in the cloud pdf environment,
So I guess it's not really a download. It plugs
into your Google Workspace so you can edit pdf sign documents.
And where is the rest of your user base coming from?

Speaker 3 (20:13):
Yeah, So we you know, at Lumin, we're really big
on integrations. We set out to solve the problem of
allowing people to work with PDF files more easily, so
merge files, signed pdf files. A lot of people are
kind of filling out forms, and we integrate with a
lot of the major players, so with Google Workspace as

(20:34):
you mentioned better, and we integrate with Microsoft Office three six,
five and we're announcing integration with Salesforce leaders. Yeah, so
it's really us integrating with those other platforms and then
allowing our users to kind of work with pedif files
and get documents signed.

Speaker 2 (20:53):
So presumably then would the majority of your customers be
working sort of in the Microsoft three six five Windows environment.

Speaker 3 (21:02):
A lot of our customers are using Windows, but a
lot of them are just accessing Lumen as a cloud service.
So they'll just type into Google search, you know, emerged
pdf files or search for Lumen and they'll use our
service directly at luminpdf dot com and they can upload
documents to that and get them signed or for platforms,

(21:23):
whatever they're looking to do on the day.

Speaker 2 (21:26):
In the Google world, you're definitely one of the big
plays here. There's another one called doc Hub. I'm sure
you know well has had something like fifty eight million downloads,
but there's this sort of cluster of companies that are
allowing you to do smart things with PDF. And after
that you go down to small Pdf, which is four
million downloads. So you and dock hub are definitely in

(21:48):
that ecosystem anyway, the big ones, this all sort of
congregates around the PDF software. Adobe Acrobat, which created that format.
It's been around for decades. But there's clearly obviously a
lot of reasons why people don't just want to use
Adobe Acrobat, which incidentally in the Google workspace and the

(22:11):
Google Cloud has almost exactly the same number of downloads
as Luhman has. So why are people not just using
the industry default, the creator that came up with this
piece of software, But it's going to all these alternatives,
particularly lumen.

Speaker 3 (22:25):
So what we do at Luhman pdf is a part
of it, but it is just a small part of
what we do. We like to help customers with all
of their workflows, so we like to drive efficiency and
a lot of that is, you know, signing workflows. So
here in New Zealand, we work with Renti, which is

(22:47):
a property management company and we actually handle about a
good portion of all of the rental agreements that are
signed in New Zealand. So Luhmann does a lot more
than just the PDF editing part. Of course, we're very
proud of our PDF editor and it gets a huge
amount of usage from people that are looking to highlight documents,

(23:07):
signed documents and share them. But it is just a
small part of kind of what Luman does. Luman helps
customers from all different aspects, all different verticals with their
kind of document workflow jobs.

Speaker 2 (23:23):
And what does it involve when you're working with pdf S,
which is a sort of a format that came out
of Adobe, do you have to license that technology from them?

Speaker 3 (23:31):
No, So a few years ago, I think it was
two thousand and six, Adobe actually open sourced the PDF standard.
And I don't know exactly they're thinking from, you know,
from it, but my understanding was was they really wanted
to build an ecosystem around that. So they had the
option to sort of keep Adobe PDFs as a closed

(23:52):
sort of closed a walled garden. But then you know
that that wouldn't allow the ecosystem to flourish, So that
wouldn't allow new products to come into the market, and
it would probably start to hander widespread adoption of the format.
And so Adobe decided to open source that format and
allow other players to come into the market and offer

(24:14):
services around that. For Adobe, the PDF part of their
business is actually a small you know, it's still significant,
but it is a small part of their business. They
offer a whole lot of other products, specifically around their
creative cloud or their document management and so by open
sourcing that they were you know, they allowed other products

(24:34):
to come into the market and kind of help build
out an ecosystem of tools which work around the PDF
far format.

Speaker 2 (24:42):
LUMANN was founded in twenty fourteen. Back around that time,
you were a civil engineering student at the University of Canterbury,
so very much going on that track into building things
infrastructure and alike in New Zealand. But what was it
then when you were studying at UNIVERSE that made you
think there's an opportunity here and I want to go

(25:03):
into entrepreneurial endeavor around software.

Speaker 3 (25:06):
I was quite interested in software from you know, early
days when I was in high school. I wrote a
few games on a calculator and kind of I was
just there was my first foray into learning a bit
about software. I then studied Civil engineering at UC where
I got I was able to take a whole bunch
of software courses and learn about programming and learn about software.

(25:30):
But I think the real entrepreneurial part didn't get started
until I left and went into the industry and realized
that some of the tools we were using in the
industry needed to be updated. There was a real opportunity
for innovation there and for me that you know, that
came around kind of managing files and just seeing how
that was done. So at the time we had you know,

(25:53):
I was working in construction. We had a side office
and we'd print out documents and kind of draw on
They were plans for bridges and that sort of thing.
So we draw the changes and then we would need
to get those back to the design engineers that that
kind of came up with those designs, and so we
either needed to take photos with our phone or we
could sort of scan those back in and email them.

(26:13):
But as you can imagine, very quickly it became quite messy.
So we had all these different plans with kind of
hand drawn you know, annotations and markups on them, and
you know, multiple versions, so it became quite complex, and
so that inspired me to create Leman, where you know,
everything was done in the cloud and you could just

(26:34):
have a single file where people were working on it collaboratively,
so single PF document and each person could be making
those annotations in real time and you wouldn't need to
sort of email around different versions or anything like that.
So it was a time where you Google Drive and
drop Box were really helping like starting to solve that

(26:56):
problem as well. And so by integrating with those platforms,
we were able to build something out which allowed people
to sort of all get on the same page quite quickly.

Speaker 2 (27:05):
Yeah, and you sort of resisted the temptation to just
build a sort of a BIM system building information management system,
which would have employed all of the knowledge you had
about the construction sector, and that would which would have
been quite a niche product. And there's been a lot
of innovation since in the last decade in that space.
But you decided to go big. Everyone is having this

(27:26):
problem about keeping on the same page literally with one document,
but you know that that was probably quite daunting. It's
like you're in christ Church, you know, tiny development house.
How do you even think about getting this thing global
exposure at that point?

Speaker 3 (27:45):
Yeah, definitely for us, Like for me at least, you know,
scale is always very important in building software. So software
works on this beautiful economy of scale where if you
can build something that has a huge impact for a
lot of people, then you can find a business model
and develop that software more. So, you know, from for me,

(28:08):
scale was really important from day one, and I think
as kiwis as well, we're sort of we're we have
to think that way because if you build a piece
of software like a PDF editor that's you know, only
for New Zealand, there's just not going to be enough
scale and enough usage here to develop that into a
real world leading product. So for me, you know, I

(28:29):
was really interested in how do we how do we
scale this thing, how do we kind of get this
thing growing really really quickly? And that's where the old
saying around you know, standing on the shoulders of giants
comes in. So we got really into integrations from day one.
So we integrated, you know, the first version of Lemon
integrated with Dropbox, Google Drive, and one Drive I think

(28:52):
it was SkyDrive at the time from day one, and
that allowed us to extend those platforms and reach a
lot of people there really quickly. And we've continued that
kind of philosophy right through with Lomans. So whenever we're
building new products, we're trying to build things that will
have an impact on a huge amount of people. And

(29:13):
we're also working really closely with partners, so we're figuring
out where we can integrate and add functionality to other
people's platforms.

Speaker 2 (29:22):
So you're in these ecosystems now drop Box, Google Drive, OneDrive,
SkyDrive back then it's fine. It's a bit like being
in the app store, but there's two million other apps.
So how do you actually then get customers to decide
to choose lumin pdf over some of the alternatives.

Speaker 3 (29:41):
Yeah, we're incredibly product focused for us, so you know,
all of them from day one. All of the profits
we made from the company, we hoard as much of
that as we could back into product development. You see
a model now with a lot of software where you know,
they build the software for a year or two, they
put it up there and then never updated again. We're

(30:02):
you know, we think it's always evolving, and we're always
looking for ways to improve the platform. Things like the
auto sync so when someone you know, when we've got
multiple people editing a PDF and all of them are
sort of drawing at once. That took us years to
get that really really refined so that it would keep

(30:23):
everyone's annotations so they could all see them in real time,
and then so that we could also save the copies
back to Google Drive or drop Box or one drive
without any errors. That's the sort of technology that takes
years to really develop and refine, so we kind of
continue to invest in product and building the best product

(30:44):
for our customers.

Speaker 2 (30:45):
And then, of course, the great thing about these ecosystems
is the transparency of what the user thinks of them.
So there's star ratings. For instance, I think you have
three point seven star rating on Google Workspace. So how
important was those those reviews that feedback from this growing
customer base. How important was that to actually showing people

(31:08):
this is the best one or one of the better ones.
I need to maybe download this one.

Speaker 3 (31:13):
Yeah, I think it's incredibly important. We read every review
that gets posted on every platform. So whether it's you know,
G two, Google Workspace, you name it, whatever platform, we've
got our product owners reading every single review. And that
means that if our customers give us feedback, like you know,
the load time was too slow or something like that,

(31:36):
then we'll be working on that and to fix it.
And the amazing thing about reaching that one hundred million
user milestone is that we get a lot of feedback
and we love we love feedback. We need more of it,
but we use that to make the platform better. So
maybe you know, you might have a really niche use
case for for PDF editing for example, you know, you

(31:59):
may be really big on you know, using our green
highlighter or something like that. If you find something wrong
with it and you give us feedback and then we
can go and fix that. And because we've got that
scale of one hundred million users, it means that all
of those niche use cases, however niche they are, will
get covered and we can build the product and make

(32:20):
it as good as it possibly can for our end users.

Speaker 2 (32:23):
Like a lot of sort of software as a service offerings,
it's a sort of a there's a freemium model there,
so you can get the free one on Google. That's
three documents you can edit per month, which for a
business is not really going to cut it fine for
me as an individual. What's your experience been like watching
customers go from maybe testing out the free one to

(32:44):
then going to the nine dollars US plan and maybe
even you know, the more sort of premium when I
think one hundred and ninety nine dollars is the top
end business plan. What's been the key to sort of
getting people comfortable with the product and then sort of
upselling them to become regular paying customers.

Speaker 3 (33:01):
Yeah, there's so we love it. We love that model,
and the reason why is I think we've we've probably
all been in this situation where we've gone out and
bought something and then we found that it wasn't quite
what it was hyped up to be. So, whether that's
a physical product like an ice cream, or whether that's
a digital product where you had to buy it ahead
of time or using it and then you kind of

(33:22):
get that disappointment of, oh, wow, this is not as
great as I was expecting. So with the freemium model,
you know, we think it's fantastic because you can go in,
you can try the product, you can see if it
works for you. Like you said, Peter, you can use
it for a few documents every month and see if
you like it in the way it works, and if
it's working out really great for you, then you can

(33:43):
kind of upgrade and use one of our pro pro features. So,
you know, we really like that model. We think that
it allows our customers to try out our products. They
know what they're buying. There's no kind of smoke and
mirrors around, Hey, this is going to be so great,
but then when you buy it, it's not, and we

(34:03):
get exposure around the whole world. So lots and lots
of people try that products, so I think that that
model works really well.

Speaker 2 (34:11):
That's great, and it is diversifying the sorts of things
that you're doing with lumin PDF you talked about. I
think it's called lumin sign, which is signing documents. You know,
the big player in that market is DocuSign. It's probably
got the most brand recognition. There are lots of alternatives now,
But what do you need to do technically to conform

(34:31):
to the standards that make a legal document a legal document?
Obviously it's one thing to technically allow someone to sign
a document that has legal standing, But is there is
there regulation underpinning that that you have to conform to.

Speaker 3 (34:48):
Yeah, definitely, And unfortunately it varies in a country by country,
so every country has slightly different legislation around what it
takes to sign a document digitally. Ins we have sort
of one of the more vague pieces of law, I
would say, around signing. So for almost any document, you know,

(35:09):
as long as you're applying a signature to that, it's
going to be legally valid. There's a few documents that
you can't sign digitally in New Zealand, like Wells, but
we have a very vague legislation. Some of the other
countries will say that you need to you know, be
signing a document in a way that uses cryptography that

(35:30):
actually applies like a cryptographic signature to the document for
it to be legally bind it.

Speaker 2 (35:34):
Going back a little bit, at some point, I mean
things were growing, Well, at what point did you realize
this is actually a really viable business. We're seeing one
hundreds of thousands of new customers coming on every quarter.
This is actually going to go somewhere. What point was.

Speaker 3 (35:47):
That we got We got really a great initial traction
with Lumen from the start. So I've built a number
of companies and apps before Lumen, probably around ten, like
small any apps and that sort of thing, and turned
out that Lemon was the one that kind of got
the most traction the quickest. When we hit the sort

(36:08):
of the million user point, which was after a couple
of months, you know, I realized that there was a
real need there and there was a lot of people,
you know, I realized there was a lot of people
looking for a solution that would allow them to draw
on documents and annotate documents really easily without kind of
having to go through the email workflows and the sending

(36:30):
documents around or downloading. So that yea one million user
point was really really the mark for me there where
I realized that, you know, we had some good product
market fips.

Speaker 2 (36:39):
And at some point along the way you did a
major overhaul of the app. What really drove that.

Speaker 3 (36:46):
Yeah, So we built the app incrementally from twenty fourteen
to about twenty nineteen. We kept adding new features, and
we just you know, like we talked about earlier, we
read all of the customer reviews, and when a customer
you know, ask for certain features, we kind of added them.
But by the time we got to early twenty eighteen,

(37:07):
we realized that, you know, things weren't as streamlined and
as smooth as we wanted them. And at that point
we could have left the product as is, but you know,
we thought we could do more, and we thought we
could build build something better. So we actually commissioned like
a full rebuild of the app from the ground up.
We kind of got rid of like all the code

(37:28):
we'd written beforehand, and we just started from scratch with
new designs for the entire platform. That kicked off a
year and a half journey where we rebuilt everything that
was Lumen and then released that in twenty nineteen as
a brand new piece of software. And I can say

(37:49):
there was a huge amount of stress kind of leading
up to that, especially those last few weeks where you
realize you're effectively going to turn off the entire service,
which at the time had probably thirty four minion people
using it, and like cut instantly over to a new
service which had been rebuilt, which did a lot of
things better, but it also didn't do everything the old
one had done. So it was tremendously stressful. But we

(38:14):
got a really great outcome out of that, and the
customers loved of the new rebuilt product, and you know,
that really triggered our graft to get to where we
are now.

Speaker 2 (38:23):
In just in time for the pandemic, which presumably saw
a bit of a bump in usage as well, with
people needing to exchange and work on documents remotely working
from home.

Speaker 3 (38:33):
Yeah. So the way I like to think of that
is that, you know, we saw digitization for many years,
So from twenty fourteen through it to twenty twenty, we
knew a lot of people were moving to the cloud,
a lot of people were becoming familiar with you know,
tools like Google Docs and Google Drive. A lot of
businesses were moving over, but there was still some that
hadn't you know that that were based in the paper world.

(38:56):
And when the pandemic happened, we really saw you know
a lot of those businesses modernized really quickly. I think
they were always going to get there. It just brought that,
you know, I just should have just forced that change
to happen a lot quicker. And so there was in
the tech world there was a bit of a sort
of a boom and a bust where people thought, you know,
that growth would continue, you know of say ADOPTU signed

(39:19):
the stock prices went crazy, like that growth would continue forever.
I think we were a bit more measured in the
way we looked at it, and we just saw it
as a lot of businesses kind of digitizing and modernizing
quite quickly.

Speaker 2 (39:31):
And at some point you also started to expand internationally
in terms of your software development. Tell us about that,
going from a small team in christ Church working on
Luhman to running teams in Southeast Asia.

Speaker 3 (39:46):
Definitely. So I was always really passionate about kind of
remote work and global teams from way before the pandemic.
So if we go right back to twenty twelve, I
was quite interested in this idea that you know, multiple
people from around the world could come together. Before Luhman
even I worked with developers in India, South Africa, Ukraine, Vietnam,

(40:09):
Philippines and really got I got a feel for how
that might work. This was, you know, before the days
of zoom and video calls, like, it was all pretty janky,
and so there was like a lot of email going
on and maybe like some Skype and WhatsApp, but it
was I was really fascinated with you know, what happens
when a group of people from you know, different cultures
all come together to build something. And so the second

(40:32):
person that started working with me on Luhman, he was
a Vietnamese national living in the States. I was living
in the States at the time as well, and we
started working on this thing kind of really really early on.
He was much better on the software side than me,
so in terms of programming, you know, the guy was
was absolutely incredible and we just built this you know,

(40:53):
this working relationship for him and you know, his family.
At the time. He wanted to kind of moved back
to Vietnam and build out a team there. So he
left the States and moved back to Vietnam and started
building out what would end up being kind of our
Vietnamese development team. And that's how we kind of we
grew out. Since then, you know, we've we've kept a

(41:16):
pretty global approach on things as well, So we've got
teams and many countries all around the world working away
and we all just come together in this digital space
to build these products.

Speaker 2 (41:27):
Yeah, and I think it's a model that a lot
of New Zealand companies are pursuing now. I know game
developers who have whole teams in South America because literally
getting getting the talent in New Zealand is really difficult.
We've got a small talent pool. At one point, it
was really attractive to come here. It's sort of gone
the other direction for various reasons at the moment. But

(41:47):
being able to reach out to a global talent pool
and not expect people to you to be coming into
the office in christ Church has got to be a
huge advantage.

Speaker 3 (41:55):
Yeah, it definitely helps our team here in New Zealander
is amazing. And you know, one of the reasons I
decided to move back from the US to New Zealand
was because I thought there was a real untapped talent
poll especially down here in christ Church, got amazing people here.
What we do struggle with sometimes is finding specialists. So
if we if we need someone that's got you know,

(42:17):
say a very a very concentrated expertise and you know,
some bizarre part of software, then say, finding that in
New Zealand can be more challenging because we simply don't
have the population to find those specialists. And so there's
really two options then to handle those scenarios when they

(42:38):
come up. One is potentially bringing someone in from overseas,
so you can kind of run recruitment and then you know,
find someone overseas that might have the specialties and then
bring them into New Zealand. I know a bunch of
companies do that and it works quite well, but it
can be challenging as well because relocation then becomes part
of your your onboarding. Effectively. For us, we just like

(43:00):
to go out into the global market and find those
specialists where we need them and then just build a
really great relationship with them and tie them back to
our operation here in New Zealm.

Speaker 2 (43:10):
One of those specialties that I'm sure you're drawing on
to a greater degree than ever is artificial intelligence. You know,
particularly with you if you if you're doing this deal
with Salesforce, you know they're all about agent force, you
know AI agents, their their AI cloud and all all
that sort of stuff. So all of that data and

(43:31):
whatever's done on that platform will be documented at some point.
So what's your journey being like sort of embracing particularly
this latest wave of AI, which is really around generative AI,
and how all of that is reflected in how we
work with documents.

Speaker 3 (43:46):
Our approach to AI was that, you know, we wanted
to really give it some time and just see, you know,
where can we get the most value for the customer
Because I know a lot of a lot of companies
that are jumped into AI really early on, and they're
released some features that they were super cool, but we're
there actually going to help businesses grow and thrive, and

(44:08):
a lot of there was a lot of kind of
marketing hype around that. So what we did is we
worked away in the background for several years and tried
to figure out, you know, where can we actually deliver
huge value to our customers with this new technology. And
so we've been rolling AI into kind of a bunch
of our products over the last few years, and we've
got some huge product releases coming up this year that

(44:30):
will be, you know, but a fully rebuilt using the
concept of AI. My personal belief is I think that
this is this is sort of a fire or light
bulb moment for humanity. We're unlocking something here that that
has never been seen before, something incredibly powerful. So I

(44:51):
think it will change the software space. But there's also
a lot of kind of marketing hype and a lot
of things that are not really AI that are being
branded as AI, and so we've got to be careful
to make sure that, you know, we're actually building stuff
that's helpful for our customer that delivers tremendous value, and
we're not just going to spray painting the outside with pretty.

Speaker 2 (45:12):
Colors, especially when you know some of the documents your
customers are dealing with, they're very important documents there deals
with with their own customers, their legal documents, rental agreements
and the like. So a lot of the AI might
be done in a different platform, but making sure that
that's all accurately represented, and the sort of auto generation

(45:34):
tools that Luman has built in as well to make
document creation and editing easier. It's got to be accurate, right,
and we do have still quality issues around hallucinations and
inaccuracies in some of these large language models, definitely.

Speaker 3 (45:48):
And we're learning how to use it as people and
as business owners as well. So you know, there's there's
a lot of experimentation going on about what's safe, how
can we use this in a safe way, And businesses
are learning where they can get the most value out
of it and you know, where they should stay clear.
But we think there's you know, tremendous potential to help

(46:09):
eliminate some of the mundane tasks. For example, you know,
even internally at Lumen we generate employment contracts for all
of our new employees. A lot of work goes into
kind of adding people's names and addresses and double checking
and making sure all the details are right. We think
these sort of tasks are perfect for AI because AI

(46:30):
is really great at you know, pulling information and falling
in fields and as long as it's constrained, it's not
off totally writing its own things, then it can be
tremendously helpful. But but as you said, Peter, there is
you know, there's a lot of risks there, and I'm
sure there's going to be a few cases where you know,
hallu hallucinations cause serious problems.

Speaker 2 (46:53):
So you're at one hundred million users. I think back
in twenty twenty one you had sixty millions, So the
growth is still really So what does this meant for
you sort of your capital journey? Have you basically funded
this from from from revenue? You said you put a
lot of the profits back into development. Have you had
to go out and raise money along the way.

Speaker 3 (47:14):
Yeah, so we're fully bootstrapped. We've got an amazing team
of people. You know, the team at Luhmann put on
a huge amount of work to build better products and
we've managed to kind of bootstrap it right through. We're
you know, we're obviously open to taking capital. I think
there's a market out there for capital. But for us,

(47:35):
you know, our goals are to build amazing products and
our goals are to think long term, and so it
doesn't always align with say the you know, the VC world,
where you're trying to build something really quickly, hype it
up and then move on to the next one. It's
a little bit different from what we try to do.
We try to build you know, amazing products that really
laughs and so, you know, looking at other ways to

(47:58):
fundraise business. In the future we'll be will be open
to it, but for now as a bootstrip business, we
kind of love what we're doing and we think we're
having a huge impact on our customers.

Speaker 2 (48:07):
Well, you've got the big three platforms, the biggest three
software platforms in the world, so you can't really go
wrong there. Just finally, Max, what would your advice be
to young software enthusiasts, budding entrepreneurs who are probably looking
at New Zealand's narrative around software, which has very much
been around business software as a service. Ven sequence, great

(48:32):
christ Church Company zero have done very well B to
B sort of software. This is sort of B to
B but it's really open to anyone anyone who wants
to edit documents. So in terms of not limiting your
ambition to that sort of market going, we're too small
to compete in that field. We need to go after
some tiny little niche like building information management. What is

(48:54):
your advice to those entrepreneurs to sort of look a
little bit more ambitiously go after that big market.

Speaker 3 (49:02):
Yeah, definitely, I think for new entrepreneurs in New Zealand.
There's some really really important things. So one, thinking global
from day one is super important. Get out there and
travel if you can, and see some of the world
and understand, you know, the needs that people have in
different markets. We try to get all of our employees

(49:25):
at Lehman, we try to get them traveling to various
different spots around the world, be it India, be it
the United States, or Southeast Asia, to really get an
understanding of kind of all of the different markets and
and the way people think about problems. So I think,
you know, taking that global perspective from day one is
super important. The other thing I would say is just

(49:47):
get out there and put something in the market, try
and talk to people because there's a lot of work
to be done. There's always kind of more software to
be built or companies to be built, this huge amount
work out there to be done. So you know, just
making that first leap of faith and taking something simple,
even a really really simple prototype, and putting that out

(50:10):
in the market is a great way to get get started.
As we sort of talked about at the start of
this podcast, Peter Luhman was probably the tenth product I
put out in the market. It was not not necessarily
that there was no one, you know, it wasn't a
sort of a first success straight away. But for each
of those products that I put out in the market,
I learned something else. So the you know, the app

(50:33):
I did before Luhmann was a file converter, and it
converted files from doc X to PDF and doc X
to other formats. And I realized that through that that
file converter, I realized that, you know, PDF is a
huge part of what people do. Over seventy percent of
the files on that platform had something to do with PDF.
So you know, why not niche down into that space.
So for a yeah, a young entrepreneur would be just

(50:55):
think global from day one and then put things out
in the market, put messages out out on blogs, and
just see what other people think.

Speaker 2 (51:02):
And so great advice, Max, I'm sure that will be
inspiring to a lot of entrepreneurs who are sort of
setting out on that same journey. How many staff you
got now, so we're.

Speaker 3 (51:11):
Just over one hundred. Now we've got teams, teams all
around the world and just over one.

Speaker 2 (51:16):
Hundred congratulations, one hundred million users. All the best for
hopefully the next fifty or or one hundred million users.
Thanks so much for being on the business of tech.

Speaker 3 (51:25):
Thanks Peter, it's been awesome.

Speaker 2 (51:33):
So Ben, what can we learn from that one hundred
million users? A lot of them probably free users. But
that premium model seems to really be working for Luhman.
It's a hundred person company, it's bootstrapped itself, it hasn't
had to take on venture capital and dilute the shareholding.
Pretty impressive model.

Speaker 1 (51:54):
Yeah, and I think not to dumplay all of that
hard work, but I think they got in there early,
and they did it well, and they got those integrations
with Google Docs and you know, back in I don't
know if it's still the case, but when you went
to open a PDF in Google Drive, it would say
do you want to open it with Lumen and you

(52:14):
could connect to Lumin and it would give you a
little more functionality there. And then building that into a
more rich, rich, featured product for document sharing was really
smart and so it was inevitable and I think that
they did it well and they got in there at
a good time and using that freemium model, which can

(52:37):
often be a double edged sword, they've really obviously been
really thoughtful about how they actually use that to entice
customers and upgrade people onto a paid platform rather than
just getting stuck in that free user only Hell.

Speaker 2 (52:55):
It's probably a bit trickier to get traction on something
like that now, just because there is such an app
economy and a plug in economy. So if you launched
into the app store now, I remember writing, you know,
sort of ten years ago about New Zealand companies that
were launching on the App Store and doing phenomenal business.
You don't really hear about that anymore because they're literally

(53:17):
millions of apps, so it's very hard to get recognized.
What they managed to do ten years ago was something
that was universally needed. We were all using pds, and
they're at a point where these cloud operators wanted to
build out their third party offerings to make it more
sticky to be on their cloud. Luhman's timing was absolutely

(53:39):
perfect there, but also their commitment to improving the product
based on feedback, and that you know incredibly in twenty nineteen,
basically throwing out the code base and saying we're going
to start again from scratch. Huge undertaking to do, but
really paid off for them.

Speaker 1 (53:55):
It makes complete sense and I do remember back in
the early days of Luhmen, I would find it about
clunky and it wouldn't quite work some of the time.
And the conversation has made me want to go back
and actually have a look and see what's new. So
excited to go and play with it. Have you had
a play with new Lumen?

Speaker 2 (54:13):
Not really, no, but the bit that really does attract
me to it. I probably use a PDF maybe five
times a month, so I do have a Lumin account
so I can sort of edit stuff. They're basically on
the on the free plan. But you know, increasingly I
am dealing with documents where you need to sign it.

(54:33):
You know, it's a legal document and it's pain in
the butt. And there are all these other solutions you
see docu sign or something, or you know, there seems
to be a lot of them. So one that goes
a bit more mainstream is simpler to use. So Lumen,
you know, pdf sign, there's the signature authentication system I

(54:55):
think has a lot of real potential to get away
from that real corporate heavy, probably quite expensive subscription for
a document signing platform. I think there's a lot of
potential there.

Speaker 1 (55:08):
Yeah, absolutely, and especially if they're able to kind of
you don't have to deal with all of the different
international standards. So if you're signing a contract with a
Singaporean company, you know you're going to meet the requirements
for Singapore and wherever else in the world. I think
that's really a useful feature that will just be innate.
Most people won't even notice it, but will be a

(55:28):
deciding factor for those who do.

Speaker 2 (55:30):
And the government in the last couple of years has
done a big push on you know, E receipts and
all that sort of thing, so they want us collaborating
and paying for things more with other countries. And when
you get over like five thousand dollars or something like that,
typically you need to sign a document for a big purchase.
So if you can integrate that into E receipting and billing,

(55:53):
that's a really lucrative thing.

Speaker 1 (55:56):
I think, yeah, absolutely, yeah, yeah, definitely the way of
future with this invoicing. Especially New Zealand's adopting it reasonably
strongly now and if it can start to roll out
better and across like Southeast Asia Europe. It's really strong
in Europe, so it does decrease the friction of international
trade as well, which I know is a grandiose thing

(56:18):
to say about a SaaS platform, but if they can
become a key piece of the puzzle for trade across borders,
like DocuSign has, but do it within within an ecosystem
of other potential useful productivity apps, then that's a really
strong value.

Speaker 2 (56:38):
Yeah. And just finally, you know, interesting their approach to
their development workforce. They've got teams literally all over the world,
a significant one I think in Vietnam based on an
existing employee who wanted to go back there. So that's
I think increasingly the model that we're seeing. Sure, they've

(56:59):
got great people in christ Church, there's a good talent
based here. Canterbury is churning out engineering students. But thinking
globally in terms of your workforce, Luman have definitely embraced it.
They've got one hundred people and they're all over the place.

Speaker 1 (57:14):
Yeah. Yeah, and gosh, so much I could say about
the benefit of being a remote work first company and
how that can ultimately help to to give you an
edge when it comes to expanding internationally, But then I
might have some you know, angry CEOs yelling at me, so.

Speaker 2 (57:34):
I won't properly anyway, Thanks, so much to Max Ferguson
for for coming on and sharing what is a true
Kiwi success story?

Speaker 3 (57:44):
Yeah?

Speaker 1 (57:44):
Absolutely thanks Max. Show notes and the tech reading lists
at Business Desk, dot co, dot and Z. Check them
out in the podcast section.

Speaker 2 (57:52):
And follow the Business off Tech on your podcast platform
of choice. We're also streaming on iHeartRadio.

Speaker 1 (57:58):
Get in touch with your feet, back and topic suggestions.
We're on LinkedIn and Blue Sky.

Speaker 2 (58:03):
And catch us again for the next episode of the
Business of Tech next Thursday.

Speaker 1 (58:08):
Until then, have a great week.

Speaker 2 (58:16):
M