Safetybelts not needed

We don’t need safetybelts in our cars.

It’s add costs and customers don’t ask for it. It’s a very rare to crash the car if you are an experienced driver. 

Do you think this sounds stupid?

If you resonate like some agile product teams, this is what you get as an answer for there software products.

Nobody asked for it. 

It’s an exception, we only do the happy flow in our MVP.

We can add it later.

Would you like to drive a car developed by a product team who took the same aporoach to car safety?

But safetybelts are mandatory in today’s car may be your objection to this.

Privacy by design is also mandatory for new development. Old software is like vintage cars without safety belts. Use on your own risk.

Who is responsible in your organisation to assure that you don’t miss critical requirements?

Endless freedom

In the best of worlds, each product team would have an endless freedom to manage all decisions about their product. Each team would be 100% autonomous in the planning and development of their features.

Development times would be much faster as each team wouldn’t have dependencies to other team. Each product team would then be small so they could be superefficient and avoid management overhead.

There would be no need for enterprise architects as the organisations architecture would be a blank paper, neither a program office as prioritization is not needed.

But in this best of worlds, the business mode canvas for the company need to be simple. Few value propositions, limited customer segments, one or two channels. 

Source: Wikipedia

Source: Wikipedia

If you have a complex business model, then you need enterprise architecture and a firm governance of the produkt teams so that the things they build works together, by purpose, not by chance.

Pizza as a service is flawed

The concept of pizza as a service is flawed, but not in the way you think.

IMG_2276.jpeg

If you seen Lucifer on TV, you know that he asks the suspect what he desire the most, and this is the question we must answer when doing business development.

As a customer, what do I truly want? Is it pizza as a service ot something else I want?

When I was 18 on Sweden, the only way to get beer or wine legally was to drink it at a restaurant. But regulations in those days were that you needed to order food to get alcoholic beverages.

This is why we had half a pizza and several large beers at the local pizzeria before going to the next place.

Today, my main reason for getting a pizza is when we are at the stable and don’t have access to a kitchen or time to prepare a meal.

Another classic reason for ordering a pizza is when you have a hangover and want lots of calories. Typical example in Sweden is New Years day that is peak pizza.

What are then the most important capabilities for the local pizzeria then. First of all, it’s the location. Second, if you serve pizza at the restaurant, you need to be able to provide alcoholic beverages.

That you have pizza as a service is less important, more important is to cook food with cheap ingredienses that could be served fast, both at the restaurant or to take-away.

If you want to share a moment with friends and family, pizza as a service is not highest priority unless you have a customer segment that crave for pizza.

Belief, fact resistance and analytic capabilities

As an Enterprise Architect, one of the skills you need is an analytic capability.

However, as seen the last weeks, this is a missing skill from a number of architects and self proclaimed analysts.

If you dissolve or reject information that doesn’t fit with your beliefs, there is huge risk that you draw the wrong conclusions and suggest wrong actions.

IMHO, with that mindset, you are not fit to be an architect or an analyst, and shouldn’t get at job in that role.

Business automation and creativity

I would like to challenge Tom Graves statement that Not every process could be automated.

Can you automate a creative process like writing a manuscript?

As a business process it’s fairly well described on a high level, but is that enough for automation?

First, if we would could feed a logline into to process, would we get out a first draft for a feature film? 

Second, if we feed the same logline into the process one more time, would we have the same result?

Third, if we gave feedback to the process about what we fought about the first draft, would the process adapt and rewrite the story based on experieces it gathered?

With AI we already have an engine that could write stories, based on the input.

The questions are then if we get the same story twice and if the engine learns from feedback.

If the AI-engine repeats history and delivers the same story from the same logline, then it’s a repetitive process. 

If it’s a different story each time, but doesn’t learn from feedback, I would more categorize it as a randomizer, probably worse than a blind hen.

But if the engine learned from feedback, it would be much smarter, more akin of a very skilled person 

So, yes in the future, all processes could be automate. The question is if they should and it is more of an ethical question than technical.