Return of the bullshit – baked beans edition

In an October 2015 post I’ve only just noticed, snappily titled No More Software Like a Can of Baked Beans: Why Software Subscription Serves It Up Fresh, Autodesk VP Andrew Anagnost bravely attempts to sell Autodesk’s move to all-rental software. This is a rather belated response, but fortunately there is no statute of limitations on skewering spin so let’s get started.

How does he go? On a positive note, top marks for creative writing! The general theme is a strained and somewhat Californian analogy in which perpetual licenses are like canned goods (bad), and rental is like fresh produce (good). However, it’s presented well and professionally written. Among the highlights are:

  • Perpetual software licenses are like high-fructose corn syrup – no, I’m not making this up. Stop laughing at the back there!
  • This is a change that is simply a better experience for everyone – everyone who likes the experience of paying more for less, that is.
  • It’s to create a better product, something tailored to customers – creating a better product seems beyond Autodesk, at least where AutoCAD is concerned. Actually, it’s to create a more expensive product. Tailoring is something we customers been doing for over 30 years without the use of rental software, thanks.
  • There will be less disruption – except a) how we pay for the product is independent of how/when the product is updated and the disruptions inherent in that, and b) even ignoring the erroneous conflation, it’s a mistake to assume that continuous updates are less disruptive. Recent history proves otherwise.
  • Companies (e.g. Autodesk) will work even harder to keep you happy as a rental customer – history gives the lie to this one, too; the closer Autodesk has got to this model and the more people have been locked into annual subscription/maintenance payments, the worse the value for money has become. It also ignores the various alternative ways Autodesk will use to try to keep you tied in. What do you think all that Cloud investment has been for?
  • Autodesk is focusing on helping customers succeed with its products and services – I don’t think so. Autodesk is focusing on trying to keep its shareholders happy.
  • Serial numbers are a terrible dehumanizing thing, rental will make them go away and relying on Autodesk’s internet expertise for Cloud-based licensing is a much more attractive proposition – serial numbers are fine, that’s just silly. There are a host of unnecessary problems introduced by Cloud-based licensing, even when dealing with companies that aren’t as crap at the Internet as Autodesk (e.g. the Redshift site won’t even let me scroll back up once I’ve scrolled past the end of the post). The idea of Autodesk disposing of serial numbers and implementing a phone-home scheme instead is pretty terrifying, and I can only hope that technical issues prevent it from ever reaching production. Mind you, the fact that some new thing is clearly unfinished to the point of uselessness doesn’t seem to prevent Autodesk releasing it these days, so who knows? Hmm, I feel another post coming on about this…
  • Autodesk will make all your customization work for you on all computers and other devices wherever you go – let’s put aside for a moment Autodesk’s total failure to even provide a usable vanilla AutoCAD on the Cloud so far. CAD Managers, would any of you care to hop in and let Andrew know what’s wrong with this picture?
  • Constant automatic incremental updates are like reading news articles daily and much more convenient than larger upgrades which are like getting a whole year’s worth of news at once – again, this makes the fatal error of conflating payment and upgrade delivery methods. Putting that aside, if we’re talking about virus definitions and OS or browser security hole fixes, then yes, automatic updates are the way to go. CAD software, not so much. Particularly software from Autodesk, given the incompetence shown to date in its attempts to make this model work. Even putting aside the practicalities, I could do a whole long post on why this concept is all wrong. Maybe I will later. Meantime, Andrew needs to talk to some CAD Managers to get some idea of how the real world works.
  • “OK, so there’s still the major elephant in the room: What about the cost?” – good of you to mention that elephant, tell me more.
  • For customers, there is real financial advantage by eliminating that huge upfront payment. – For some customers, yes. Not so many, though. Short-term customers are the minority. What about the millions of long-term users who would have their annual costs blown sky-high by falling into your rental trap? Andrew, I see you mentioned the elephant in the room and then tried to avoid meaningful discussion of it, giving the impression you had addressed the issue without actually doing so. Sorry, but I noticed. Care to try again? Tell me more about how you expect either a) customers to be better off by paying more, or b) Autodesk to be better off despite customers paying less. Pick either one of those and run with it, I’m sure it will be entertaining.
  • “And if you don’t need a product for months at a time, switch it off, and then switch it back on. It will be there ready and waiting for you” – strange, that kind of flexibility seems to work for perpetual licenses too, at a fraction of the long-term cost of rental. No guarantee that flexibility is a reality for rental products, though, because the vendor may not provide that product when I need it, or may have racked up the prices to exorbitant levels, or may have introduced new incompatibilities or other technical problems. Oh dear, the boot is very much on the other foot with that argument.
  • “After three years, software becomes obsolete…” – er, no. Many people (myself included) are happily productive using at least some software more than three years old. Some of it works better than the newer stuff. Hands up all those people who couldn’t possibly live without the latest version of Word or Excel, for example. Anyone? Didn’t think so.
  • “…and the pace of obsolescence is rapidly increasing” – if we’re talking Autodesk software, then the pace of obsolescence is doing the opposite. AutoCAD improvement has slowed almost to a halt, for example. There is little in any of the last few releases that gives an AutoCAD 2017 user a significant productivity advantage over an AutoCAD 2013 user, say. And anyone using AutoCAD 2010 or earlier has a much more efficient Help system than that provided in any of the last 7 releases. I guess that’s the kind of anti-progress that happens when you sack a bunch of knowledgeable people every few years and divert too many of the remaining resources to trendier projects that you end up junking anyway.
  • Customers of Autodesk can continue to renew their maintenance contracts for as long as they want – except that Carl Bass has now indicated otherwise. Andrew, maybe have a word with your boss and get back to me on that one?
  • “The company is always listening to how to improve the transition and setting out for the long road, not the short win” – except rental is all about the opposite: short term savings that cost big in the long term. And don’t get me started on the irony of claiming Autodesk is “always listening” while promoting an all-rental scheme that goes against the very clearly expressed wishes of customers.
  • “It’s this beautiful kind of world where things are connected and work together better” – does it have rainbows and unicorns, too? Strewth. Come off it, Autodesk is rubbish at CAD interoperability, even among the AutoCAD-based products. Why should anyone who’s been struggling with poxy proxy objects for a couple of decades believe that paying differently is going to act as some kind of magic spell to make everything exquisite in CAD Connectivity Kingdom?

Here’s the TL;DR version of my response to Andrew’s arguments if you can’t be bothered reading all that:


What are the real reasons Autodesk is going all-rental?

  • Autodesk wants to charge us long-term users three times as much money for the same thing and leave us with nothing at the end of it.
  • Autodesk thinks we’re all stupid and don’t own calculators.
  • Adobe did this and made it work, and Autodesk thinks it can do likewise despite significant business differences, much higher prices and an untrusting customer base.
  • Autodesk has run out of motivation and/or ideas to improve its traditional cash-cow flagship products, to the extent that customers increasingly no longer see value in upgrades or maintenance.
  • Increasing income by product improvement is way too difficult; price gouging and spin is much cheaper.

I’ll conclude with my own strained analogy:

Autodesk spin is like a tin of baked beans. No matter how attractive the packaging, the end result is just a bad smell.


  1. This rental scheme is just a way of raising prices without making that original purchase so big. If it really will make more money, Autodesk was foolish for not doing it sooner. The market will adjust though, but in the past that has meant staying on some older adesk product version. Now, the adjustment will be to non-Autodesk products, and stockholders might notice that.

  2. “After three years, software becomes obsolete…” Autodesk self-indicted itself on this one by not changing the DWG format for five years — instead of the usual triannual change. This may be where Mr Anagnost might have derived the 3-year lifespan statistic.

    • Autodesk is to be commended for keeping the DWG format unchanged for 5 releases, that’s about the only customer-friendly thing it has done over that period. There hasn’t been a need to change the format, as Bricsys has proven by using the same format and doing a lot more with it.

      Anyone want to take bets on how long Autodesk will continue to make life easy for customers by sticking with the 2013 DWG format?

      *Dr Anagnost

Leave a Reply