Exploring Chess, Manipulation & Product-Led Growth with Claudio Descalescu

My guest today is Claudio Descalescu. He's the chief growth marketer for ArchBee. ArchBee is backed by Y Combinator. It's a tool that allows you to easily and quickly document product documents, APIs, and SDKs. They've got around 400 customers and are projected to reach about $2 million in USD in revenue for 2023. In this episode we discuss how Claudio uses chess, manipulation, product-led growth and surveys to improve products and gain insights into customer behavior.

Claudio has held business development roles and data analyst roles at UiPath and IBM.

Claudio, welcome to Innovators Can Laugh. Thank you. Great to meet you, Eric.

Yeah, so I want to jump into a couple of your hobbies real quick because I find them fascinating. First one is Jiu-Jitsu and the second one is chess.

So when did you get into these?

Were you kind of nudged by your parents to get into them or were you interested by yourself?

Interesting question. So probably for chess, somehow the parents kind of guided me, but not in the way you would expect. But Jiu-Jitsu, I've discovered it late.

I mean, I started doing it before everything closed down.

But yeah, it was later in my life and I realized it's a great way.

I mean, for Jiu-Jitsu, you either look at it as a physical exercise or as a spiritual thing just because it has so much history or just a way to fight. It has so many angles you can look at it. So it's great to do that.

But chess, actually, my father and I were playing chess at one point and he was just winning game after game. And I just decided like, okay, I need to do something about this. So I went to a school, a chess school by myself. And the teacher said like, yeah, I don't think you're too old for this. I was 10 years old or something. Usually you start chess a lot earlier.

Oh, this was a physical school.

This was an online?

Yeah.

No, no, no. I was 10 years old. I went to my local chess school. And the teacher said like, you're too old. We usually start with kids at five years old and we try to make them grandmasters and stuff like that.

I was like, yeah, but can I join?

Yeah, whatever. And he didn't pay attention to me. In one year, I was actually one of the better ones at the school just because I was probably I had the motivation. I wanted to go back and win some rounds with my father and other people that I was playing with. I was just brutally beaten.

I was playing with the person and they had like, I think 50 wins over me and I had zero. And I was like, this cannot continue.

The person was an adult or was it an older kid?

Yeah, an older kid actually. I was playing with my father and with this older kid. Those were my opponents. But after that, I went into chess professional. I took it for about five years or so, actually more than five, six or seven. But after that, I went to college.

What do you mean professional?

Like at a certain age, do you go and you say, I'm going to play in these professional tournaments and they have to look at your background and let you in and say, okay, now he's a professional?

I had a card that was saying I had pay intuition or something to the club. I went to two or three professional tournaments, chess tournaments.

I won one, but it was a small local one. The other ones, I didn't do that great.

Hey, does that card get you any discounts?

Can you go to a bar and tell the bartender, hey, you're looking at a professional chess player, seeing my card right here?

No. It doesn't do anything for you. I actually didn't show it to anybody. Probably that's why I don't know about the discounts. Okay.

Well, I would think maybe if you're like in the passport line trying to get into another country, there's this long line and you tell the security agent, hey, professional chess player right here, gray master.

Yeah, it didn't work for me. Okay. All right. Just thought I'd say. Okay. So I also saw something interesting on your blog here and you said that it's important to read at least one book about manipulation.

I wonder if you can expand on that and what book would you recommend for people to read and write when it comes to manipulation?

You did your research. So I would just recommend the book that I highlighted in the article. It's a Romanian book. Everything about manipulation, I think is the title. But the background story for that article is that I was reading that book when I was in college. So I had a roommate that didn't read anything.

I mean, he was that type of person.

Wasn't into reading, right?

So I finished the book, put it somewhere in the room. We were sharing a dorm room and he picked it up, started reading it. And while reading it, he was mind blown about what he was reading there, about the strategies and how things work. And he was like, wow, this changes my life.

And I was like, yeah, it does, right?

You should read more. And that's why I wanted to put this into writing, just to have something to go back to and remember the story. I found it really powerful when I saw it was the reaction. I wouldn't expect from that guy to...

What's this guy doing right now?

Is he like this master criminal or is he an honest person with integrity?

What's going on with him?

The latter, I think an honest person. I think he got married and he moved back to his hometown and he's being manipulated as we speak. Okay. I'll probably lose a lot of my listeners here if I say this. So let me try to put it in the most political friendly way as possible.

I look at my son and my daughter, right?

My daughter's three, my son is five and a half. And she's already starting to kind of like, not manipulate him, but I feel like for girls it just comes sort of natural and instinctively. And manipulation is not the right word, but to be able to get daddy to do what I want.

She does it in such a way where she hugs me and she'll kiss me and she'll ask, or my son, it's more about the straight up ask.

And then he throws a fit and he gets all angry about it, right?

So I feel like the women in general have just by nature, just by nature itself, have this leg up on us when it comes to, again, lack of a better word, manipulation. Okay.

What was the name of this book that you wrote about?

I guess, is there a Romanian translation for the name of the book?

Yeah. It's Totul Despre Manipulare Bogdan Ficeac, I think is the author.

But yeah, that's basically the book. It's like a small two hours reading. Okay. It's great. All right.

Well, I want to talk about another book that, and it's not necessarily a book, but I think it was a case study you read about product-led growth. And this book had sort of a really big impact on you, or at least on the way that you were trying to grow ArchBee and perhaps some other companies that you worked for.

And you were telling the story and many of us that were there found it really fascinating. I read the story, I shared it with a few of my colleagues that I work with and it seemed like, hey, this is sort of simple, but it's ingenious at the same time.

And what would be your recommendation for people who are looking to try to find who is their ideal ICP and how can they grow faster using this method?

So I guess first talk about what is it and then how would you utilize it?

Yeah. So I've read a lot of things. I think I have a unique position. I just realized this. I've basically joined a startup after it was recently founded and while going through Y Combinator as the first hire. And the first hire was a marketing hire. So kind of a unique position I found.

Which means you're doing like 50 things all at once, right?

At least. At least.

Yeah, let's be honest. Yeah. Yeah.

But this also allows me to experiment a lot, right?

So you just need to break down how can I build this growth engine?

What type of fuel can I put into this?

How can I make this work and continue to have the growth that we need to basically be a startup?

A startup means that you're growing at exponential rates more or less. So at one point I realized, okay, maybe it's just about the larger problem, not about the product, not about how we onboard people.

It's about what you mentioned, like what is our ICP?

How people talk about us because the wording that we're using on the page is not exactly what people describe about the product.

So I went, I've read so much, but the one that was the recent one and clicked most with me was an article about how superhuman tried or found their product market fit. As a startup, this is the first step. You need to have product market fit. And after that, you can add the marketing engine that channels on and so forth.

But first you need to have product market fit to understand your ICP, who is the larger audience. And Claudio, for those that don't know who or what superhuman is, I don't want them thinking that this is a superhero. So just kind of briefly explain what superhuman is.

It's basically an email app that offers some extra features that they're basically geared at least after they're not pivoted, improved towards busy people like CEOs, executives, so on and so forth to allow them to go over the email faster and improve that time you spend on email. So that's their main goal basically. So superhuman is basically an email client at the end of the day.

Yeah, if you're getting more than 75 emails a day, this might be something you probably want to look into and test out. Okay. So you read this case study by the founder of superhuman.

And what was he saying in this article?

So the question, he presented the whole methodology. So as any good marketer out there, I just stole the methodology and applied it in my case. And the main idea here is that in order for you to actually, let me go back a bit, to find product market fit, it's not a black or white kind of situation as I found it. It's more of a spectrum.

You'll have KPIs or metrics that you would look into and say like, okay, we have product market fit. But at the same time, you don't really feel it.

I mean, you see like, okay, we're targeting, we're building what we think will solve a problem, but it's not really that thing. So it's always a spectrum. So what they did, they've sent a survey that had a couple of questions, but the most important question was about how the user will feel if they wouldn't use superhuman anymore.

So the questions sound something like, how would you feel if you wouldn't use ArchBee anymore?

Somewhere along those lines. And you have three questions. I love that question because you get the true fans of your product. Yeah.

And you put the person in the context of thinking, hey, okay, do I want to use this app?

Should I use this app?

I mean, it changes the mindset when you address this question.

And I've heard this, after I've sent the survey and went on some calls with some customers, some of them were like, what was that email about?

Are you closing down the business?

I mean, it really puts them in an interesting, at least, state of mind.

Like, what's that about?

Are you... Yeah. Yeah.

You know?

That's what I'm trying to say. So you have these three answers. Very disappointed, somehow disappointed, not at all.

And after that, you have some follow-up questions, depends on anybody else that wants to run the survey, something like, what is the main benefit of ArchBee that you would find?

Secondly, what type of people do you think would benefit from ArchBee?

And usually, this is an open question. They would usually identify with the job description, job function that they add here. It's an open text question, but they usually would write something that's related to them. So instead of asking them, what's your job title, which you should already know by then, you ask them what is the larger audience that you think would benefit.

And finally, something like, what do you think we could improve on the product?

There are multiple questions that you can ask, but these are the four ones that I've sent.

So after that, like any other survey, it's great, we have data, what do you do with it?

So that's where I think this framework really helped at least part of my activities as a marketer, because I took the answers. And of course, you'll have random answers there, but somehow a benchmark would be like, if you have 40% of the users would actually be very disappointed, you've kind of hit product market fit, because that's almost half of the user base.

So half of your user base is actually very disappointed if they wouldn't use ArchBee. It means that you're basically doing something good and the people that are using the product are the ones that you actually want. But what happens in a startup is that you don't have 40%.

Yeah, I was going to say that, realistically, you're probably what, 10 to 20% maybe?

Yeah, around 20%, at least that's based on all the users that replied to the survey. In my case, it was around 20 something percent, which is more or less similar to what it was in that article with Superhuman.

And I was like, great, what do we do from here?

And in the article, they explain like, okay, but you need to segment.

And this is like a marketing thing, right?

You always need to segment. So what I did was, okay, we kind of know what is our ICP. We did some research in the past about this. We know that product managers, technical writers, maybe developers will exclude founders because founders need to do everything. So that's not really our ACP ICP. So let's look at these job titles.

And when I filtered just by those job titles, the very disappointed percent actually went up. So we were around 40%. So it means that our assumption about the ICP might be correct. So maybe we should target product managers, technical writers, developer relationship specialists, mainly the person, a person that's hired to connect with developers when you're building a developer tool.

Right?

So that's more a niche thing, but they're into documentation a lot.

So okay, that's great. But what I did next, I mean, again, product market fit, it's a spectrum. You can look at the data, say you have it, you don't have it. But that at least the way I've used the data, I would say, okay, if these are with these job titles, actually kind of reach that 40% mark benchmark. Let's look at what the benefits are for them and what they want to improve.

What they want to improve, I'll just send to the product and guys, we need to work on this. Consider this. Let's see what we can do. It seems like that. Yeah.

I mean, because instead of having this laundry list of improvements that you could make, right?

And even requests by other customers, you're just really sort of segmenting the improvements that really matter to your OACP there. So that's fantastic. So now you're sharing this data with your team.

What's been the result?

Or when did you do this?

How long ago with ArchBeef?

Two months ago, I think. Okay. It's actually really, really recent. And I think actually the survey is ongoing now.

I mean, we got so much value that we're sending it two or three months after somebody signs up with the product. So we give them two months, two or three months to get into the product after they purchased a subscription. And after that, we send them the survey ongoing.

Initially, I thought it was a one-time thing, but it's too much value to not send it. Yeah. Yeah.

How has this changed what you're doing from a marketing standpoint, Claudio?

Yeah. So I've looked at the benefits.

So I had four questions, right?

One was the segmenting question.

How would you feel if you wouldn't use the platform?

The one regarding how we can improve is for product, but the other two benefits and what other people can actually benefit from this. I used that information to build a new landing page, which was addressing more the audience or the audience, let's say, not necessarily the ICP, that actually resonated and was very disappointing. They didn't use RTP.

So I just filtered by that, built some mental, not mental, codified, because it's open text, it means I got a lot of random things, but I read every each one of them. I codified them. Okay. So it means that for these users, publishing is the most important feature. For these ones, the writing experience is the most important feature.

So I've basically did a count on how many times on this codified column that I've built I had, and I've just translated that into three new sections on the homepage, on the page that was addressing exactly the main things that they were seeing. And what I did was not to write about those things myself, I just copy pasted some of the things that they said and added and expanded. I love it.

Yeah.

No, I love that. I always feel like, hey, the best copywriters are those that just look at the comments and the reviews by their customers and use, not necessarily verbatim, but sometimes you can use it verbatim, word for word, what they're saying. Okay.

So is this the homepage that you're describing?

Because when I went to your homepage, it did look like a lot of the copy and text on there, it was like in this friendly handwritten font, but it looked like those were things that customers had said about the product. Is that- It's another homepage. I'm running some ads.

So the way we actually iterate on messaging and what we add to the homepage is I'm running ads to the audiences that we think will resonate with this and I'm sending them to this landing page. I can share it with you and it's a lot different than the homepage and see what happens from there.

Are we getting more engagement?

Are we getting more conversions from that compared with the previous version?

So before we update the homepage, we're basically iterating by, and by the way, I've also used the same wording, not the same, similar wording in the ads, definitely much more engagement compared with the ones that I was reading, I wrote. Absolutely. Yeah. It's definitely more engagement. It resonates.

And I want to mention this, we were actually reviewing some of the inter calls that we had with clients and just to pick the keywords that they were using, they use, like the keywords are so different, documentation CMS, knowledge hub.

I mean, there's so many ways they address this because each company has a different knowledge management problem. Knowledge hub will resonate so well with users in the US because that is a common term that was commonly used as like a central platform for document sharing and knowledge sharing.

So yeah, knowledge hub would definitely resonate for people in the US.

Yeah, absolutely.

But I love that you mentioned this because I felt at one point that there's a gap, you know?

We can talk in English, I have a good understanding of the language, but the nuances are pretty different. And the way we want to describe the platform doesn't necessarily... Yeah. And I never do this, Claudio, I never do this. What I mean never do is I never plug in the company I'm working for or client or whatever.

Right?

But one of the things that you can do with personalization is that you can have different messaging based on where the user is coming from based on IP. So if you have like different messaging that resonates with US visitors, then you can simply have that headline appear.

What did you want it to say for visitors from the US versus maybe from the UK versus other countries in Europe?

And that's something that OptiMonk, that you can do with the platform and OptiMonk because it's a personalization platform. So a couple of other questions.

So before and after, before you started running these experiments, what sort of lift or difference are you seeing from the new landing page in terms of conversion rates, like signup rates?

That's exactly it. It's the percentage of signups and it's complicated because you see the conversion rate improving. We have three and a half percent now from two and a half or something. So there's a big improvement, but the number of signups are not necessarily growing.

So when you show this to somebody, it's like, yeah, but did we actually improve?

Well, the idea is that not all the time you can get more signups, but you can get better signups at the end of the day. So because this will, you'll see this in a month or two because they will get better qualified, pre-qualified, I would rather say, will get into the product and you'll see the results later than now.

But now what I can show you is that we just improved the conversion rate, not necessarily the number of, because at the end of the day, how many product managers and technical writers are in the world?

There are so many.

How many can you convert to?

Yeah. Yeah. What's going to matter is the retention aspect. How many of those people are actually staying on long and actually are addressing that question as, yes, I would be greatly disappointed if I can no longer use ArchBee. Okay. A few more questions.

So what do customers really love about ArchBee?

I know there was a few different things that they like, but what is the number one value that they get from it?

I spent months trying to answer this question. But the main thing about ArchBee is the editor. It's really modern compared with the other documentation tools you have there or collaborative tools like Notion, Confluence, so on and so forth. But even recently I've talked with a client and it's more about the experience, which is like, so how can you translate that into, yeah, we have a better experience or something.

But actually this is a really recurring thing, like it's a combination of how easy it is to write, how simple it is to use if you don't want to use advanced features. But we do have advanced features, but that's so well built into the product. Only the ones that need them will get there. So I'll say the editor, but it's the experience.

I mean, anybody from technical and non-technical can... I have to agree with you on that. When I first saw it, I saw a video of it and my first thought was, oh, anybody who loves or likes Notion is going to love this. That was the first thing that was in my head. I use Notion. I think it's pretty cool, but I don't love it.

But there's a lot of people who love it. And I thought when I saw your product, I'm like, oh man, anybody who loves Notion is going to love this. Yeah. Yeah.

Very, very cool. Okay.

So, if I were to give you $10,000 to help grow ArchBee in some way, where would you spend that money, Claudio?

I'm happy to have a good budget for the marketing activities. But if I had an extra $10,000, I would use them for an event. Of course we have some spent for the classical channels, paid channels. But if I had an extra, I will just use that money to go to an event because I know my ICP. I know exactly what type of events I would look for.

And I wouldn't use the money to get the booth or something like that. But you'll have the expenses to get there and join the conference.

But maybe I'll set up something for, I'll reach out to some people and maybe prepare, invite them to a nice dinner and out of that $10,000, I'll just pay for the dinner and get some people to really connect, not necessarily have a booth and hope that somebody will stop by and stuff like that. But that's what I would do with $10,000.

And that money will allow me basically to go to any, no matter how large the event is, you can do this with $10,000. You have some expenses with the travel and paying for a dinner, shouldn't get that. Okay. Okay. Cool. Other question for you.

What is a lesser known tool or app that is bringing you a lot of value?

This could be a Chrome extension or it could just be like a tool you use all the time.

Yeah, that's a great question. But the real lesser known are usually like some Google Sheets type of thing that automatically does something for you. There's a import from web Google Sheet add-on that I've recently used. And it was really cool because I was trying to improve the copy for some ads for Google search, basically. And I just wanted to scrape the search results for a specific keyword.

And this add-on allowed me to add a keyword and they scraped the title and the description. And what I did, I just took those titles, mixed them up real good and generate my own copy basically, based on the existing search results.

And the reason I did that is if Google ranks those pages, it means that the keyword that I want to pay for to get traffic will match the intent, if that makes sense. So if you're going to use the same wording and stuff like that, you're going to get a better keyword to add copy, match. Okay. So the name of this extension is... Import from web. It's just like that.

Import from web, it's a Google Sheets add-on. I can share it after. We'll include it in the call notes. So I guess if you're doing some research and maybe you go to a website that has, maybe even a competitor, you can use this tool and you put in a keyword and it will show you import all of the copy related to that keyword or surrounding that keyword.

They have multiple use cases, but the one that I've used is just that, just scraping Google search results for keyword. But I saw they have other use cases, but I was trying to build this myself and I saw like, oh, this is an interesting tool. All right. I think I need to try that out. All right. Last question for you, Claudio.

Something weird that you obsess about, something weird that you obsess about. I think the best person to answer this question is probably my wife, but... Mine too. Yeah.

Probably, I like my obsessions, so I don't have the power to recognize that I obsess over something. But I think I have my setup, so I can work from anywhere like many others can in this industry. But when I'm home, I want to have my setup. I have two external monitors. I have my keyboard. I have my pad. It's my setup. I'm kind of obsessing. I actually, I'm sharing this with...

I just said like, this is my desk. You're not going to touch it.

Yeah, but I need to clean it.

No, I'm going to clean it. This is mine.

Yeah, I think that's what I obsess about, my setup. Okay.

Very, very cool.

Hey, Claudio, thank you for being on Innovators Collab. I will include links to this interview in the show notes, including links to ArchBee. You can also catch this interview on YouTube. Just look up Innovators Collab on YouTube. And if you like it, tell others about it. That's how we go. Thanks for everybody listening until next week. Thanks for listening