How can non-tech companies (e.g. traditional manufacturers) build great digital products on the first try?
There are hundred methods out there! Yes, but most "digital product" approaches are driven by tech-first companies (e.g. Google, Facebook, Tesla). This works well - but companies that have never built their own software are doing it now and its painful learning process for them.
I explored a few ideas - from simple interviews to working prototypes - in order to solve this. Feel free to contact me for more details.
Non-Tech Companies can't build software
From the perspective of a non-tech company, building software is expensive and the result inferior. Why?
- Most "Product Owners" don't have good product or technical skills. They're average project managers and the organisation doesn't appreciate agile in higher-up levels. A Feature Factory appears.
- The tools used (Jira, DevOps, …) are too complex and focus on the tech too much. Further tools (Miro, Product-Software) will just make the collaboration more painful.
- The developers and business people only collaborate on paper. Teams need to align over time, mixing their skillsets. The "handover" or any collaboration only scratches the surface.
- Finding and paying the right people is hard. Teams need to level up, external contractors are used and the organisation is already grown-up around non-software people.
I've explored two concepts in order to solve this:
Solution 1: Using AI for requirements
- An AI assistant guides the whole team - from Product Owners, to developers and stakeholders.
- The AI is trained on past requirements, agendas and common known product strategies.
- The AI can shape questions around product strategy and moderate this process
- The AI can breakdown complex requirements into workable tasks.
- The AI helps with moderation and translates requirements between the roles.
This is already possible. I've extended GPT-3 with custom weights from written requirements and with pre-defined prompts to replace myself as a consultant. It's especially great to keep a structure within a team over a longer period of time.
Solution 2: Software defined project management
Modern cloud infrastructures are setup with code. This means that an engineer writes scripts that for example start a database server in azure in an automated and repeatable way.
The idea is to transfer this concept into project management. We write a script that defines how a project management software should work. For example, we define the states of a task, which fields are required and how the teams works together. This would work across project management vendors.
With this central definition we can do two things things
- Organisations can easily define how projects work across tools
- We can render the projects in a unified interface making it easier for every participant to join a project
A developer might only want to see a few fields and details like all tasks in the current sprint. A product owner has a different mindset so he would render the project based on the business requirements. All these definition are still defined in code.
External contractors or other departments could even have "their" own templates so that they can just map the definition to their process and onboard new people faster.