October 17, 2019
My name is Kevin and I’m 27. I’m French and I live in Toulouse, a beautiful city in the south of France. I’m the founder of ScrapingBee, a web scraping API that I co-founded with Pierre de Wulf my lifelong friend and partner.
After high school, I went to a French university to study computer science. I’ve always been passionate about computers, the typical story of the nerdy teenager spending time programming and building websites. It was really clear to me that programming and building “computer things” was what I wanted to do with my life, so I naturally pursued a bachelor’s degree in computer science.
I really hesitated to pursue a master’s degree in computer science but was feeling bored. I felt like most topics we were studying were very theoretical, and I wanted to build real things.
So I decided to start working for a startup, Fiduceo at 21. They were doing “bank account aggregation”, like Mint.com in the US. Technically speaking, one of the biggest things we did was “Web scraping”, meaning gathering data from the web. I found this interesting because you can do so many things by extracting data from the web.
I learned so much during my first year. It was a really small startup, we were 10, so I had the chance to do lots of different things. Of course software development on all kinds of different tech that were new to me, but also technical account management with small customers etc.
Our office was in a tech-hub in Toulouse with lots of different startups. Every week there were lots of meetups about entrepreneurship and startups, and I talked with a lot of different people there. That’s where my interest in creating my own company started.
After a year we’ve been acquired by a big French bank, they forced us to move to the HQ and some people quit. So I joined the big French online bank, took a few more responsibilities and worked on lots of new projects. I stayed there for two years. It was also really interesting because I learned new things there. How “enterprise” software works, how big companies works, regulation, compliance ...
That’s also where I learned about office politics, bureaucracy, long processes, endless meetings… Welcome to big co :)
After two years I had enough and I decided to quit. I didn’t know exactly what to do, but one thing was certain, I wasn’t happy as an employee, and I needed to do something on my own. I wrote a book about web scraping during my first year and did some freelance gigs. For those who don’t know, web scraping is the act of gathering data from a website in an automated way.
There are many use cases, like news aggregation, monitoring prices on E-commerce websites, market research, monitoring social media accounts…
After I was done writing the book, my partner Pierre quit his job too. We had been thinking about starting a business together for a long time, and we both decided that it was the right time to do it.
Prior to ScrapingBee, Pierre and I worked together on multiple projects involving web scraping at scale, such as PricingBot, an E-commerce price monitoring SaaS and Shoptolist, a B2C universal wish list app.
In our jobs, both Pierre and I had to deal with web scraping, I don’t like the word “expert” so let’s say we knew a thing or two about this field.
In all our projects, there were two technical things that we had to solve over and over again:
Headless browsers are like regular browsers, but can be launched on servers, without the need for a graphical interface. The most famous is Chrome. The problem is, it is really resource-intensive in terms of CPU and RAM and really complicated to manage at scale.
These two problems are really common when you decide to scrape websites, and we decided to leverage our experience with this and package it into an API.
From day one we released a landing page with a signup form to gather beta users and feedback. In the three months it took us to build the MVP, we were able to get 200 subscribers to our newsletter, half thanks to Ship by Product Hunt:
And the other half by marketing our landing page on various communities, Reddit, Facebook groups:
The MVP was done quickly because we had a lot of existing code from our previous projects. Once we were done coding the API, we soft-launched to our newsletter to get our first beta users.
A small percentage finally signed up, maybe 15 to 20% for different reasons. The first one is that we didn’t communicate enough with them during the MVP building phase, and the second one is that our landing page wasn’t (and still isn’t) perfect, and some people misunderstood what we were doing.
During this quick “beta”, we fixed a lot of bugs, but more importantly, we worked a lot on our infrastructure to be as cost-efficient as possible. It is well known that “premature optimization is the root of all evil”, but being bootstrapped meant that we could not afford to launch a business that was costing us money. So we worked hard to have a scalable and profitable business from day one.
We considered that the beta allowed us to validate the idea, so after some fixes and optimisation, we have decided to take things to the next step: ProductHunt and heavy content-marketing.
In our first month, we got something like 200 sign-ups, and 10 paying customers, so a 5% conversion rate, which is really good, and much better than our previous project.
Content marketing definitely. Both Pierre and I are enthusiastic bloggers, we love to write and share what we know. So we did exactly this to launch ScrapingBee, and it worked like a charm.
Most companies do content marketing for the sake of it. But we really love doing it, the last blog post we made was 3500 words long about programming and web scraping. And people loved it, especially programmers. We got tons of shares, tweets, and engagement.
One big thing that worked for us is building relationships. As software developers, we tend to often think about our product only through code, we try to automate everything and avoid interactions with people.
And it’s a shame! Most of our big customers became customers after one or a few calls and several follow up emails.
In the beginning, it was complicated to get people on the phone, but we quickly put a banner in our dashboard so that people could schedule a call and get free API calls in exchange for a 15mn calls.
And it worked like crazy, immediately after doing this we booked several calls per week, we learned a lot about our users, and we closed many deals.
We launched ScrapingBee beta in July and started accepting payments in August. In terms of traffic it looks really great, on average we have 5000 unique visitors per month for the last three months.
We also went from 0 to $2000 in MRR and are profitable!
We have lots of things to do, we just released an “API-store”, which is a store where people can use ready-made APIs that returns formatted data to scrape popular websites like Instagram, Google, without having to maintain the scraping code.
People have shown a lot of interest for this, and we had many feature requests for new APIs so we need to work on this. We need to be careful in order to build what people really need, not what they want :) Meaning before developing anything, making sure people are going to pay for the new API.
That was for the product. We have lots of things to do and to experiment about growth. We’re just started to audit our SEO and work on our strategy, regarding link building and content. There are so many things to do in this field!
Then we also started cold outreach, and are seeing the first results but this is a slow process. We are targeting SMBs up to 1000 employees. These companies can take up two weeks to answer a simple email, we don’t have the same time-frames and it can be difficult to adapt.
Then we also need to improve our landing page, our copywriting, the email sequences with send…
I don’t think we’re going to get bored in the near future :)
Yes, I learned a lot! First thing is, creating a SaaS is not the difficult part. The most difficult part is distribution.
How to get attention from people when there are thousands of SaaS products launching everyday. When you have people’s attention, how to sell the product.
Creating a SaaS is not the difficult part. The most difficult part is distribution.
I also learned the importance of so-called “soft-skills”. There will be many ups and downs in your business. Being resilient in the long run is one of the most important things.
Patience is another very important soft-skill. Things take time. For example, some SEO strategies will take 6 months before seeing the first results. Some enterprise customers will take two weeks to answer your email, two more weeks to test your product, and two more to make a decision and buy it.
Our tech stack is really classic, we use Python / Flask for the web app, Stripe for payments with Chargebee. Chargebee is an amazing tool, it handles VAT and other complicated tax stuff for you, provides great reporting, lots of automation are built-in (sending automated emails when a payment fails, when a credit card expires… those things happens ALL THE TIME). For hosting we use both Heroku and AWS.
Basically all our project management is made inside Notion. It is really a game-changer for us. We work remotely, and having such a tool to keep track of everything is really awesome. The thing we love about it is that it is really modular.
Everything is based around blocks, you can have tables, to-dos, pages, embed everything you want. We also use Notion as a CRM, with reminders.
Then of course Slack! We use it to communicate but also to centralize important notification about the business. For example, when a lead fills a form, we got a Slack notification, when a payment fails etc.
The other really useful tools we use is Calendly / Zoom for scheduling and taking sales calls. Calendly is amazing, it’s directly connected to my Google calendar, and people can book a call inside my calendar themselves, no need for the classic conversation “Are you available on monday 3:30 PM GMT+1 ?”
Then it will automatically send them a Zoom link for the meeting, and reminders!
In France we have this wonderful “investment” / “accelerator” company called “The Family” which provides capital and education to entrepreneurs. They have a Youtube channel called StartupFood, where you can find lots of talks about any subject with experts.
They provide really good advice and insight about startups. The thing is, their mindset is all about VC funded startup that seeks huge exits. Then of course there is the Ycombinator online startup school.
I also like to hang out on communities like Indie hackers or various SaaS facebook groups and see how others are doing, learn from them and exchange.
For specific skills, like marketing, sales, copywriting there are tons of resources online, both free and paid. But don’t spend too much time taking online courses, learn by doing!
Watch the Jeff Bezos video about his “Regret minimization framework”. https://www.youtube.com/watch?v=jwG_qR6XmDQ
Then do it!
We are not looking to hire right now! But we are interested in freelance content writers.
You can go to our
website here: https://www.scrapingbee.com
Here is my Twitter: https://twitter.com/SahinKevin and
Pierre’s twitter: https://twitter.com/PierreDeWulf