Wolfy

We learnt from failure, this is how eagerworks started

Wolfy lets users quickly find contact information from targeted prospects. Users just needed to enter their ideal customer profile and they received verified emails to feed their outbound email campaign. This is the story of a lead-generation platform built by the co-founders of eagerworks, while they were starting eagerworks.

Ruby on Rails icon Ruby on Rails
Project cover

Challenge

Get from an idea to a real SaaS product that users would pay. Leverage public information from the internet to create a massive scraper that would feed Wolfy’s search engine. Create an algorithm that would find a user’s work email.

Solution

We designed and implemented a web application where users could apply advanced filters to find their target prospects. This was powered by a massive scraper and search engine that could handle complex, real-time queries

Outcome

A platform where users could search more than 200 million people and their work emails, using data publicly obtained from the internet. Users could start a free trial and later subscribe to a recurring plan using their credit card.
client image

"With Wolfy we really understood how creating a product is much more than just writing code. Thanks to having launched our own tech startup, we are a better agency today."

JP and Santiago were working together in a prestigious software agency and decided they needed their own adventure.

They started off with Wolfy (a product) because they saw it as a win-win scenario. If the startup was to be a success, they would keep their energies focused on scaling and growing the business. On the contrary, having walked the same path as other technological entrepreneurs do now, they could apply the lessons learned on future clients.
 

CHALLENGE

The idea behind Wolfy was to use all the public information that’s already available on the internet to create a search engine where people could find information about companies and their employees.

One of the main technical challenges we faced was to create a massive scraper that could process all the information we needed while being able to run periodically in a cost-efficient way in order to keep Wolfy’s information up-to-date.
Another important requirement was to find a person’s work email address given their name and the company where they worked while being polite to email service providers.

One of the most important challenges was finding product/market fit. Wolfy targeted the Latin American market and what we found was that in order to use a product like Wolfy, most of the companies needed to change at least some part of their sales pipeline. They were using traditional tools like buying outdated databases and making phone calls, instead of using more modern approaches like cold emailing. This made the sales process difficult because we first had to show customers that they could get better results by introducing new tools/processes to their sales pipeline.

 

 

SOLUTION

In order to obtain all the information we needed, we had to create a massive parallel scraping architecture.

The project used Python and Scrapy for the scraper part since it allows for quick iteration and has a great community behind it. 

The backend was implemented in Ruby on Rails, which served as an API for our frontend. For the frontend we used Angular, since as a Single Page Application fitted perfectly our use case: less than 10 different screens with lots of expected user interaction on/between them. In retrospect, it was a great choice since the web application performed almost as fast as a desktop application, without any lag/delay between page changes. Also, the code turned out to be easily maintainable since everything was divided into not-so-big components.

In order to handle the amount of data that Wolfy had to query and to improve speed, we had to architect a sharded database using PostgreSQL. Each shard was responsible for handling data related to one specific country.
In order to improve scraping speed, we used a pool of thousands of IPs that were rotated between scrapers using proxies.

OUTCOME

Wolfy was one of the first products that we crafted at eagerworks, where we took an idea and we created a product that customers were paying for.

We learned a lot of lessons that up until this day we apply on a daily basis to our customers’ products.

Applying the lean startup methodology to our own product lets us see it working first-hand. We could iterate really fast, giving value from day one. Requesting feedback from customers, prioritizing features based on real value to them, finding the right metrics to assess product success, among others, was a fundamental part of the product.

It was our first time raising capital and we learned things like when to raise private vs public capital, how to make a good and effective pitch, what’s important to an investor, and what’s not. We went through all that process, so now we can transmit our experience to future entrepreneurs.

Another enriching experience was deciding our business model, either B2C or B2B. We knew this crossroad was going to lead us to two different sales approaches and strategies, different brand communication and marketing, and so on. We went for the B2B model after concluding it would be the most efficient way to scale. In our case, our decision wasn't too hard, but we do know many entrepreneurs who face this crossroad too, and we use all our experience to help conclude which option is most suitable for their project.

We were able to create a platform where users could apply simple filters to search over millions of data points to find their target audience. A massive data pipeline was designed and implemented to scrape millions of web pages periodically, to maintain Wolfy’s information up to date.

The technical experience we gathered on Wolfy, opened the doors to one of our most important clients: The Appraisal Lane, where we assembled a team of 5 people to solve their scraping and data processing challenges.