Our AI Writing Tool Won 'Startup Of The Year' And Has Customers In More Than 150 Countries

Matej Guid

•

Published: March 13th, 2022

save

Matej Guid

Founder, InstaText

$50K

revenue/mo

Founders

Employees

InstaText

instatext.io

from Ljubljana

started March 2019

$50,000

revenue/mo

Founders

Employees

Discover what tools Matej recommends to grow your business!

platform

trello

social media tool-icon

YouTube

Twitter

productivity tool-icon

Slack

Trello

payments

Paddle

crm

PipeDrive

other

Zoom

Discover what books Matej recommends to grow your business!

📗

Want more updates on InstaText? Check out these stories:

Hello! Who are you and what business did you start?

Hello, my name is Matej Guid and I am the CEO and co-founder of InstaText, an AI-based online writing assistant that helps us write like a native speaker.

For non-native speakers who use English in their professional communication, our AI-based online writing and editing tool helps them rewrite and improve their texts. Unlike conventional grammar checkers, which mainly correct grammatical errors, InstaText also enriches the content and makes it more readable and understandable.

InstaText was Slovenian Startup of the Year 2020. Since the product launch at the beginning of 2020, customers come from more than 150 countries all over the world and there are more than half a million active users.

We bootstrapped the business from $0 to more than $50k monthly revenue in two years and no end in sight.

my-ai-writing-tool-won-startup-of-the-year-and-makes-600k-year

What's your backstory and how did you come up with the idea?

My background is a combination of entrepreneurial experience and academic research. Shortly after completing my doctoral dissertation in the field of artificial intelligence, I became a professor at the university. During all this time, I was also the founder and CEO of my first technology startup.

It started with a real problem in academic writing:

The initial drafts are often poorly written and need to be rewritten completely.
Editing an academic paper takes a lot of time and effort. Sometimes entire paragraphs need to be rewritten completely.
There is no help available and you need it in real-time.

There was a clear need for a tool that could help us write better English in real-time.

The last decade has seen significant advances in AI. Deep learning enabled significant breakthroughs in language technology. However, it was the breakthrough in reinforcement learning, widely known as AlphaZero, that enabled powerful technology that continuously improves and leads to better and better performance.

So I asked myself: "Could we set up improving text as a game?" If a computer that knew nothing about chess just a day ago was able to learn how to play the game better than any human and to do so by only playing against itself, could such an approach also be applied in computer language acquisition? For example: can we maximize the readability of the texts?

The idea emerged: to create an online tool that would utilize artificial intelligence to provide instant recommendations and ideas to improve the content of a text written in English. Needless to say, this is easier said than done.

However, I had a novel idea of how to use the emerging technologies to really solve the non-native speakers’ problem quite effectively. I tried the new technology on some of my own academic papers and it turned out that the envisioned approach was really promising.

Take us through the process of designing and developing your tool.

After obtaining some very promising results (and it took a long way to arrive at this point), I presented this concept to Marcus, a language technology expert from Klagenfurt, Austria, who immediately recognized its potential.

We set to work and only a few months later (in 2018) we engaged a front-end developer. Early in 2019, we also engaged a back-end developer. Fortunately, we had a basic infrastructure and some budget available from my previous company, we also had more than a decade of entrepreneurial experience.

In May 2019, we offered the initial version of the product for testing by some of our friends and other people from different fields. And there was a lot of enthusiasm about it, even if the overall user experience was far from perfect.

By September 2019, we already had an MVP - a minimum viable product. Perhaps not polished enough for the people to buy it, but good enough to learn how to arrive at a sustainable business model.

Describe the process of launching the business.

In March 2019, we founded a company to obtain funding from a state entrepreneur fund. Our application was successful and we obtained a grant of slightly more than 50,000 € - however, we didn’t receive all the money immediately and we had to accomplish specific milestones in the upcoming two years. But it turned out that something was at least equally important as the grant: access to mentors and various educational workshops.

Aim for a small team of skillful people. They can have a great impact even when working part-time. Be patient and know that sometimes less is more.

In September 2019, we opened the product to the public, to learn:

What would be a suitable business model,
Who are the most promising target customers,
In what ways we can improve the product.

Note that there was no advertising at the time.

We engaged a professional salesman, who spent part-time trying to find out what could work. It was commission-based only - we couldn’t afford to pay a salary - but with a huge commission of the entire year of revenues for each customer.

We learned from one of our mentors that to be able to sell a product, the price range can be either:

Up to $200 - to be able to sell online.
More than a couple of thousand $ - to be able to afford a salesperson.

In other words: It is almost impossible to sell products that cost about $800 - make sure that your product doesn’t fit into that category!

So the goal was to find out in what context our product can be sold for a couple of thousand dollars. In other words, which companies or institutions would be eligible for some kind of team licenses.

It was very important to “get out of the building” and talk to customers. There was an important book that helped us learn how to talk to users and potential customers, recommended by one of our mentors: The Mom Test. With a meaningful subtitle: “How to Talk to Customers and Learn If Your Business is a Good Idea when Everyone is Lying to You”. The point is that people don’t give you useful information if you ask them hypothetical questions. It’s much better to let them talk about their past experiences and see how they solve particular problems.

Since launch, what has worked to attract and retain customers?

Our customers come from the following segments, I also list the benefits InstaText brings them:

Research and academia: significantly increase acceptance rates and shorten the time to publication, boost productivity.
Business and marketing: communicate effectively, write clearly, produce high quality and efficient sentences, and be understood.
Copywriting: improve text interactively and quickly get ideas on how to improve the content, increase writing quality and learn how to write better English.

Our first customer came in October 2019 and it was a translator. She bought the license online, no sales were involved. By the end of 2019, we only sold 6 licenses. However, I was not worried at all - quite the opposite, we have learned a lot in this period. Note that we were focused only on learning at this time. The product was ready in January 2020 and then started another phase, in which we focused on sales and revenues.

The important decision was taken in December 2019: which customer segment should we focus on? We chose academics, for various reasons:

We understood this segment best.
We knew that they are very actively involved in writing academic papers in the first quarter of the year.

We started to organize meetups for academics (live). Although many people were invited, usually no more than 3-5 actually attended the meetup. However, those people were our initial “evangelists” who spread the word and helped us organize presentations or connect us to the right people at their institutions.

Our salesperson completely focused on the academic segment. He organized presentations at various faculties and research institutes and we went there together. It was me who was presenting and answering questions - unless these questions were about pricing and purchase procedures.

In this way, we sold over $30,000 in just two months: February and March 2020. This was huge progress compared to the six licenses sold in 2019, right? :)

The corona crisis came in April and May 2020. Live presentations were no longer possible. We quickly switched to Zoom. It helped us a lot later, because suddenly the entire region became available within minutes, without expensive (not just cost, but also time!) sales trips.

In May 2020, we became the Slovenian Startup of the Year. This gave us an additional advantage on the local market and a lot of publicity.

What about online sales?

We had online sales right from the beginning, however, it was difficult to establish that channel:

Online advertising was too expensive to obtain a positive ROI.
We only offered bank transfer payments at the beginning.
Our pricing model was difficult to understand.

Our goal was to enable global sales without having to worry about accounting and administration. So we planned to integrate Paddle.com, an online global reseller, which was a perfect solution for us. However, with a lot of priorities in the development process, we were able to achieve full integration only in November 2020. That is when we were finally able to sell globally, with a credit card and PayPal payments, without having to issue a single invoice!

In the meantime, I focused on completing various courses on Digital Marketing and Growth Product Management. With the help of one of our mentors, we fixed the pricing model and we were finally ready to sell globally.

I started with a small advertising budget in December 2020, however, we were reinvesting all online earnings into scaling this channel. According to one of our mentors, our growth was “insane” at times.

It was important to find out that we have a practically 100% retention rate in B2B sales - not only that all institutions were renewing their team licenses, each renewal was also an opportunity to upsell.

In this way, in the upcoming year we were able to:

Double our revenues of B2B sales (still having only one salesperson),
Grow our online sales from scratch, the revenues from online sales even surpassed the B2B sales.

Overall, our sales grew from $130k in 2020 to almost $600k in 2021.

How are you doing today and what does the future look like?

We more than quadrupled our sales in 2021 while growing profitably, which was an amazing achievement - note that we are still bootstrapping, we haven’t taken any investors. Moreover, we help hundreds of thousands of users worldwide to become successful with the help of their written communication - and this makes the entire team happy. There is a clear purpose in what we do.

InstaText is now making its way into web applications and closer to the moment of writing. The upcoming browser extensions will enable InstaText in several important web-based apps, including Gmail, Google Docs, Outlook, Slack, Facebook, Twitter, LinkedIn, WhatsApp, and several others. Technically oriented academics will also benefit from LaTex support in Overleaf web client for academic writing.

The international team consists of top experts in artificial intelligence, language technologies, natural language processing, and software development - among others, of course.

We are creating impact and it is becoming apparent - the reviews of InstaText are great and testimonials by the users are fantastic. Written communication is the key to success for many and we really help people around the world. I would love for tens of millions of people to become successful with the help of better-written communication by using InstaText.

Get to know your customers. Never stop doing this. You can outsource almost any task in a business, but you should never outsource learning about your customers

Through starting the business, have you learned anything particularly helpful or advantageous?

We knew from the beginning that the speed of our learning will be crucial for success. If I have to emphasize one point, that would be the importance of access to mentors. I’ll give you a couple of examples.

One of our mentors told us that a company can compete in one of the following only:

Price
Product
Integration
UX (user experience)

We didn’t want to compete on Price (at the time Grammarly started to get big and there were already cheaper versions of Grammarly on the market). We couldn’t compete in Integration, because we were still a very small team. But it helped us realize that it is the Product where we really differentiate. And not UX.

So instead of inventing our own UX, we started to follow best practices from other tools: Microsoft Word, Google Docs, Google Translate. People are already familiar with their UX solutions. Once we understood this, many people started to acknowledge that our product is actually very useful, they recognized the value.

The otherwise thought by one of our mentors was that you can be either Best, Cheapest, or Local. With possibilities such as “cheapest-best” - for example, Mercedes in the automotive industry: Renault cannot make cars of such quality, and Ferrari cannot make cars at such low costs. And generally speaking, there is nothing wrong with being the cheapest - however, your strategy on the market will be totally different.

This was useful in two ways:

We decided to strive to be the best (not the cheapest), which helped us determine our pricing.
We could still take advantage of being local - at least in Slovenia and perhaps Croatia, getting some advantage in generating sales.

What platform/tools do you use for your business?

We use common productivity tools such as Slack (for internal communication), Trello (for planning), and Zoom (for meetings). I already mentioned Paddle, which is extremely helpful to sell online globally. I can add Autopilot, which really facilitates marketing automation. And PipeDrive for sales.

Actually, there are several more - the tools play an important part in increasing our productivity. In my experience, it really pays off to choose the most relevant and powerful tools.

What have been the most influential books, podcasts, or other resources?

I have spent a lot of time learning from various online and offline resources in the past three years. Let me recommend you one pleasant-to-read book and three great online courses that really made a huge impact on me.

Book: The Mom Test: How to talk to customers & learn if your business is a good idea when everyone is lying to you by Rob Fitzpatrick
Course: How to Build a Startup by Steve Blank
Course: Digital Marketing
Course: Growth Product Manager

Advice for other entrepreneurs who want to get started or are just starting out?

Here are some suggestions that really worked well for us:

Aim for a small team of skillful people. They can have a great impact even when working part-time. Be patient and know that sometimes less is more. In computer science, there is a famous saying, "What one programmer can do in a month, two programmers can do in two months." :)
Try to get the best advisors. It is extremely important to do the right things at the right time. Learn to ask the right questions and make sure you implement what you have agreed to. But do not just follow what others suggest - they often have limited information. In the end, it's you who makes the decisions. Argue constructively to get the right feedback. Sometimes an idea can be good if it is refined a little and then implemented well.
Focus on what you are good at and let others on the team focus on what they are good at. Work as independently as possible to reduce overhead and maximize the efficiency of individual team members. Set up weekly online meetings to coordinate everyone and create weekly plans together.
Get to know your customers. Never stop doing this. You can outsource almost any task in a business, but you should never outsource learning about your customers. Keep communicating with them, trying to understand them. Learn how to ask the right questions.
Product-market fit is not a binary concept: once you find a good marketing segment, keep looking for a better one.

Where can we go to learn more?

If you have any questions or comments, drop a comment below!