By Florian Plank

How we plan for GenAI—task by task

Why Task-Based Analysis is our most pragmatic tool for a generative future

Data & AI

NoA Ignite — The digital consultancy that delivers

At the moment, the generative AI boom is impossible to ignore. New tools launch daily. Business leaders scramble to respond. Some see opportunity. Others brace for disruption. And yet many are still asking the same basic question: Where do we even begin?

At NoA Ignite, we’ve developed a practical answer—one we’ve already put into practice across a range of projects. Our Usable AI™ framework is designed to help teams move beyond experimentation and into execution. Not by focusing on models or features, but on human needs and business value.

One of the most effective tools within that framework is Task-Based Analysis (TBA). It’s not theory. It’s not tied to a specific model or tech stack. It’s a structured, bottom-up method that helps organizations understand how AI can improve the way work gets done.

If AI is going to transform your business, it will do so task by task. Not all at once.

Here’s how the process works:

1.IDENTIFY KEY ROLES

Start by selecting three to five of the most critical roles in your business. These could be internal or external—employees, suppliers, even customers. If you’re unsure, run a quick prioritization exercise with your team.

2. DEFINE KEY TASKS

For each role, list the tasks that make up their day-to-day work. We use a “human toolbox” of 10–15 common task types, which are then tailored to fit the role. “Writing” becomes “writing copy for ads” or “writing code.”

3. RANK BY REPETITION

Place each task on a canvas, ranking it along a horizontal axis based on how repetitive it is. More repetitive tasks go to the right; more variable or judgment-based tasks stay to the left.

4. RANK BY VALUE

Now sort the same tasks vertically by their business value. Define what “value” means for your organization—efficiency, customer impact, revenue, or something else.

5. MAP AI POTENTIAL

Using a “genAI toolbox,” match each task with a generative AI capability. Add these AI-enabled versions to the canvas using a different color. Draw arrows to show how tasks shift with AI support.

6. ANALYZE THE FUTURE ROLE

Step back and look at the full picture. What does this role become when AI enters the workflow? Which tasks are ready for automation or augmentation? What needs to remain as it is, and what can be reimagined?

This process doesn’t take months. It gives you something tangible in hours—a shared understanding of where AI fits, what it can improve, and where to begin.

Pretty much all knowledge workers can benefit from using GenAI now, but most will need training. Task-based analysis of jobs is helping businesses identify opportunities.

— Andrew Ng

WEF 2024

Start where it counts

Building a pragmatic AI strategy? Begin with the task.

Curious about what TBA could unlock for your team? Let’s talk.

Florian Plank: Head of AI; (+49)17682465930|florian.plank@noaignite.com

Related insights

See all insights

Article

The future of GenAI is Usable AI

Data & AI

Article

Designing AI That Works for People

Data & AI

Article

Connecting the ocean’s boldest ideas with those who can bring them to life

Data & AI,

Article

Award-winning data in motion

Sustainability & Impact

Article

Fast Company names NoA Ignite one of the world’s most innovative companies

People & Culture

Facing complexity? We’re ready to listen.

Big idea? Strategic challenge? Emerging vision? Whether you’re crystal clear or still shaping the path, we’re ready when you are. Leave your details, and we’ll get back to you shortly.

How we plan for GenAI—task by taskHow we plan for GenAI—task by task

Start where it counts

Related insights

The future of GenAI is Usable AI

Designing AI That Works for People

Connecting the ocean’s boldest ideas with those who can bring them to life

Award-winning data in motion

Fast Company names NoA Ignite one of the world’s most innovative companies

Facing complexity? We’re ready to listen.

How we plan for GenAI—task by task