Yash From Momentum (00:00)
Hello and welcome back to Building Momentum, the show where we peel back the curtain on the exciting and often chaotic world of building a successful SaaS business. I'm Yash, your host for the show where every episode we bring you the stories and strategies of founders who've been in the trenches, conquering churns, scaling their teams, and building products that people and businesses love. In this episode, we'll be chatting with Christophe from Doc Analyzer. Doc Analyzer is a next -gen AI -based reliable agentic tool
that offers workflow automation, intelligent document conversations, and OCR, among other features. We're excited to hear their story and the lessons they've learned along the way. We'll be dissecting the wins, the losses, and everything in between. So buckle up, grab your headphones, and get ready to dive in the world of SaaS. Hey, Christophe. Thank you for doing this. How are you doing today?
Christophe (00:51)
I'm good. Thank you for having me, Yash
Yash From Momentum (00:54)
Absolutely. Awesome. So I wanted to start just to try and understand what's the big, bad, hairy problem, challenge that Doc Analyzer is trying to solve.
Christophe (01:07)
Yeah, sure. So everyone of us, of course, have been very excited when ChatGPT has been launched. We are all trying to chat with the tool on doing some extraordinary stuff. And past the first dialogue on week of amazement of how much this new kind of AI has reached, the big question was, what do do with that? There are, ofcourse, generative AI. There are some evident
Yash From Momentum (01:13)
Yeah.
Yeah.
Yeah.
Christophe (01:36)
usage that we can think of on the first one where to have a summary, to generate the text for SEO, to generate easy imaging for marketing and so on. But past that kind of a very simple usage, are just a big question how we can do better than just generating image or some small piece of text. So this
Yash From Momentum (01:48)
Yeah.
Christophe (02:02)
or what the analyzer is about is to explore specifically in the domain of documentations of documents when you have huge amounts of documents, what you can do with AI, generative AI that is a bit more complex than just having a summary or having a conversation, a chat with the document. So we are trying to create agents that can replace you in your work
if you work in the domain of document. So there are many people who are working with documents. have accountants, they are working with invoices, you have architects, they are working with the construction permit, you have lawyers who are working with contracts. And all of them, have already their kind of workflow. They do nearly the same kind of things every day. So the goal here is to try to understand for each of these power user of documents,
processes that could be simplified and be done by AI agents instead of them. If we succeed in that endeavor, of course, we are going to save a lot of time for all these people. So that's free the time to do something else.
Yash From Momentum (03:18)
So most SaaS products fall in either of these three categories, whether they are either they are saving time, saving money, or helping the customer make more money. So if you had to put Doc Analyzer into one of those boxes, would you categorize that as we're in the business of saving time of our customers? Got it.
Christophe (03:38)
Totally, we totally on that category without any
Yash From Momentum (03:43)
Understood. tell us a little about the journey of building this, right? How did you start? How long has it been? What's the size and scale of operations where we are today? What's the status quo like?
The idea behind this is, just so that you have a little more context, is the idea behind understanding this about Doc Analyzer is for the audience, for the people who are watching or listening to this. They will be able to put everything that you're about to say in that context. This is where the product or the platform is as on net.
Christophe (04:20)
So you're asking about how I get to this idea to build the Doc Analyzer ? I get it right? OK.
Yash From Momentum (04:25)
Yeah. Yeah. Yeah. So how did you decide to build it and then where is it today?
Christophe (04:32)
Yeah, so basically I'm an IT guy. I've been doing IT all my life. I've been creating a few companies already all my life in the mini sector of IT, open source, more recently in blockchain. I've been CTO also in many companies which I've not created. So I get this kind of depth about IT.
generative AI came out, I was totally trying to find the right sector of that particular new trend in IT. I should pick up to try to bring my own solution. So I choose document because it's kind of the both easy way to start with it because, mean, of course, the simple application is just what we call
a rag in the AI talk, it's just you take a document and you take for example, a GPT of OpenAI and you just put the context of the document and you just chat with it, you give the context to the AI and you can have conversations. But then when you start to go a bit further, of course, there lots of complications into that. So this is where my experience of having created companies before and having
SaaS before is, of course, very useful. And then it came up to the simple recipe of finding a good niche to start with on something not too wide. Otherwise, you need to solve a problem for some people. So you need to define who is your customer you are going to help. And really try to focus on that problem. So even if we didn't get that
Yash From Momentum (06:02)
Mm
Christophe (06:30)
clear visions one year ago when we started, the more we talk to our customer, we took a lot of customers, it's very important. If there is something that you should not do, is not to talk with your customer. So I'm still answering most of the support questions myself, even if I get some help with some stuff which are more evident sometimes, because this is where you learn the most actually. You listen to them and you listen to their problem and say, yeah, sure, I can help.
Yash From Momentum (06:55)
Yeah.
Christophe (06:59)
this way. And then it gives you, of course, you are going to help one customer this time this week, but it also gives you an idea how you are going to change your business and put new features online for the month to come. That's very
Yash From Momentum (07:01)
Yeah.
Got it. And where are we today, right, in terms of the size and scale of Doc Analyzer? How big is the team? How many customers do we have? How many docs have been analyzed?
Christophe (07:25)
Sure, sure. So we are still kind of a small startup. We started last year, built from scratch in the spring. We launched in July. There was a slow start, which is totally normal. People should not, by the way, it's something important maybe for your listener, you should not be discouraged because at the first weeks of the first month when you start to put the product online,
there are very, very few customers. To be honest, in doing the last summer, there was nearly no one using our tool because no one knew we exist as simply as that. But then when you start to get people on the world of math start to propagate, you can pick up quite fast, actually. So at this stage, we are a small team of five people.
Yash From Momentum (08:01)
Yeah.
Christophe (08:21)
two developers, one marketing, one designer, and myself. And we gain about one customer a month, which is still a modest amount of customers, but it's interesting to accumulate them. And then your game is going to make them satisfied. Working hard on the churn rate, so every person that gives
Yash From Momentum (08:47)
Yeah.
Christophe (08:50)
that says to you, OK, I want to unsubscribe, to cancel my subscription, you should go after them. Not so much to save them, because most probably they already got the idea. You're not going to save them. But you need to understand the reason why they are leaving. Because then this is where you can fix your tool, not for them, but for the next customer that there is less reason to leave. Or you will always have some people who leave because they could have money crush, or they could change their job.
Yash From Momentum (08:57)
Yeah.
Yeah.
Christophe (09:18)
Maybe the need for the tool doesn't exist anymore. it's totally perfectly normal to have some churn rate. But this is something which is very important to
Yash From Momentum (09:25)
Mm
No, fair point. Because as much as retention is important for a SaaS product, it's also important to understand why is churn happening. So why are people, in whatever percentage, is it just revenue churn or is it customer churn as well? is the usage decreasing or is the number of users also decreasing? How many people are there who have active subscriptions but are not active users?
that they will churn in the next few weeks or months or quarters, it says that they have not looked at their trade cut statements yet or something like that. So it's extremely important to start measuring that. But one of things that I want to touch upon, something that you mentioned early on, that, and this is a challenge that happened with me as well, and given that you build multiple businesses,
happen with you with Doc Analyzer is that when you're building something, you tend to think that you're building, you're working on something that's extremely important and you're getting attention from, you team members or the people that you're speaking to, right? So people in your family, everyone's asking you, what are you working on? What are you building? And you keep sharing about whatever it is that you're working on and you're getting some amount of attention. As soon as the product launches.
as soon as the website's up, as soon as the beta is up, you start to realize that you don't have it. Just because you've built something, it doesn't mean that people will start coming in or walking in. so the amount of effort that needs to go on distribution channels, the amount of effort that needs to go into engaging with prospects, whatever the go -to -market channels are, is almost as important as building a solution. And so that also brings me to my next question, which is, how do
How do you at Doc Analyzer think of your go -to market as a strategy? So how do you define your customer? And then how do you go after them? So what are the channels that you use? What are the experiments that you created that worked or that didn't work?
about that as a journey? When we are acquiring, I think you mentioned one customer a month. So what are the channels that you're trying to use?
Christophe (11:39)
100 customer, OK. So our typical user would be a power user, someone who is a bit in all this category we spoke about, accountant, lawyer, and so on. You always have in these categories
a smaller category which is very familiar with IT, which is kind of always trying the new tools, the new way to handle their specific early adopters. So clearly, that would be this power user, early adopter that clearly our audience right now because we are still a new things. By the way, this relates to the channel, right? Because I noticed
Yash From Momentum (12:15)
like the early adopters.
Hmm.
Christophe (12:32)
For people who get disappointed by the tool, most of the time is related to the fact that it's complex to understand what AI can bring to you. And sometimes it's complex, even if you kind of chat with it, and it's much more natural than older interface, because it's still a natural language. You're addressing the computer with a natural language. There are ways to talk to with
LLM which are not exactly the same like I would talk to you for example, because they have a very short context window and they have some quirks. Some people get fast into this way of making the most of addressing with LLM. Some people just don't understand so well. They could be frustrated very fast and they can start to argue with the LLM, which of course doesn't bring any solution to their problem. And so
Yash From Momentum (13:04)
Yeah.
Christophe (13:29)
So the frustration can mount very fast with this kind of interface. So the way we do acquisition at this time is two ways. The first way is SEO, because this is something which is very important for us. And the second way, and we are just starting with that, is going to use influencer marketing on the training, making some video training.
which are permitted by influencer in the subgroup that we defined that could potentially be our customer. We don't do paid marketing at this stage because it's just too expensive in my opinion. But that's the stage for the growth we are in now, which is still early.
Yash From Momentum (14:14)
Yeah.
Christophe (14:24)
We, of course, will change that strategy later on.
Yash From Momentum (14:29)
Yeah. Got it. And so I also saw on your website, and one of the things that you mentioned earlier in the conversation was also the fact that anyone who works very heavily with documents are one of the prospects or possible customers. So people who are in HR, people who are in admin, people who are lawyers, accountants, and so on and so forth.
Within that, so as a founder, if I'm building a product that appeals to, let's say, two or three different definitions of ideal clients, what are your thoughts or how are you going about it at Doc Analyzer? Are you saying that we analyze documents for everyone who needs it? Or are you one of those where you say that, we'll do, like, we'll go niche down on a particular...
ICP segment like acquire a very large chunk of a very small market and do that extremely well from a product standpoint or Are you hedging your bets in the sense that okay? I may have smaller parts of but three or four different markets so that you are not you're not sort
risking yourself, this is a particular market scenario changing and you're sort of a platform that appeals to four or five different definitions of ICPs. What do you sort of go after?
Christophe (15:51)
Yeah, we define our server especially for documents. Indeed, it could apply to many kind of businesses. But we focus though
on small and medium businesses. That's very important. And also, geographically speaking, we mostly target the US customer base. But even though we also have customers all over the place, people who speaking this, because we don't put so much effort at this stage to localize our interface, for example. This is not something we are interested at the moment. We prefer to go to features and stuff.
Yash From Momentum (16:16)
Yep.
Yeah.
Yep.
Christophe (16:33)
when we get to the point where our feature is more stable, maybe that would make sense to do some localizations, but not at this stage. So that is two strategy we have. Specialist of document for the English -speaking
Yash From Momentum (16:49)
All right. And so another thing that I want to understand from a product management standpoint, right? So pricing a little bit as well. So there are a couple of ways that you can have like a cost -based pricing or a value -based pricing or just competitor benchmarking and stuff like that. Given that...
AI and everything that AI enables is so new and so early that almost everybody is working on categories that didn't exist, right? Because they are all categories that existed, but they're making that significantly more efficient and effective, right? So in those cases, like, and I also saw that you have a very generous free plan. How have you thought about pricing for knock analyzer?
What are the kind of experiments that you are running or is it, let's see what happens. Can you talk a little about that?
Christophe (17:36)
Sure, Sure, of course. So we start even with a cheaper price than we have now, because we start with four. When we launched, the basic plan was only $4. The problem was $8, so very, very cheap indeed. And we had a generous 3 .0C. Also, now we raise that bit, because it's now $6 on the $12
CTE for basic on ProPlan. But there's also some package you can buy during the subscription if you are using expensive models or if you are using expensive features like OCR and so on because it's just impossible to have a sustainable business if we pay a lot on this little subscription coming in. So I'm trying to be as cheap as we can because I don't want to restrict too
Yash From Momentum (18:30)
Yeah.
Christophe (18:36)
the small independent crowd for which the pricing is very important. And we don't yet target so much enterprise plans. We will open that in probably this autumn because usually it's better to have this independent small businesses. We don't think too much when the plan is cheap to buy because enterprise, have
a sale process will be different. Of course, they can pay much more money per month, but it's also very low. The acquisition time is much longer. It's not eight seconds. It can be several months. So that's like complexified a lot of things. Also, you need to have sometimes certifications. They want this stamp about security. This is costly. So it's better. My advice
Yash From Momentum (19:17)
Yeah.
Yeah, yeah.
Christophe (19:34)
for founders to wait a bit before going chasing enterprise because you risk spending a lot of time on the calls, many, many, many calls. And at the end, they will refuse you because you are just too small yet. So you need to grow a bit to look a bit more like a tiger than a cat to
Yash From Momentum (19:50)
Yeah, I've been there and it's just extremely difficult because just filling out like vendor onboarding or filling out their RFP or RFQs, that in and of itself is a two or a three day activity for founders because there are questions that you have not thought about, about your own product, right? And
Christophe (20:12)
Yeah, yeah,
Yash From Momentum (20:15)
There are things that you also don't understand. you essentially have to start and go on a journey of understanding what are they talking about? What are they asking for? And then sort of, and then it's like, even the journey of decision -making is extremely long. Like there are multiple people, there are decision -makers and champions and influencers and everything, right? So it's very hard.
Christophe (20:23)
Mm -hmm, mm -hmm.
It's hard to get there. It's hard to get there. I've been living that also in other startup, which are bigger, and I saw the process. So I know it's bit too soon. And I received a very great email from big organizations, say, we love the Canalizr. We want to use it for 200 users, which is great. But the problem is that you can still answer the call to have some preview of what they could need.
Yash From Momentum (20:41)
Yeah. Yeah.
Christophe (21:03)
But even if the person, the champion that loves you, loves the product, there's still a lot of green light that needs to pass on green later on the process, the chain of command before it works. So that's complex indeed. So yeah, we focus on this small business enterprise for now on making them happy. I mean, this is the key point of every business.
Yash From Momentum (21:11)
Yeah. Yeah. Yeah. Yeah.
Christophe (21:32)
Make at maximum the maximum customer happy. There will be some unhappy, of course, but this is the goal.
Yash From Momentum (21:36)
Yeah. Yeah, no, absolutely. That sort of puts it very simply. just as many people as possible who are using your product if they are happy, there's nothing more that they expect from customer support, from product, from usage, from whatever urgent, important, frequent use case that they have. So.
Interesting. So, so Christophe, this brings us towards the end of our conversation. And so one of the things that we typically end our conversations with is I ask you a question that was asked by a previous founder, which is something that they are currently trying to solve for at their organization. And then we'll hear your question as well. But so this question is from Zach, who runs a platform called It's Not.
Christophe (22:15)
Mm -hmm.
Yash From Momentum (22:28)
And Dart is basically an AI enabled project management platform where they are building agents who are like people who can also accomplish certain low -level tasks. And over a period of time, those agents will get better as well. But think of it as a project management system where you have people and you also have agents, and you're assigning tasks to people as well as agents. And that's how you're working on your day -to -day.
One of the questions that he has that he's trying to currently fix is given that they're trying to build something that doesn't exist in its final form or in its mature form yet, it's very early stage, how, in a very small team, right? So there are a team of 10 people. How do they make a decision on what to work on? So do they work on things that their customers are asking today? Or should they work on things that they want to build, which is where they want to take, dart?
which is a direction that they want to take it out in. So in the fight, and I'm sure this must be happening with you as well at Doc Analyzer, is in the fight between building faster horses or building a car, what do you sort of go after? Because a customer is not going to be able to imagine the things that are possible. So what are your thoughts on
Christophe (23:49)
Yeah, of course, yes. It's very tricky question actually on every team or every SaaS has same problem indeed. My take on that is I try to do both.
meaning that I take all this input from customer as potentially feature they want. And kind of take it half. So half I just put aside because for several reasons. They could be too complex to implement for the potential number of users that want them. So you have to sacrifice sometimes. mean, you cannot make everyone happy at the same time, maybe later, but not at the same
Some are just bad ideas in my opinion because they are making the product illogic. Because there needs to be some clearance. If you only take user input, because they are different users, you may end up with an animal with the head of a lion on the back of a rhinoceros, on the legs of a tortoise. So it kind of maybe not totally make sense.
Yash From Momentum (24:55)
Yeah. Yeah. Yeah. wow.
Christophe (25:02)
So I just select half of the feature that user want that could work in my strategy basically. Because I have my strategy, but so I will add on my own strategy with the feature that user wants. So I still want to go nearly the same visions I want, this notion of AI agents that are super smart about documents. But I would make
some detour to make happy in the meantime to customers. So there are some satisfactions going, ripping down the software.
Yash From Momentum (25:43)
It's so funny the image of a lion's head on a rhino's body with legs of a turtle. And so what's the question, Christophe, that you have? What's something that you are currently trying to solve for at Dock Analyzer that we should ask our next guest?
Christophe (26:01)
Yeah, so my question would be, in what time do we have, simply? Because we are still in the old way of doing SaaS, of course, this new AI, knowing that we are all waiting for AGI that may obsolete us all. So the simple question, like, what's your take on the date in every classical
Yash From Momentum (26:21)
Yeah.
Christophe (26:28)
recipe to build a startup and
Yash From Momentum (26:30)
Got it. Got it. That's an interesting one. I would love to see my next guest struggle to try and answer that question. But thank you. Yeah, no, absolutely. And thank you for this question. And thank you for joining in and answering all of my questions. And thank you to everyone, all the people who joined in, whether you were watching or listening on YouTube or Amazon Music
Christophe (26:39)
Yeah, I will watch to see that also.
Yash From Momentum (27:00)
or Spotify, wherever you are, you'll be able to find the link to Doc Analyzer in the description. If you work with a lot of documents and if you want to save time, you want to automate certain workflows, you want to have conversations with your documents, I strongly recommend you go ahead and check out DocAnalyzer .ai. Do give it a try and thank you everyone for joining in. We'll see you until next time.
Christophe (27:23)
Thank you.
.png)
.webp)
