1. The question moved from "which tool" to "who decides".

In January I was mostly asked what to buy. Now I'm asked who should own it, who signs it off, and what happens when the person who championed it leaves. That's a more mature question and a harder one.

2. Software spend became visible.

Renewal costs have climbed far enough that directors are looking at the bottom line properly for the first time. Once you look, you nearly always find licences nobody opens and two tools doing one job.

3. Vendors started asking about practices.

The interesting shift this year is on the other side of the table. Software companies are increasingly aware they don't understand how architecture practices actually make decisions and that this, not the product, is why the sale stalls.

The bottleneck is still organisational, not technical. The tools are ready. The practices, mostly, aren't - not because people are resistant but because nobody owns the decision, nobody has time to run a proper pilot and nothing survives contact with people who are busy ono a project.

I said this in January. I'd say it harder now.

I thought AI adoption would be high by now. I was wrong, and not by a small margin.

Across the interviews I've run this year with practices of all sizes, the picture is consistent: there is an enormous amount of noise, and very little of it has landed in anyone's actual workflow. People are experimenting. A few individuals are getting real value. But for most of the practices I spoke to, AI is not yet a meaningful part of how work gets done.

That gap between conversation volume and genuine use is the most interesting thing I've found all year. It means the practices moving deliberately now are not late. They're early, and they still have room to do it properly rather than in a panic.

If you're sitting there feeling behind - you're probably not.

A few things I want to flag, because they change what you get from ADDD.

A weekly show reviewing five new tools.

Rather than tools appearing in the database with a one-line description, they'll be reviewed properly on camera each week. The database now sits at around 1,900 tools, and it deserves better than just a list.

An annual State of AEC Technology report.

Published each March. What practices actually run, what they pay, what they've dropped, and how you compare. The benchmark doesn't exist at the moment, which is exactly why the interviews this year have been so revealing.

Fixed prices on everything.

No more bespoke proposals for standard work. You'll be able to see what something costs before you speak to me.

The through-line is the same as always: better decisions come from better information, and there is far less independent information in this market than there should be.

A new website is coming which will cover everything above - including self auditing tool, my half day workshop, software audits, technology strategy and long term support advisory work. There will be a range of options that suits all sizes of practice.

I am really excited about this as it will deliver true value for architects.

That's it for this week - back to normal next time, with new tools, a new job role worth knowing and the usual posts.

One ask while I'm away: if AI has genuinely changed something in how your practice works this year, reply and tell me what. I'm collecting these, and the honest answers are more useful than the confident ones.

See you next week, Thanks for reading!

Allister