Thoughts from Agile India 2025 — the finale
I recently presented at and attended Agile India 2025, the last one.
There has been a precipitous drop in Agile conference attendance from the COVID-19 pandemic which hasn’t recovered.
Beyond Budgeting still makes sense but I don’t really hear about it much in practice. I suspect “business agility” (vs “productivity”) framing is not compelling in the current macro-economic environment.
I’ve been interested in dialogic approaches to transformation but I’m starting to wonder whether emphasising dialogue risks a tendency to focus too much on the “air war” and not enough on the “ground war”. Between “having the right conversations” vs “having the right experiences”, I suspect the latter might be more important.
It occurred to me that Lean Startup won’t make sense in organizations that don’t actually pay attention to impact.
Impact Intelligence looks interesting.
It was interesting to hear what’s actually useful to build a non-self-referential AI B2B product that customers are buying.
The highest leverage cost improvement for tech interviewing is reducing the time required from senior engineers, without reducing the quality of the assessment.
What’s currently the best way to do LLM prompt refactoring?
When there’s an active fire, don’t worry about capability building, put out the fire! When the fire is out, then you can focus on building capability.
There are two aspects of what I consider a “high-performing team”:
- You trust that they care about doing the right thing for the organisation.
- You trust that they have the competence to do things right.
This combination means that you can point the team to a vague, open-ended problem space and trust that they can handle it effectively.
When it comes to “engineering intelligence”, don’t start with tools. Note that the DORA report relies on surveys. Who’s the customer? What are the goals? What questions are you trying to answer? Capture the relevant metrics manually and validate that they’re useful first. And then, and only then, look at automation to help you scale and sustain the effort.
Overemphasis on tools also reminds me of the Taiichi Ohno quote: “Data is of course important in manufacturing, but I place the greatest emphasis on facts.”
Going back to the theme of impact, I’m thinking “product development intelligence” would be more interesting than just “engineering intelligence”.
Innovation = useful novelty at scale.
The product development version of SQDCM (Safety, Quality, Delivery, Cost, Morale) seems to be Delivery, Impact, Health, Cost.
