Marybeth Alexander turned an internal project into a company built to last

Taking Over an Internal Project to Build a Company for the Long Run

Marybeth Alexander is a cofounder of KnowledgeOwl, software for managing a company's help content and knowledge documentation. She and developer Pete Grigg took over in 2015 a product they had built together at their previous employer and made it independent. It is a case of building a software company that runs without outside investment, on the strength of experience gained in customer support and a product that already had users.

From a Shelved Internal Project to a Paid Product

The starting point was an internal startup program at the survey software company SurveyGizmo. In 2011 Marybeth, who led the customer support team, took part in developing HelpGizmo, a comprehensive support tool that handled everything from receiving inquiries to resolving them. When the internal startup program ended, that plan lost its momentum as well, but the support team still needed a better tool for organizing its work knowledge.

In 2012 the development team in the support division revived HelpGizmo as a '20% project' that let people spend part of their working hours on a separate project. Its purpose was knowledge management inside the support team, and in 2013 the documentation team began using it for the product manuals published to customers. As Marybeth became product manager and Pete led development, the direction took shape as knowledge documentation software that could be sold independently.

The first paying customer came in April 2014. The next year SurveyGizmo's CEO offered the two of them the chance to take over the product and run it as a separate company, and they accepted and renamed it KnowledgeOwl. The product was already profitable at the time of the takeover, and the two, already familiar with customer demands and how the product worked, now also took on management responsibility.

Keeping the Product Focused on Knowledge Documentation

The gap Marybeth found in the market was between professional technical writing tools and comprehensive customer support systems. The products she looked at were often advanced tools for professional writers, or add-on features bundled into another support system. KnowledgeOwl focused on knowledge documentation while connecting with other work tools, and it held its product scope even against requests to build forums or simple inquiry-handling features.

The product helps a company write and classify documents and then publish them as a searchable help site or an internal knowledge space. It can run user guides published to customers and working materials visible only to employees, supported by an editor, search, access permissions, change history, and appearance settings. For support staff it becomes material for answering repeated questions, and for documentation staff it becomes a workspace for continuously maintaining information that many people rely on.

Revenue comes from subscription fees paid by companies. As confirmed in September 2026, the Basic plan costs $100 a month and includes one knowledge base and one author, with additional authors at $25 per person per month and additional knowledge bases at $50 each per month. There is no limit on the number of public or private readers who read the documents. It is a pricing model matched to a work structure in which a small number of people manage the documents and many customers or employees read them.

Paid demand expanded after the company became independent. According to the company's official history, revenue and the number of active customers each tripled in 2015, and revenue doubled again in 2016.

Customer reviews were one of the growth paths that spread word of the product. In a 2024 Indie Bites interview Marybeth discussed growth through customer satisfaction and reviews, and the company has a record of winning Capterra's review competition in 2016 against 63 other companies. It is fair to see this as a structure in which the reputation earned through customer support becomes material for new buyers' decisions.

The company also widened its points of contact with writers. Its 2025 materials present Owl Academy, which offers training and workshops, and The Not-Boring Tech Writer, a podcast for technical writers. Alongside selling the product, it ran content that addressed the writing and documentation work customers actually do.

Revenue from Big Customers and Autonomy of a Small Team

The scale Marybeth described in a May 2024 interview was about 450 customers and a team of about 10 people in total. At that time the only people who were both employed and owners were the two cofounders, and the rest of the members worked with the company as contractors. KnowledgeOwl's small-scale operation worked by dividing roles between the cofounders and outside contract staff.

As the customer mix grew, concerns about revenue dependence emerged too. According to Marybeth, the combined fees from about 170 early customers who were kept without a price increase were similar to the fees from about 20 enterprise customers. She was wary of situations in which a big customer's demands steered the product direction or one customer's departure shook cash flow badly. The goal was to cover operating costs with revenue from self-signup monthly subscribers alone, but she acknowledged that at the time revenue from enterprise customers was still necessary.

How a 25-Hour Work Week Took Hold

The 25-hour work week was a solution that came out of the early hiring process. With no outside funding, the company added people when customer revenue could cover the cost, and among the early hiring candidates was a customer who already used the product. When this candidate proposed that sufficient hourly pay and a guarantee of at least 20 hours a week would mean they did not need to take on another job, a collaboration premised on short working hours began.

This approach matched the company's finances and skilled workers' needs at the same time. KnowledgeOwl secured experienced people by setting hourly pay high while limiting the required hours, and members earned a living while gaining time for other activities. According to an explanation Marybeth published in January 2024, the company went on to set 25 hours a week as the standard for its regular roles and stopped hiring for positions that required longer hours.

They also gave people a choice in how the hours were arranged. Someone could organize the time to fit their life, working 4 to 5 hours a day five days a week, or about 6 hours a day four days a week. The same conditions did not, however, immediately apply to the founders, and Marybeth said that even in 2023 she was working about 30 to 40 hours a week. Applying to the founders themselves the working conditions offered to the team remained a problem that still had to be worked out.

The company also set up hiring and onboarding procedures that backed up the short working hours. According to a 2023 account from team member Erica Beyea, the company explained the hiring stages and schedule and sent the main interview questions in advance. After joining, an Asana task board laid out everything from first-day setup tasks to reading documentation to understand the company's direction, and it presented assignments spanning the first six months. Expectations were communicated in writing so new members could know what to learn and when.

The way the team gathered opinions was concrete as well. When a member raised a problem or an idea, a vote chose three topics to discuss at the weekly meeting. The company covered course fees and materials for learning that helped with the work and allowed up to 20% of weekly working hours to be used for learning. The way of operating revealed in Erica's account gave individuals a say over their work while the company provided the time and money needed for shared discussion and learning.

A Company Shaped for Long-Term Operation

The direction of building a company that would run for a long time was there from the early days of independence. An explanation Marybeth sent to prospective customers in 2015 contained the goal of building a service customers needed and stable jobs for members, with no plan to resell the company. In a 2025 retrospective she also stated that the business had been run on money paid by customers, without ever taking outside investment.

In the process of turning this direction into systems and policies, she also pursued B Corp certification. Preparation began in May 2020, the company created policies and held discussions within the team, and it received certification in December 2023. The assessment score listed in B Lab's public records is 80.7, making the company's impact on customers, members, the local community, and the environment a subject of evaluation in how the business is run.

The product expanded to match changing ways of using documentation as well. The 2025 tenth-anniversary announcement presented recently released features such as an AI chatbot that uses semantic search and the analytics tool Owl Analytics. The trend is that finding and answering with the needed information and reviewing how documentation is used were added to the existing features for writing and organizing knowledge.

In the 2025 retrospective Marybeth stated that in the long term she wanted to make KnowledgeOwl a company owned by its members. At that point it was presented as a goal still to be worked out. In a business that began by taking over a product they had built, how to widen ownership to the people who run it together is the next challenge in this case.