What happened to the promised 'regular new features' benefit of switching to new subscription model?

[Paul King]

I seem to recall the big carrot offered to induce us to switch away from dongles to the new and more expensive licensing regime, was that this way we would get regular new features coming through via incremental automatic service patches, without having to wait for a whole year for the full version update. 

It seems that so far we have been getting no new features coming through during the year at all, and the first annual version update 28=>29 was so limited in architect-relevant new functionality that it resembled an incremental service patch far more than a version update.

Operating system used: Windows 11 H2 25

[Mathias Jonathan]

Since it's a now captive market with the subscription model, all I can say is a traduction of a french maxim: "Promises are only binding on those who believe them".

[Jp1138]

What was the last version to include significant modelling upgrades? Archicad 23 with beams and columns and the opening tool?

Later updates seem very minor for a company that must be earning a lot of money... Where is it being invested? AI stuff? What are GS programers doing all day at the office? - No ofense, it´s just that advances are so slow, you would think there are only three poor guys there putting out fires.

[Bricklyne Clarence]

---

@Jp1138 wrote:

What was the last version to include significant modelling upgrades? Archicad 23 with beams and columns and the opening tool?

Later updates seem very minor for a company that must be earning a lot of money... Where is it being invested? AI stuff? What are GS programers doing all day at the office? - No ofense, it´s just that advances are so slow, you would think there are only three poor guys there putting out fires.

---

They don't even do the "AI stuff" themselves. 

Like Cinerender and other similar features before it, it's basically just a clamped on tool developed by other people (the Stable Diffusion model they're implementing), with minimal effort on GS' part in (.....maaaybe) developing a bridge and maintaining said bridge.

Even their own implementation of that AI model into their program is suspect as is,....and as anyone that uses the AI visualizer will tell you - what, with how it creates new geometry out of god-knows-what information it's taking in from the 3D window, as well as just generally badly misinterpreting things you'd think that GS' own intervention would cut down on.

The only things I can think of where the  money from the last 5 versions has gone into, include perhaps the MEP addon/designer (....and the seemingly never ending iterations of them trying to get even THAT working correctly), the quietly forgotten SAM tools, and that Apple VR Vision pro goggles thingie, that I've LITERALLY never seen a person using in real life. What a misfire it was to invest development resources into that venture. Oh, and also DDSCAD.....but I've been informed by some folks that the development for that is separate from Archicad's development and therefore development resources are neither shared nor starved one way or the other.

I imagine they're also spending a penny or two in their project "Aurora" thingamajig, and probably AEC industry seminars and Industry awards shows in Cabo or Las Vegas are not cheap nor free to host or to attend, respectively.

If I as an architect had this level of production and output to show for the fees that my clients were paying me, I'd be starving before too long.

To be (slightly) fair to Graphisoft, in the past several years, every single company who's software I use that's made this switch to subscription-only licensing has seen their output in terms of new tools and development take a nosedive not long after.

You should see how the Trimble folks (Sketchup) and the Vectorworks people are a getting reamed by their users for having had several sub-standard version releases - with scant improvements in any area - over the last few years.

It's what we said would happen with GS if they decided to go this route.

Once you have a client's money already assured to land in your pocket, regardless of how poorly you work, then what's your incentive to actually try and to produce a really great, or better product then before?

[Lingwisyer]

The only people who benifit from a subscription model on something that is not live service, are the shareholders... The statement is often that it provides financial stability so that the company can focus on providing quality updates, but if that was being done in the first place, many people would be paying for upgrades, or future potential improvements that they trust will come, anyway... with there still being an incentive for the developer to produce meaningful updates.

[Bricklyne Clarence]

Well they (the shareholders) may be benefitting in the short term, but they'll be the ones losing out big in the long term as they continue to shed long-standing loyal users.

Taking a look at the Nemetschek stock price performance over the last several months, that "long-term" may not be as long as they think it will be.

And you're right.

If Graphisoft had been handling the program much better and treating their customers far better than they currently are, this need to switch to subscription-only licensing model as a means of securing financial stability wouldn't be necessary.

They're being forced to go this route because majority of users long-stopped upgrading to every new version they released because they were not seeing any appreciable return on their investment in terms of new tool improvements  they need.

And in response now they (GS) are resorting to a model that's guaranteed to drive away even more customers, as they head down a road that will see them improve the program (or have the incentive to do so) even less than they're doing now. 

[Lingwisyer]

Slow and behind closed doors.

[Claudiu Serban]

It's definitely not a programming problem (tho it kinda is in the whole industry).

The problem stems from the Nemetchek behemoth trying to suck all the money possible from their customers.

Firstly, they went on aggressive acquiring of new companies and now they have like 3 Architectural solutions (Allplan, Vectorworks and Archicad), each with their own market/customer base.

Then they want to somehow upsell their other solutions to these customers. When it came to arhicad, they tried to facilitate a direct connection with Scia, which held archicad back for like 5 years, cause probably there are 2 people in the whole world who use AC structural analysis tools. Then they bought DDSCAD, and tried to upsell MEP/HVAC to engineers working with Archicad architects, using IFC interoperability and MEP Modeler. This failed as well, so now they gave Graphisoft DDS and they spewed the half baked MVP called MEP Designer.

Its all so easy, but they can't see that Revit is winning BECAUSE its ONE PRODUCT. The big firms dont want to spend time with hotlinking IFCs, updating interoperability addons, display issues, etc. They can have 1 license per seat, and that seat can switch between architect, structural or MEP/HVAC engineer. Not only that, but the addon market is much more appealing, and thats why Magicad is the golden standard when it comes to MEP/HVAC design/drafting.

So basically, they keep trying new stuff to grow their numbers, nothing works, they leave new features half baked to start even newer features that will be abandoned half baked in a couple years as well.

[jl_lt]

To the revit mep and structure "it's one product argument"  I would add:

It's one product that has more than 20 years of a head start, with massive resources poured on it, marketing campaigns and user-base penetration that starts even before kids enter architectural school, and an extremely high cost of software switching which still would be high even if graphisoft had a competitive alternative.

One of my structural engineers became mildly interested in archicad when he saw me using it; I immediately fantasized of a nice little world in which I could interact with said engineer in the same software ecosystem. 

So I asked in these very forum, eyes filled with hope, it they had promotional materials about archicad structural capabilities that I could show to said engineer and close the deal. The answer: no, we don't have any of that.

World shattered 

[Wlad222]

Unfortunately, Archicad is not improving at the pace many of us would like. While the software remains a strong BIM platform, progress in key areas such as performance, interoperability, visualization, automation, and modern workflows often feels too slow compared to the speed at which the industry is changing.

[Holger Kreienbrink]

I understand that you don´t see the benefit now. But this will change this year: Forma Connection, MCP Server, Design Intelligence Platform, Open Collaboration Platform just to a few ones. They will be in your hands soon. 

[Karl Ottenstein]

Users really need fundamental things that they have been waiting decades for, particularly:
https://community.graphisoft.com/t5/General-discussions/Central-Attribute-management-Roadmap/m-p/639...

[runxel]

Holger, with all due respect, most oft these things nobody has ever asked for – maybe you remember the SAF debacle. Your statement is hardly the "gotcha" you think it is. Bringing this up is rather tone-deaf and weakens your standpoint with your user base massively. I hope you understand that.

[Holger Kreienbrink]

ok , let me explain

MCP is the connection to all kind of automation tools:

you want to create your own library parts, cause the library content sucks - MCP enables AI to create custom library parts

you want to have clean dimensions, automatic dimensions, etc - MCP enables AI to connect to Archicad to do this

you want Archicad automatically exchange wall types - MCP will get you to that solution. 
And many things more, we have not really thought about. 

You are looking for a special command or route - with AI and MCP you can vibe code it. 

MCP is one of the most important technologies for the future. 

Open Collaboration Platform targets all the data exchange you have in every project. You are the only architect with Archicad and all other engineers  are using Revit. And IFC is....not the preferred solution. OCP will connect you to the engineers. 

Or your engineers are using Allplan. Same thing. Without the need to export and import IFC. 

And Design Intelligence platform targets the design phase. It will help you to decide on data when designing. On top of your knowledge it will help in finding the best solution for your design and for story telling to your client. 

AND: We are also targeting a lot of smaller, daily problems in Archicad. 

I know about SAF. A good idea, a good technology. But the reality is different. Accepted. 

[Bricklyne Clarence]

---

@Holger Kreienbrink wrote:

ok , let me explain

MCP is the connection to all kind of automation tools:

you want to create your own library parts, cause the library content sucks - MCP enables AI to create custom library parts

you want to have clean dimensions, automatic dimensions, etc - MCP enables AI to connect to Archicad to do this

you want Archicad automatically exchange wall types - MCP will get you to that solution. 
And many things more, we have not really thought about. 

You are looking for a special command or route - with AI and MCP you can vibe code it. 

MCP is one of the most important technologies for the future. 

Open Collaboration Platform targets all the data exchange you have in every project. You are the only architect with Archicad and all other engineers  are using Revit. And IFC is....not the preferred solution. OCP will connect you to the engineers. 

Or your engineers are using Allplan. Same thing. Without the need to export and import IFC. 

And Design Intelligence platform targets the design phase. It will help you to decide on data when designing. On top of your knowledge it will help in finding the best solution for your design and for story telling to your client. 

AND: We are also targeting a lot of smaller, daily problems in Archicad. 

I know about SAF. A good idea, a good technology. But the reality is different. Accepted. 

---

The big problem for you guys is that these "smaller, daily problems in Archicad" are actually the things that your users would have preferred you were spending a significant amount of your development resources and time addressing.

'smaller, daily problems...' ...typically tend to pile up and add up to bigger (billable) man-hours wasted on clunky workarounds and infuriating wall-headbutting roadblocks.

Those big ticket, big picture items you  mentioned are great for the future outlook, I'm sure....and they will definitely make for great marketing brochure/pamphlet material for the sales guys.
But these are not the things we automatically gravitate to, to help make our workflow better.

And the problem is that when they don't take with users, with a particular version release, we all know what happens next in that they get left behind, abandoned and forgotten.
And as a user suddenly you're left (STILL) without the actual tools you wanted addressed,....while at the same time these other new tools you never asked for, and which you guys spent a lot of time developing, suddenly also getting left by the wayside leaving everyone with nothing.

Ultimately all leading to what feels like a whole big waste of a version.

And that's what a lot of the past several versions have felt for a lot of us - particularly if you weren't clamoring for Design Option,......or.......um.......yeah, .......that's about it.... .....over the last several versions,.... in terms of architecture-focused design tool improvements.....(since at least version 23) - if you weren't clamoring nor asking for that, then the last 5 or 6 versions have been more or less a waste as far as most of us have been concerned.

The last MEP/SAM/DDASCAD and AI-focused versions.

[Jp1138]

This is exactly what this whole thread is about 😓: there´s a huge disconnect between the users and the direction of the software - or at least the most vocal users.

I´m all for MCP, but the program has been stale since before Chatgpt took the world by storm, so this doesn´t add up. And GS should have enough resources to add many more little things every year while developing bigger things at the same time.

When Rotate objects with a keystroke (very useful as it is) is considered a mayor improvement for a version change, something is not working. That should be the small print of others improvements, not the main event.

[Jp1138]

With all due respect, all those things are great, but very far from the use of the program a great deal of your users do. And if the main tool isn´t working correctly (Archicad), then we have nothing to share with nobody on those magnificent platforms or the AI won´t be able to resolve things the program can´t do.

GS has left behind the small and medium firms and bets everything to be the new revit, it appears. I hope I´m proven wrong soon 😓

[Patrick M]

my understanding was that there are/were features that were being locked out by bimcloud basic (primarily the useless AI tools).
only with the subscription + bimCloud would you get these features. 
...
I don't know about 'regular new features'... maybe they meant 'regular, new features'?

[Lingwisyer]

Has that not always been the case? The difference being that the self hosted differences were all to do with management, which you can easily do without even on a medium scale. Being locked out of AI tools and the main subscription bimCloud features is not surprising as that entails hosting and compute costs...

[Mathias Jonathan]

Yes, Design Options are really important in our actual workflow! This is the kind of game changing tool we need.

To be honest, I feel a little torn. I’d almost like to ask for more modern tools like new doors and windows, but given developments like MEP (from what I’ve read on the forum, we don’t use that part of ArchiCAD) and the time it took for it to start meeting user needs, I’m wary of such major changes without a significant investment from Graphisoft—and therefore a strategic commitment. And given that feedback on the unfinished nature of new features has often halted development (Dynamo, if you’re listening!), I’m not sure if we’ll ever see this kind of “new feature” or if it’s even desirable given the current development model.

[Bricklyne Clarence]

This is indeed the conundrum that Graphisoft puts its users in.

We want them to implement new features and the improvements we need, to make the software better for our use.

But given their recent history in badly handling much needed improvements, releasing them half-baked or half-finished, and then in a lot of cases, completely abandoning revisiting those improvements to fix the bugs and address the issues brought up once users actually put those new tools to use (or doing so, super-belatedly*), no one has any faith in them anymore that they'll handle any new improvement with any higher degree of proficiency, concern and the responsiveness really needed to improve the program.

(*in some case and tools, abandoning new tools for as many as 10 years or 10 versions apart  (or longer) from the release to when they refresh it to address issues. In fact, 10 years is about the average time they take for a lot of the major tools  -> Curtain wall (released at v12; refreshed at v22), Complex Profile (released at v10; refreshed at v23) Stair and Railing tool (.....Still waiting for a refresh...), Door, Window, Slab...(never improved...) etc).)

But the thing here is that we DO need them to improve some of these abandoned tools and/or develop new tool features and functionalities that are badly lacking - despite their poor recent history and record in doing so.

More than anything they need to improve their development model and methodology.

They  need a better user feedback acquisition and addressing system than they currently have and are relying on.

They're badly lacking in better information as to what people need improved (see : This thread ....and the utter disconnect that still exists between what they think we'll get excited about that they're working on, and how the actual response plays out when they pull back the curtain. And this isn't even the first time this kind of reveal and response disconnect is happening on this forum.)

And most importantly (....I think), they SERIOUSLY need to develop thicker skin to accept criticism in good grace and the constructive spirit in which most people intend it.

I think they've cocooned themselves so deep in their silo over there at GSHQ for so long, ....and with a kind of 'Fallout shelter' mentality, where they don't listen to, nor accept anything they perceive as negative coming back to them in the form of feedback from users, and only read their good press.

And now we're seeing the long-term destructive results of relying on such a blinkered echo-chamber strategy for so long a time.

Continue down this path and an unstable revenue stream forcing them into an unpalatable and unlikeable licensing model will be the least of their worries.