Last month, a group of predominantly British architecture studios penned an open letter to Autodesk criticising its poor performance over recent years. The group, including Zaha Hadid Architects, Grimshaw, Rogers Stirk Harbour and Partners and other prominent studios, condemned the increasing cost of ownership of Revit, its constantly changing licensing arrangements, and failure to keep pace with product development.
I didn’t pay much attention at first. Our studio doesn’t use Revit, open letters always strike me as the Chihuahua of lobbying activities (all bark, no bite) and frankly the world has bigger problems on its hands right now.
But my interest was piqued last week when Andrew Anagnost, the CEO of Autodesk, responded with a blog post comprising equal parts apology and rebuttal.
The apology is almost instantly forgettable: platitudes on product development, proclamations of customer care, promises to do better. Classic corporate speak. In contrast, the rebuttal is much more exciting. Most memorably it includes the rather snarky claim that for the seventeen original signatories to the open letter, annual Revit licensing fees represented just 0.63% of their 2019 revenues.
This claim instantly struck me as dubious:
- In the UK, an annual Revit subscription costs £2,796.
- ZHA, of whose 600 employees let’s say 500 are Revit-using architects, would have an annual Revit bill of £1.4m and should therefore be able to generate annual revenue of £222m.
It turns out my suspicions were correct, because in 2019 ZHA actually generated revenue of only £56m. Surely Anagnost got his decimal point in the wrong place, but even if he did my guess is it wouldn’t have altered his essential conclusion that Autodesk software does not represent a substantial cost for architecture businesses.
Now I should point out that while I’m hardly rooting for Autodesk, I’m not necessarily cheering on ZHA and co. either. Yes, architects are my people, but it’s famously hypocritical of Patrick Schumacher to complain about the cost of software while simultaneously championing unpaid architecture internships. Indeed, the seventeen signatories to the open letter are all Big Business and I am sure perfectly capable of negotiating some sort of enterprise bargain with Autodesk.
Instead, I’m curious to examine Anagnost’s claim that neither software generally nor his software specifically are substantial costs to architecture businesses. And instead of mega British studios bristling with employees, I’m going to do so within the context of small architecture businesses here in Australia.
- What percentage of revenue does the cost of software actually represent?
- And how does the cost of Revit compare to other essential software platforms on which architecture businesses depend?
For my analysis, I’m going to use a fictitious architecture business called Architects McArchitectsFace. This very serious business employs five people and enjoys annual revenue of $650,000. It comprises a sole director and four architects, each of whom uses every software service to which the business subscribes (i.e. they require five licenses for everything). It pays for real software packages at current prices.
First up, here’s the software list to which Architects McArchitectsFace subscribes:
- Documentation: Revit
- Presentation: Adobe Creative Suite
- Business management: Streamtime
- Rendering: Vray
- Design: SketchUp
- File management: Dropbox
- Email, word processing and spreadsheets: Microsoft 365
- Task management: Trello
- Accounting: Xero
- Communication: Slack
- Video conferencing: Zoom
And here is its annual software bill (in Australian dollars):
- Revit = 5x $3,410 = $17,050
- Adobe Creative Suite = 5x $924 = $4,620
- Streamtime = 5x $720 = $3,600
- Vray = 5x $480 = $2,400
- SketchUp = 5x $410 = $2,050
- Dropbox = 5x $210 = $1,050
- Microsoft 365 = 5x $206 = $1,030
- Trello = 5x $165 = $825
- Xero = $780
- Slack = 5x $110 = $550
- Zoom = $350
Which comes to a grand total of $34,305 or 5.3% of Architects McArchitectsFace’s $650,000 annual revenue. So a significant minority, far more than Anagnost’s claim of 0.63% and twice even the 2.5% I estimated above for ZHA.
Thus, I argue that software does indeed account for a substantial portion of the revenue of a small architecture business. And perhaps as well it should. The packages listed above drive a large chunk of an architecture business’s daily activities, from design and documentation to communication, accounting and file management. These things are just as important as the benefits obtained from paying for rent or office equipment or insurance. Which for Architects McArchitectsFace aren’t as expensive as its software:
- Rent = 4.4% = $28,600
- Office equipment = 3.2% = $20,800
- Insurance = 2.0% = $13,000
I’ll leave it up to you to decide whether the above software packages are indeed more valuable than the computers on which they run. I guess these days they’re certainly more valuable than the office in which the computers sit.
But regardless, perhaps most poignant is that Mr. Anagnost’s software on its own equals the cost of all the other software packages combined. It turns out that Revit is in fact really expensive. And it turns out that on this issue at least, I’m with Schumacher and his mates after all.
- Matt Hickman; Leading architecture firms pen open letter to Autodesk over rising costs, sluggish development; The Architect’s Newspaper; July 2020.
- Andrew Anagnost; Autodesk and the Architecture Industry; Autodesk; August 2020.
- Buy Revit; Autodesk; accessed August 2020. Curiously, the subscription price in the UK is actually higher than it is in Australia: $3,410.
- Zaha Hadid Architects; Growjo; accessed August 2020.
- Jim Dunton; Profit dips at ZHA despite record turnover; Building Design; April 2020.
- Patrik Schumacher; Brexit: a chance to roll back the interventionist state and unleash entrepreneurial creativity; Archinect; July 2016. For more on Schumacher and unpaid internships, see previous Panfilo article, Redefining Success.
- This is based on an employee to revenue ratio of $130,000, which was the 2019 average for architecture businesses with annual revenue under $700,000. Source: Benchmarking Data and Research.
- This in subsequent figures are based on 2019 averages for architecture businesses with annual revenue under $700,000. Source: Benchmarking Data and Research.
- Revit logo; image sourced from Logo Lynx.
- Software costs; author’s own image.
Why on Earth you propose that combination of software? Office365 is $19.95/month, and it includes Teams (a better communication platform than Slack), Teams again (a better conferencing solutions than Zoom), OneDrive (an easier to use file management than DropBox), and Planner, which is the same as Trello. Save yourself $2,500 and use better tools with your Office365.
Thanks for the suggestion Meee, I’ll pass it on to Architects McArchitectsFace!
What I took away from the open letter was more to do with the cost of Revit going up, but very little improvement to the database and functionality. I work on the engineering side and have to admit I am not familiar with the changes to the architectural functions. However, I would agree that the changes that have been made are minor or not very useful. I believe that all users have been begging from the start to have a better text editor. They improved the stability of the text, but is it really that hard to incorporate a proper text editor? If AutoDesk were to make updates and changes to the software that resulted in better project or document management and increased efficiency there would not be a concern of the increased cost.
I’ve always thought of the Autodesk suite as being classic examples of monolithic software with decades of mostly defunct legacy code weighing the whole thing down. I bet a proper text editor would involve undoing 20+ years of patches and hacks and fixes!
Good writing. Enjoyed reading it. 😊
Another perspective to this is how Autodesk products compare to the market alternatives. They might cost more, but on the innovation side everyone is pretty much the same.
I think firms do have choices, but at the same time need to re-examine their own business model in this fast changing world.
Yeah, it does feel a bit like you’re damned if you do damned if you don’t. I don’t use SketchUp Layout as a documentation tool, but I’ve always been impressed with how elegant SketchUp is as a software environment: it does one thing really well at a decent price… like all good software should. Imagine if it were scaled into a proper BIM tool!? That would be a real alternative.
I think in your example the cost of Revit Suite was just over 2%. That maybe what Autodesk is representing.
I think there is a whole discussion on the cost of production for any project (something the AE industry is still grappling with) and learning investment. The customization side of the conversation is equally as challenging as the software true cost of operations.
As the others have commented, more focused development of communication features like a text editor or text that can be aligned to a specific point would be nice! Even a lowly program like AutoCAD can do that!
Very good read,
I hardly believe that Zaha Hadid have 500 Revit single subscriptions in use. Your assumption are wrong there in the numbers of revit and also the cost per unit. Big companys have special contracts with autodesk so you pay less money and you also buy network licenses for example you buy a package with 100 licences which can be shared between employees. Why i know this i work in a big company here in switzerland we currently brought a 100 license. And for a little more bucks per year you not only get revit you can get the AEC Suite including 3dsmax, civil 3d, navisworks, infraworks, autocad.
Yeah, I made some big assumptions about real figures. I’m sure ZHA use a network license for their office. Of course, whether it’s 100 or 200 or 300 seats does make a big difference to the cost analysis.
Your example is good idea but i think a business of 5 full time employee would need to turn over a lot more revenue to stay a float.
Most people vastly underestimate expenses.
Largest expense is wages.
Average wage in Australia is 75-80k.
5 people will cost 400k
On top of this you will have rent, hardware software, book keeper/accountants wages,
Cost of sales, travel, phone, junior assistant
and various taxes.
Company will likely need to be closer 1 million plus revenue or would go broke.
Higher revenue would mean a smaller spend on software.
Real cost here is opportunity cost.
Better software means faster work, more projects, more revenue, less employees, less wages
Sadly, architecture is rarely that lucrative. As I mentioned in the article’s footnotes, average revenue per FTE for small architecture businesses in Australia is $130k.
Also, senior architect employees at that scale might be paid at the $75-80k mark but more typical would be $60-70k.
Sorry but you’ve got your figures very wrong here.
I think you might have glossed past this part, but if you look at the original open letter, they cite very early that all 17 signatures spend $22m over 5 years with Autodesk, that was their way of chest puffing which is cool, fair enough.
That works out at a total of $4.4m per year, and an average of $258k per signature (including ZHA), which averages at 80-90 licenses per firm at 3k per license. These guys are tiny as accounts to Autodesk and would need to be 4-5 times bigger to qualify for talks as enterprise accounts.
They’re not, but if ZHA were the odd one out and they did amount for 1.7 million of that annual 4.4 million, then that puts those other 16 firms at an average of 45 licenses per firm, which would be 135k annual spend with Autodesk. I’m not about to go looking up the revenues of all 16 other firms, but Andrew did, he really did not get the figures wrong, we are talking sub 1% percentages to revenues around 0.2-0.6%.
I think you went wrong at the “500 out of 600 staff would be revit users” part, that seems wildly out of proportion. I don’t profess to know anything about them but online reports put ZHA at around roughly 400 staff, of which typically in engineering for example roughly 100 would be design software users with the remaining being admin, sales, support, procurement, management, directors, team leaders, HR etc etc and they would be teams themselves at that level and non design software users. But honestly that all doesn’t matter because ZHA just don’t amount for 40% of the stated spend of all 17 signatures on the open letter.
Look I know the software is at a high price point and I know it has development stagnation issues, now isn’t the time to go off into why and comparisons with Adobe/MS etc but the figures remain as they are, the software really is sub 1% of revenues.
Thanks for your thoughtful input Neil, I acknowledge my calculations wouldn’t necessarily survive a Royal commission.
As I understand it though, even your calculations aren’t entirely accurate because Anagnost used USD for his calculations, where the software is substantially cheaper than in the UK. Rather than debate the point, I think it’s enough to recognise that he and everyone else is spinning the same mostly opaque data set to serve there own purposes.
All that aside, the comparison to other software types is entirely reliable and based off current recommended retail prices. And in this instance at least, Revit is substantially more expensive.
Now, it might not be helpful to compare Revit with some of the other smaller packages on my list, but I believe the Adobe comparison is useful because as a suite, it is also a monolithic one-stop-shop for a number of creative industries. For Photoshop et. al. to cost 25% of Revit strikes me as imbalanced. What are your thoughts on this?
I’ve referred to Autodesk as the Evil Empire for ages. They aquire more and more software programs and: 1) Significantly mark up the price; 2) Gut it for a functionality missing somewhere else and discontinue it; 3) Just scrap it entirely so it represents no competition for anything in their line.
Also, at this point, they’re are so much about animation that I think the architecture side of things had before the red-headed step child of their programs.
But Revit is what we are stuck with. To bad the original company sold out to Autodesk.