
Blueprint for Building AI Agents That Perform Real Jobs
Documentation as a marketing tool and how to write useful documentation that promotes technology products to both humans and machines.
In this article, we will examine documentation as a marketing tool and how to write useful documentation that promotes technology products to both humans and machines.
Documentation is often treated as an obligation rather than an integral part of the delivery process by both developers and product managers. Everyone agrees that good documentation is important for others to understand how a complex system works, but it is rare for documentation to be embraced and promoted in the way it deserves.
Writing good documentation and promoting it publicly provides not just one, but at least three benefits:
Consider a practical example in Norwegian. A Google search in Norwegian for how to record the financial transactions of a Christmas party will return a guide on how to perform this exact task in the accounting software Fiken:
Using an AI service like Perplexity also provides a banner for the Fiken documentation above the answer:
Traditionally, developers do not enjoy writing documentation. It is often a checklist item in the “definition of done” that is required as part of developing a new feature.
When the documentation is completed, it is often forgotten in an internal documentation repository, and the only time the documentation is used is when a new developer is onboarded to the project or when some external consultants are hired to conduct a software audit.
Promoting documentation as a first-class citizen, as important as the feature itself, not only helps to motivate developers to write the documentation but also serves as a goldmine for the marketing department.
To achieve quality documentation, at least 10% of overall resources should be allocated to its development.
Stripe is a company that is celebrated for its documentation and serves as a good case study for how to write and maintain exceptional documentation.
Without insider knowledge, it is safe to say that Stripe’s documentation is not only created by developers and technical writers; there is a whole team behind the production to make it accessible and relevant.
Key takeaways that any company can learn from Stripe and Fiken include:
Involving developers in the documentation process allows them to evaluate the features they create from a user’s perspective.
Developers often develop tunnel vision when working on a feature, and writing documentation provides an opportunity to audit their work, helping them identify shortcomings they may have missed during development.
Much has been written about how, in the future, most users will be machines and how AI is stalling the adoption of new technology.
In a world increasingly driven by AI, documentation is no longer just for user convenience. AI agents are constantly scanning and analyzing online content. Comprehensive and well-structured documentation provides excellent training data for AI models, making your product easily understandable and accessible to these AI systems.
This improved understanding translates directly into better search engine rankings, more effective AI-powered recommendations, and ultimately, increased product visibility.
AI is powerful, but requires data to learn. When new features are introduced, AI tools can be slow to recognize them due to limited training data, as nobody has yet used the feature. Detailed documentation acts as a reliable source of information for both human users and AI systems, enabling the AI tools to incorporate the new features into their suggestion models.
Want to learn more? Contact us today!
Blueprint for Building AI Agents That Perform Real Jobs
6 inspiring IT staff augmentation success stories: How companies scale fast with nearshore talent