Overview: Curious about the efficacy of no-code product development tools, I cofounded and growth-hacked ApplyAll.com: a product that helps job seekers land interviews.
Skills: Project Manager + Research + Design Futures
Curious about no-code
I’ve been extremely curious about the recently growing wave of “no-code” product developers. One of my superpowers is building prototypes and getting them in front of users so I can create tools in a “lean” way, and no-code would seem like the perfect opportunity to take this a step further. I come from a coding background, but through work experience, I’ve realized that building full-stack solutions are time-intensive and not conducive to lean and iterative product development. In this regard, ideally, you’d want to push the coding back as far as you can until you have a semblance of an idea with product-market fit. No-code offers an even more intriguing premise; what if you could theoretically eliminate the coding altogether and continuously employ iterative product development.
A Vehicle to Explore No-Code: Founding ApplyAll.com
The past month, Uber laid off a couple hundred marketing executives. Somehow I stumbled upon an internal spreadsheet these employees were using (it’s now ‘public’ here), and after some digging, I realized the process used by ex-employees to apply for new jobs was painstaking and often unfruitful. When you consider that the current employer response rate hovers around 5-10%, it becomes hard to win at the statistics game when it takes hours to fill out unique applications and cover letters.
Welcome to ApplyAll.com
ApplyAll.com is a bot that makes it easy to apply to hundreds of jobs in seconds. I spent this month designing and building out the interface and back-end. From user flow mapping to graphic design and marketing, this sprint required pulling together a lot of my learnings from previous projects.
We used a tool called Bubble.io to create our user-facing website and manage each customer’s profile. It uses a fairly intuitive database structure that uses similar object-oriented logic to PHP and SQL, except without all the syntax and code jargon. Setting up the database around each “user” object with all the data types associated with it was fairly easy, and we were set up within a few days.
The front-end took a little bit longer to figure out. It uses somewhat of a drag-and-drop interface, but Bubble’s responsive design can be quite finicky. At least there was a lot of support on online forums and good documentation on the product.
The app functioned in a simple way for users:
- Select roles you want to apply to
- Upload your resume
- Finetune your search (by locations, job type & blacklisted companies, and other notes)
- Buy a package of 100 applications or 250 (upsell for a resume review)
A quick explainer video below:
As we finished building our user-facing app, we were running into trouble with calibrating the bot that would ultimately send the resumes to job applications. The solution to our troubles started of as a joke between me and my co-founder, but quickly turned into a winning rapid prototype strategy; we hired a fiverr contractor to manually take the resumes and work with the bot to apply the customer to tailored jobs.
To the user, everything seemed like it worked. Behind the scenes, we had patched together manual and automated services to deliver the product.
And it worked! We had users pay $12 for 100 applications and 20$ for 250 applications. Our first few paying customers immediately started hearing back from companies. Initial feedback even sounded like this:
From here, we kept iterating (adding features like blacklisted companies you don’t want to apply to, resume reviews, etc.).
Looking within our networks for the first users: We had a 100% interview response rate for our first 20 customers. These were people from our networks we convinced to use the tool.
Blasting the Uber Spreadsheet for more leads: Moving forward, we went back to that Uber spreadsheet and sent emails to thousands of applicants pitching our service. We had over 500+ respond. Through free upsells, and personal customer service, we were able to capture 40% of them.
Money-back guarantee: As our product reached more people and the same results kept flooding in, we realized that we could implement a money-back guarantee at a minimal expense to us. In fact, the money-back guarantee could help us with the many prospective customers who were reaching out/nervous about putting down money into a company that didn’t seem to have any track-record. Out of over 2300 customers, only 12 requested refunds.
FullSession: We incorporated FullSession to record user sessions and identify where leads were dropping off. FullSession also gave us interactive heat maps and a visual way to track product effectiveness.
Google Adwords and Analytics: This deserves an entire blog post of learnings, but the Google AdWords experience was one of the biggest learning experiences through the process. It seemed intuitive, but narrowing down and refining an effective SEO + advertising model that had a return on investment proved quite tricky. We were able to onboard Fiverr SEO article writers to help out with our search rankings. Given the chance to hire one person, I would have brought on a marketing head because it really is a full-time role.
Advertising: We had a few concepts we never quite got to releasing, but it was a fun exercise.
From product innovation to execution innovation
My focus would be on marketing and execution innovation. I’ve usually let myself get carried away with making my idea unique, but I want to see how innovating on other fronts could lead to success. I’ve been tracking many companies that offer simple enough services, but innovate and create an impact based on alternative forms of innovation, and over the past month, I’ve had a fanatic goal to see if I could build something and scale its impact by growth hacking.
Growth hacking is an important skill I want to have in my toolbox, and I’m currently focused on finding ways to get ApplyAll more traction. The idea itself is fairly simple; instead, finding product-market fit, and, more importantly, getting people to actually pay for a service, is an interesting challenge I want to tackle.
- Go Bubble! My current favorite for no-code product development.
- Incorporating product feedback using no-code takes 1-2 days vs. 2-3 weeks for similar full-stack changes. It’s a no-brainer in most cases.
- No-code would be perfect for client work. It’s quite common to finish a contract project but remain in this constant charade of being called back for “maintenance” work or future bug fixes. Instead, no-code can enable you to even train the client to make fixes using the software by themselves. It makes product development work more resilient and accessible.
- Prioritize no-code in situations where visual UI isn’t as much of a priority. Think internal development products for organizations or tools where quick functionality is more important than bespoke UI. The bulk of no-code development effort on Bubble was focused on creating responsive UI, which could be frustrating at times.
- No-code has its limitations (at least right now). You can obviously get more custom work by employing a developer, but there is a hard tradeoff to make in terms of time and effort.
The bottom line: With ApplyAll, no-code was the perfect way to develop product-market fit. We could get a prototype out in the wild, find customers, incorporate feedback, and hone in on a solution people are willing to pay for. In the long run, however, we could see ourselves developing a bespoke, coded solution in parallel and eventually shifting over the platform when the time is right. It would give us more control and flexibility on the product, and be more cost-effective for a product that scales to much larger audiences.
Using tools like Bubble provided me with a radically new way to think about product development. I’m setting my sights on exploring more no-code tools, and I’m even more excited to see how these platforms evolve and mature.