Kirnu's logo

Introducing Kirnu, an automated approach to churn prevention

Aleksi & Oskari on February 17th, 2021


  • Kirnu is a churn prevention tool for SaaS companies. The platform utilizes the latest developments in sequential modeling and deep learning to produce customized recommendations to undertake to fight churn
  • High degree of automation and resilience to change make Kirnu stand out from its competition
  • With Kirnu, we wanted to make advanced churn prevention technology available to all. This is highlighted in Kirnu's easy set-up
  • We are approaching beta test launch and expect to get there during Spring 2021. Join the wait list to secure your place among early users!


Kirnu is a churn prevention tool (web application) for SaaS businesses of all sizes. We designed Kirnu to make SaaS companies' fight against churn as effortless as possible, while maintaining applicability to all use cases. The platform utilizes the latest developments in sequential modelling, deep learning and churn prevention analytics to deliver customized recommendations for preventing individual users from churning.

At the core of Kirnu's platform is the Churn table (simulated data)

Example of Kirnu's Churn table

Note: the screenshot above illustrates Kirnu with traditional churn prediction selected. Users implementing an uplift model with Kirnu will have additional fields in the table for estimated uplift and recommended action. You can read more about uplift models from our blog here.)

The table enables you to get a broad view of your customers' health in terms of churn probability with a glimpse. More specifically:

The table is easily sorted, so you can for example sort your customers based on probability and focus on the customers with the highest values.


Kirnu's deep learning algorithm relies on the data it receives to make predictions on churn probability. In general, the data can be divided into two groups:


Having seen the importance of churn on SaaS companies, we built Kirnu to make data-based churn prevention effortless and accessible. As a result, Kirnu's set-up and maintenance are made as easy as possible. All you have to do is add a JavaScript snippet to your application (copy & paste) and provide selected additional data through our API. Additionally, we wanted to provide a solution which targets churn specifically instead of as a byproduct of other features.

Kirnu solves many challenges unaddressed by competitors. For example, let us segment the churn prediction / prevention solutions on the market into two by asking whether they offer:

Companies in Group 1 provide an algorithm to which users are to provide the data, either manually, through data import, an API or an integration. Their platform analyzes the data and provides predictions. Rather similar to Kirnu, no? Well, yes, to some extent. Using solutions provided by companies in Group 1 requires you to have all data ready and neatly organized. What if you do not collect event data? Then you cannot use it, simple as that. Kirnu provides a predictive algorithm equipped with tools for collecting event data.

To provide a solution to the shortcomings of solutions in Group 1, companies that provide tools for event data collection have emerged. They offer event data collection either through:

Solutions in the former group (event trackers) require you to install event trackers, which track predefined events in your application. This is a laborious manual task, that requires some programming experience or alternatively the use of a slower, but less-programming-experience-requiring solution to manage the trackers, a graphic user interface. Companies in the latter group (auto-collection of events) however, have solved this problem by introducing "auto-collectors", JavaScript snippets that automatically capture event data. We consider solutions of this type as the most sophisticated among churn prediction software. The biggest drawback of currently available auto-collectors is their ability to adapt to change. If the structure of a web application changes, they easily lose track and need to be manually configured.

We wanted to improve on this, and built Kirnu to be resilient to change from the beginning. With Kirnu, you do not have to worry about changing the structure or contents of your application. Our algorithm adapts to the changing environment on its own and is able to connect the dots on its own. This is why we dare to call our solution "automated to the extreme".


We (Aleksi & Oskari, two brothers) began working on Kirnu at the end of 2020. A lot has been achieved since then, and we are now approaching a private beta launch. We estimate the beta launch to happen during Spring 2021. As of now, we are fine tuning bits and pieces before beta launch to maximize test users' ease of use.

We hope this introduction got you as excited as we are! If you want to stay tuned, make sure to secure your early access by leaving your email below. We will also regularly (not more than once a week, however) post updates on the launch schedule.

- Aleksi & Oskari (feel free to contact us at [firstname]

P.S. We are looking for test users within the SaaS field who want to reduce their churn. If you are interested, drop us a line at and we would be happy to have you on board.

Sign up for early access

Enter your email below to hear from us (maximum once a week) and to secure your place among early users!

Thank you for your interest and welcome! You can also follow us on Twitter @kirnuio
Oops! Something went wrong while submitting the form.

P.S. we expect to be beta test ready during Spring 2021 and are looking for test users. Drop us a line at if you are interested in reducing churn in your SaaS application and getting exclusive access to Kirnu.