Episodes

Andrew Fogg

Andrew Fogg

Andrew Fogg is the founder of Import, a tool that every growth hacker should have in their arsenal. In this episode, we dive into data scraping and how to use that information as a catalyst for growth

TOPIC ANDREW COVERS

  • Import IO is a technology platform that allows users to turn any website into rows and columns of data without writing any code
  • Users navigate to a website and show Import IO a couple of example pages, pointing and clicking on the data they want
  • Import IO learns from these examples to create an API that works for all web pages on that site
  • Users can then use the API to get data from the site, populate a spreadsheet or database, drive live web data into an application, or use it for other purposes
  • The technology behind Import IO is constantly evolving, and now there is a version where users don’t even have to visit the website or point and click to get an API
  • The topic is a new feature called “Magic” that automates extracting data from websites
  • The goal is for every website to have an API and for the data to be returned in a clean format
  • The motivation for creating this feature is to solve the problem of websites that do not have APIs
  • The feature is intended to benefit businesses by providing a way to access data from websites in an automated manner
  • And a whole lot more

LINKS & RESOURCES

WATCH THE INTERVIEW

READ THE TRANSCRIPTION

Bronson: Welcome to another episode of Growth Hacker TV. I’m Bronson Taylor and today I have Andrew Fogg with us. Andrew, thanks so much for coming on the program.

Andrew: Hi there. It’s a.

Bronson: Pleasure. Yeah, absolutely. We’re glad to have you here because you’re the founder of Import IO. That’s Dub Dub Dub Dot import dot IO. And in my opinion, it’s a tool that every growth hacker should have in their arsenal. So I think people are going to want to pay attention to how to use this tool and the things we talk about, because I think it’s going to be a big bonus for them. But first, tell us about import IO. First, how do you see it? Do you say import IO or just import? How do you talk about it?

Andrew: We say we say with you say imports. Okay. Very occasionally. Some people say import here. But that’s not that’s not what we say. And I think you’re going to we’re going to lose the IO at some point saying, I’m sure. And we often internally just talk about it as imports.

Bronson: All right. So import or import IO, those are kind of two imports.

Andrew: Are imports.

Bronson: I like it. All right. So tell us, what is import?

Andrew: I say it. Say import is that we’re a technology platform that allows you to turn any websites into rows and columns of data without writing any code. Okay. So you basically so any websites, any websites, data? No case. And basically, you as a user, you navigate to websites and web browser, you show imports, a couple of example pages, a couple of example web pages from the sites you point and click in the browser showing where the data is in the page. We import then learn from these examples to create an API on a platform that will work for all the web pages on that site based on those couple of examples. And then you can use that API to get data from the sites, populates a spreadsheet or a database drive live drive web data into a into an application or like an iPhone app or a business process. And so, so that’s kind of how it works. And we’re constantly evolving the learning technology behind imports. And we’ve got we have a version now whereby you don’t even have to do any kind of point and click training. You don’t even have to visit the websites or you don’t even have to visit the website in order to get an API for it. You just put a URL in on the home page of all of our website and it gives you an API back.

Bronson: So it kind of automatically just finds out where the data is and how it’s structured and just spits back to you.

Andrew: Exactly right. We call this new feature Magic, and that’s how we markets it automagically. Nice word for it. And it’s where a lot of our development is going as basically automating all of this. So that I mean, the vision for what we’re trying to do is that every website will have an API and will require no training. So effectively you’ll just be able to throw a URL for web for for a web page that’s imported and you’ll get the data back that you want. Yeah, exactly. Categorized nice and clean. Ready to be, ready to process. No big physical long journey. But that’s why we can know.

Bronson: It’s an awesome vision. I mean, that’s a cool place to be going because right now you mentioned the word API a few times. And just to be clear, a lot of times, you know, websites, they will create their own API, some of them, not all of them, which is great when you can tap into it. But what you’re essentially doing is kind of a back door where now you can have an API, you can have this application programing interface to tap into the data of a website, whether they created it or whether they want you to have it or not. Is that correct?

Andrew: Yeah. So we don’t I mean, so back backdoor, sometimes it sounds a bit kind of black cat security stuff. We think of it more about like being inside and outside. So, yeah, we’re so you’re quite right that we’re solving a problem whereby if you if there’s a website that doesn’t have an API, most websites do not have APIs. If there’s a website that doesn’t have an API and you want the data that’s fit or you want to interact with that website programmatically, then then you kind of you kind of have to either petition them or wait for them and very often you will never get one. Yeah. So what’s important to do is to kind of build that API without. Without having to wait for anyone.

Bronson: Yeah. And now this all sounds very techie and geeky, but here in a minute, we’re going to show people why this matters to just business in general. So we’ll get there in a second. It’s not just an experiment in programing. This leads to some bottom line growth kind of stuff for sure. But let me ask you this first. Why did you start it? Were you in a situation where you needed a product like this or you just kind of saw the need for it? What was it that sparked this in you?

Andrew: The kind of the history of imports sort of goes back a couple of years. And I was working at a bank with my co-founder. This is back in 2008. It was Royal Bank of Scotland which which at the time was the world’s second largest bank. It’s significantly smaller now, and it’s no longer the world’s second largest bank. And we were basically we in an innovation team and we were trying to how it was just on that cusp of right around the financial crisis whereby a lot of things started getting difficult in the financial services sector. We were working with some the sales teams who were basically had this bottleneck around bringing new clients on, which is in order to bring in new clients and they needed to kind of do background checks to make sure they weren’t bringing on like a fraudster or a terrorist, etc.. And they had some internal data that they could check. They had data from data providers like Bloomberg and Reuters and these guys that they could check to see if, you know, if company APC was a front for terrorism or not. But actually, the most useful source of data for them were websites. Hmm. But what they were doing was they’re having to. They had a whole team of people. They had a whole room full of people. And basically doing the onboarding of this of of customers for the banks, opening lots of tabs, going to lots of websites, doing the same search again and again and again. And David and I would just sort of, you know, we were trying to solve this problem. And I remember being in this one particular meeting room and it was like, what if we could just like, you know, you could sort of firebug because it was then switch to this tool in a web browser that allowed you to say, you know, to see the code behind a web page and highlight things on a on a page. Like what if we could just like instead of having to. What if we could just automate all this so that instead of these this team having to visit all these web pages in a web browser, we could just get the data from it automatically. We could have the business people, not the techies. We could have the business people building the integrations to the websites because there was about there were about 100 websites that we would always search. They were kind of like stock exchanges, financial regulators and the techies of the IAC departments didn’t know about had to be like, okay, what data do you want? But the the business uses the people who are doing this day in, day out, knew how to navigate these websites perfectly. And so, like we, we could build a tool that would allow people to develop the business used in this case to navigate into a website, get the data that they want, do that once, and then basically just automate that process for them so that they could just type in the company name once and then get all the that they needed from all the 100 different data sources that the bank was using to validate that a customer was not a terrorist, basically. So that was what we’re doing in banking. And it a very it was very it became very apparent very quickly that this is a very pretty powerful technology. And the application for it was massive. And so what we left financial services and very happy to say and started import basically to make this capability that we built available to everyone bits manufacturers retailers, iPhone app developers, my dad etcetera.

Bronson: Yeah, no, it’s great. It seems like the real heart of what you’re doing is taking something that used to require programmers or used to require a public API and putting it in the hands of the business people. Even with the new thing, we talked about the auto magical kind of, you know, thing. I mean, it can’t get more simple than that. Just type in the URL and you’re going to have structured data and I’m sure that it’s just taking it to the nth degree. How can we put this in the hands of normal people, not just programmers, and you end up with what import is and where it’s going. So I think that’s a really cool tool for that alone. I remember the first time I saw it and that’s what struck me as, Wow, anyone can do this. Anyone can get the data they need. That’s the game changing aspect of it, I think. So you gave us a great example, the World Bank, and how that you guys used it internally there. But a lot of people watching this, you know, they’re not vetting people that are opening up accounts to make sure they’re not fronts for terrorist organizations. So their use cases for import are going to be very different from that. Do us a favor. Give me some of the use cases for different businesses, different kinds of situations. Help it like really bring this home. How can we in our various businesses use import to get something done with a business process growth, something how other people use import?

Andrew: Okay. I can take you through three very simple example, three examples and we’ll sort of grow on them. We’ll start with a simple one and then maybe it’s more complex and then we kind of expand the vision that way. So a very simple use case that we see is, is basically iPhone apps. So like. An example that I like comes from. The British Red Cross. These guys, they were building a first date app for the iPhone. And what they wanted to do was to include hospital waiting times in in the iPhone app. And the hospital waiting times is basically it’s the same sort of problem as like store locators or store opening times. They’re in lots of different places. You know, you can you go you have to go to a store locator, essentially, but a hospital locator to find the hospitals nearest to you. And then you’ve got the waiting times or in that very basically, they just wanted the the opening times of these hospitals in their iPhone app they used in. There wasn’t an API for this data. The data was available on the web that was freely available on the web and they built an API using imports to it was a live query. So with import you can basically pull down all the data at once to use offline or you can use it live and feed a query and to say this is my postcode where with an error in this case it which is going to be a postcode where the nearest hostels to me and what to their opening times and they built an API to do that. And then in their iPhone app they were able to kind of display all of the waiting times. Very great.

Bronson: Example. Yeah. Just helps us get our head around it.

Andrew: Something building on that is one that we’ve actually seen a couple of times in manufacturing around. There are kind of a couple of different versions of this, but it’s basically pricing strategy where manufacturers and web products are concerned. You are in a you work in an electronics company or you work in a products or manufacturing company in the marketing departments. The product division says, hey, we’ve got a new hard disk drive for you. We want that. We’re going to release this next quarter. How much did it cost? That’s your job as a marketeer to answer the pricing question, like how much should this new hard disk drive cost? And what we’re seeing, what we’ve seen a number of manufacturers doing is basically pull pricing data from all of the comparable products, hard disk drives. It’s a real use case which is Western digital on this and whereby and what they’re doing is they are the pulling like data for like 16,000 hard disk drives down and they’re getting all of the product information about how disk drives have its storage capacity, whether it’s an internal hard disk drive or an external hard disk drive, whether it’s a network what’s known as a NAS network, hard disk drive. And, you know, all these features, dimensions, how big it says color, etc., and then how much does this cost? And so they’re using this this data to kind of build a model that aligns them with this new product to say, okay, we reckon that’s this new product should we should pitch it in at this price.

Bronson: Gotcha.

Andrew: So that’s something a bit more kind of sophisticated and then kind of going on a little bit more. So what we see is there’s a nice one in the recruitment industry whereby recruiters are able to use data via AI to pull jobs off the careers pages of their of their customers so that if you are sitting work, if you are recruitment, this is basically Leeds. So it’s like a Leeds pipe for the recruiters. You’re a recruiter, you come in on Monday morning and you’ve got 100 clients that you’re looking after and unbeknownst to you and one of your top clients who you’ve got a great relationship with but you haven’t spoken to them this week, you weren’t planning on speaking to them this week. They just posted ten new jobs to the careers page on their websites. So what they’re using input for is basically as soon as this happens, the data gets pulled from the correct from the careers page. The website files straight into Salesforce so that the sales guy gets notification to site hat in a ring ring mark is there you know your friend, the talent recruitment is kind of on the other side and the client give him a call because he’s got ten new jobs today and you’ve got a bunch of candidates he’d send over. So it’s kind of again, the difference example, but it sort of gives you an idea of the diversity of things that we’ve seen people doing.

Bronson: Yeah, that’s the word that came to mind is diversity. This isn’t just a single use case. I mean, this is data matters in a lot of different situations and that’s really where import comes in. All right. What’s the single most clever thing you’ve ever seen anybody do with import? Maybe. Clever, funny, clever, stupid, clever, awesome. You know, however you interpret it.

Andrew: Dead people.

Bronson: Explain these.

Andrew: So it’s quite a it’s quite an interesting use case. I have some friends. And it’s two brothers and their father is a real estate agent. And they you know, it’s like any good real estate agent, always looking for leads, always looking to know when.

Bronson: They’re think I.

Andrew: Go see coming on the markets. And what they did is in the area in which their parents live, they went and found the death records and they started pulling the records about the deceased from the kind of public records. They then use that with very often what basically property comes onto the market this way because someone dies and then the property gets sold and then the assets get distributed among the beneficiaries. But basically they were using their scraping data about dead people to provide leads ahead of time to the real estate agent. Father And yeah, it worked. And I know they’ve the guy, you know, obviously it’s a sensitive subject, but he was able to get in with some of that. So always be the first to know about this stuff.

Bronson: Yeah no that’s that’s a great example is definitely clever if nothing else, that’s for sure.

Andrew: But what else have I seen? I’ve seen people from a kind of growth hacking perspective. Sure. I have seen this is a super hack example, but a beautiful use of the technology. And sometimes when you’re if you’re starting a new business or, you know, you’re looking to grow a new business, you you effectively, if you’ve kind of worked that product market fit you sort of, you know, the kind of companies you’re after and you can generate a list of by a thousand websites of like who kind of perfect target customer. Very often the challenge is then like emailing, you know, knowing who to contact and thinking about kind of reach out strategies, etc.. And sometimes the biggest challenge is just working out the email address. And so there are there are tools out there where you can kind of that where people sort of share sounds, guys can share email addresses, etc., and this sort of thing. And one thing that I saw, which was very cool, was using imports too, in two ways. First was to run a thousand URLs through imports looking for MailChimp subscription buttons. Mm hmm. Automatically clicking on those MailChimp subscription buttons within with an email address. Very often you get a hello, thanks for signing up and then using that and then replying to those emails. You kind of like how you don’t know me, but because the logic being of the person who was doing this was to say in a very often there’s a human on the other side of hello at large, an income is actually kind of managing these things. Yeah. But you know, it would take forever and would be a pretty not very pleasant task to do that for 1000 kind of websites. And if you I mean, this guy is he’s he works in the distribution team at 500 startups now. So his problem is he has to go in every and every week. That’s a use this new company who are thinking about how do they you know, how do they reach out to a thousand companies? He’s got to generate a brand new list of a thousand websites for them every day and then every day and then reach out to them. And he’s using import to do that.

Bronson: Yeah, no, that’s a great example. And so they gave us that one because that’s the kind of stuff our audience wants and needs to hear. I actually saw something online where you combined import with using Mechanical Turk and you know, the example is just kind of a funny example I think with sweaters or something with Christmas jumpers. Christmas jumpers, yeah. Yeah. Well, a jumper is if it’s a British puzzle. But talk to us about Mechanical Turk, because I’ve always been kind of obsessed with Mechanical Turk. I love using it. It’s almost like this untapped thing that nobody knows about, even though it’s right there in front of us. And it’s amazing. Tell us what it is and then how that might work with import, because I think that’s going to spark some crazy ideas and people when they realize the overlap of these two.

Andrew: Yes. So Mechanical Turk for those. Well, let me let me rephrase this. Mechanical Turk is it they explain themselves as being a market place for micro tasking. And what this means is that there are the market kind of give you an idea of what this actually means. It’s like a have a mechanism of how it works is there are requests and there are workers. And as a requester, what you can do is submit. Basically, if you create like a little form that describes a task, it’s like a single web page, like a Google form, something like that. And then you load it with data, and it’s basically the kinds of tasks that people load up to Mechanical Turk or things like, I know it’s been used a lot with detecting adult content in images on social networks. So basically what the request to do might be applied large social network. They will fire in a light of images that have maybe been flagged. And then basically they’re asking the workers to say, does this image have adult content in it? Yes or no? Or sometimes I never knew that this was being used for training kind of things like object and image recognition algorithm and say, you know, does this image have a dog in it? Yes or no? And you can create training sets like this. So as a worker, you can sit there and be like and go through a bunch of images being like, yes, this image has a picture. It has a face in it. Yes. No. Yes, no. Yes. No. And that that your human responses are kind of fed back to the requester. You can do things. It’s essentially it’s and all of this is available over an API so that there’s like a nice user interface for it whereby you can sort of upload tasks, etc.. Anyway, let’s step back a little bit. I remember Mechanical Turk when it came out. I read the theory as I’m like, God, this is awesome and same here. Like a kind of an army. If like people they have like I think half a million workers online working around the world and all available over API to do the sorts of things that only humans can do. And when I thought about it like this, amazing, etc.. But I never used it. I just knew the theory. But so I went to the Christmas sweaters. Exercise was really useful for me because I actually went through using it and having used it. There’s a little bit of a learning curve, but I’m used to it, I guess. Yeah. So it’s kind of brings it kind of brings to life. I’ll kind of explain. Next, the Christmas sweater example. Basically, we do lots of content stuff where we talk about data and why we tell stories with data. We do this because largely data is boring and it only in order to engage people. We’ve got to kind of make it interesting somehow, because otherwise I’d just be saying it took some animal to snip sleep. Yeah, it was Christmas just gone and there were lots of Christmas sweaters available. It really is become a thing these days to kind of wear bad Christmas sweaters. There was a website that we found that was selling that all they were doing was selling Christmas sweaters. There were like 16,000 Christmas sweaters all secondhand that they were kind of selling. We used imports to basically pull down all of these Christmas sweaters into a dataset. The image of the Christmas sweater, the title, that’s the name of the Christmas sweater, the description. It had some dates around it, like the year. Mm hmm. But there wasn’t that much kind of, you know, text. That is not there’s no there aren’t many numbers. There’s not much you can do with images. And it’s quite hard to build. Basically, I wanted to know about the Christmas jumpers themselves, and the data from the website wasn’t going to tell me very much, but the images themselves, I was like, you know, I wanted to ask questions like, How many snowmen are there in an average Christmas, Christmas jumper? How many times does Father Christmas appear on an average Christmas jumper? You know, snowflakes kind of, etc., etc.. So what I did was uploaded these images to a kind of tech and asked the workers to basically a couple of very simple questions, saying, how many times for Christmas? How many simonovic? How many reindeer? And then with a couple of thousand of these, I could sort of say the average Christmas jumper or the ultimate Christmas jumper looks like this, etc.. So with with import, you can you can use imports to pull data down from the web. Some of the data will be kind of numbers, but you can do analysis on. Some of it may be text and there are some tools that you maybe if you want to get numbers on, unlike paragraphs of text analysis, there are some tools out there, but certainly with things like images, etc., you actually have to kind of get data about images. Your best bet is to get him sorted out. Yeah. And it’s super cheap. Yeah, it’s super affordable. I mean, I got I mean, you get a thousand images done cost I think $50. Yeah. And I got them back. As soon as you upload your data, you start getting you start getting the data back immediately you get feedback on your data and it was finished within an hour.

Bronson: Yeah. And the reason I like to import Mechanical Turk together is because with import you’re creating an API where there wasn’t one for website data. With Mechanical Turk, you’re creating an API where there wasn’t one for micro human actions. So you put them together and now you have two APIs, two things that are currently were no interface to deal with before. And when you start combining those two, you have the data, you have the people, and you can just create some really crazy machines of your own using this data. But because of what import and what mechanical Turk can do. And so I think it’s just up to people’s own creativity to how they could use these two these two tools together to create some really amazing growth machines. So I’ll leave it up to our audiences imagination to go from there. What you can do based on the sweater idea. The jumper idea.

Andrew: Sweaters.

Bronson: Sweaters. There you go. So as you grow an import, let me ask you this. You know, you’re in growth mode yourself. You’re getting new customers. You know, those kind of things. What’s been your best growth strategy yourself? How do your new customers hear about you? How are you acquiring them? What’s that look like for you?

Andrew: So I think the best thing about growth is you’re constantly experimenting and you’re trying different things. And I think this is I think this is because different growth strategies work at different stages of a business. So one of the things that works really well for us in the early stages was going to conferences, actually going to conferences that had starts at competitions. We entered like nine of them and we won eight of them. And, and that. We’re sitting on some of the wards in this and this office and basically we went to the starts competitions can give you if you pick your the conferences that you’re going to, it can give you great exposure. I think it’s quite important we we kind of pick these things as like marketing tools and we went since every single one absolutely determined to win it. So we were kind of like, you know, whatever it takes to win it. Practicing the pictures and practicing the demos. I mean, we learned a lot and got a lot better at that. But then, you know, 8 to 10, nine strike rate was pretty good. So that’s a you know, it’s kind of evergreen. So that’s the great strategy that works for us very well and year one but doesn’t work in year two because we’d kind of targeted all of these conferences once already, and you can’t do that in year two and it’s someone else’s gain apart from anything else. So that was something that works really well for us. And another thing, and this is something that I’ve seen a lot. This is something I actually kind of learned from our used from inputs, lawyers uses inputs that is used a lot by growth hackers to find people, you know, you have a product market fit, you just need to basically tell people about your service. And if you target them super well, then it doesn’t feel like spam. It’s it’s like, okay, this is yeah, this is and this was actually what I was looking for. And so we’ve seen we’ve seen a is a very some cool examples of that as a company called storefront based in based in San Francisco that.

Bronson: Basically came on growth hacker TV and talked about how they use them it actually really.

Andrew: Was that was that matter.

Bronson: It was it was yeah.

Andrew: High time that I was watching. And so what Matt was doing and to remind anyone who hadn’t seen that episode, they were so storefront that kind of like the way I think content is like an Airbnb for commercial spaces. So there are some people who’ve got commercial space, there are other people who want commercial space on a short term and they do pop up shops. There’s a marketplace for pop up shops and they had a they had product market effect people. It was a problem. They had a great solution. It was just a matter of letting people know. And they used imports to go and find both sides of this market. So they were the where the the people with commercial space were concerned. They were going and saying they’re going on things like Yelp and other kind of directories looking for people like the bars and nightclubs or art galleries very often who sometimes have this dead space. And they were using import the crates of an API to these sites to pull data down, putting things like name, telephone number, email, address, location, etc. and then putting it into a kind of marketing campaign. The distribution channels that they used ranged everything from the traditional email to SMSs to etc. and it was it was kind of all kind of very templated. The reason it worked so well was because they were pulling so much data down from the websites about people. They could say, Hi, Brunson, I walked past your nightclub from a category I read every time there’s a template. Hype Brunson I worked past your nightclub on Fifth Avenue. Have you ever thought about doing this? And this is what I’m doing? And I they did the same thing for the people who wanted retail space. So they were looking at Etsy. They’re pulling all this information down from people’s Etsy profiles so that they could do a similar thing. So it was like name top selling products to the area which they’re located in. So they could say, Hey, Brunson, you, I just saw you. I just came across your Etsy profiles. See that you’re I love your homemade teddy bears. I see that the Barack Obama that is like killing it with that product. Excuse my language I love that product. You know, have you ever thought about selling it? And in a retail space, we’ve got three spaces in your area and Chicago and etc. and so it was what it said. Watching users use input to do things like this informed some of the kind of reach out strategies we’ve done. So we’ve did something around Odesk whereby while we’ve made import available to lots of non-technical users, it still helps technical users because previously technically users would have to spend between 4 hours and four days writing a web script that if you go on Odesk looking for a technical freelancer, you can find people who say this is scraping and we used import to basically build an API to Odesk to find a list of a couple thousand people who said they were proficient. Web scraping. We then used import to correct another API that used the messaging system on Odesk to invite them to a job. So it wasn’t spam. We were actually paying people to kind of build things, but so we put some jobs upon Odesk that were real jobs that we wanted people to do to build snipers using imports, but then invited these people, these, these kind of guys started out as to say, hey, would you be interested in applying for this job? Which is kind of how it works on Odesk. And for every job that we did we messaged I think like ten people, maybe a bit more than that and it was another way of kind of marketing because.

Bronson: They were your users, because if they can use import for other clients on Odesk then it saves a lot of time, makes them more money. And so it’s a brilliant way to use import actually to find your target market. Now those are great examples. I love the storefront. I love the one you just gave. Let me ask you this last question. You know, closing out here, this is an incredible interview. I know people are going to learn so much just thinking about all different ways they could use import. But what’s the best advice that you have for any startup is trying to grow.

Andrew: And hire the right people, which is a pretty standard startup advice. But I think it’s especially important, like growth is concerned. You’re looking for you’re kind of profile of person, someone who is not too precious, who has the attitude of like shit, let’s give it a go, let’s try it. It might work. If it doesn’t work, yeah, we’ll take it time. No worries. No big failure, no shame. And if it does work, we would put a visionary person one out. That’s not the kids that work or not. If it works, let’s do version two and put some more resources into it. And and I think there’s a I think getting that psychological profile that the kind of person who’s relaxed about experimentation and having fun, I think it’s really important because then they can help you build. They do the experimentation and that you can then as a larger as an organization build this is around the winning strategies that that can help you then scale out and then there’s growth. People remain at the kind of cutting edge trying experiments. Most of them don’t works and them do and they’re amazing. And it’s all just a learning kind of a constantly learning.

Bronson: Yeah, that is such great advice and I know that’s great advice because literally that describes the people I work with and we’re in the process of hiring right now and it describes who we’re looking for. I mean, we’re looking for people, like you said, that are not too precious, that are not afraid, that are willing just to do something. And so you’re absolutely right. I mean, I couldn’t have said it better myself. So I’m really glad you give them that advice. Well, Andrew, this has been an incredible episode of growth narrative. Thank you so much again for coming on.

Andrew: Thank you. Brunson Thanks for your time and thanks everyone.

Are you an entrepreneur who is trying to grow a startup?

Get the strategies, motivation, and in-depth interview with all the details every week!