Search Unity

  1. Welcome to the Unity Forums! Please take the time to read our Code of Conduct to familiarize yourself with the forum rules and how to post constructively.
  2. Have a look at our Games Focus blog post series which will show what Unity is doing for all game developers – now, next year, and in the future.
    Dismiss Notice

Unity financial results

Discussion in 'General Discussion' started by Kamyker, Aug 10, 2021.

  1. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
  2. undevable

    undevable

    Joined:
    Dec 25, 2020
    Posts:
    140
    The profit is $215m while the revenue is $273m, while the expenses are ~58m in 3 months. Margins are great, around 78%.
     
  3. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
    "Total operating expenses: $365m" not 58. $215m profit is before expenses.
     
    undevable and Ne0mega like this.
  4. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    12,147
    Unless I'm misreading the table, it looks like it is operating at loss, and spends more money than it earns, and lost 255 million in 2021. and 50 million in 2020.
     
    NotaNaN and april_4_short like this.
  5. april_4_short

    april_4_short

    Joined:
    Jul 19, 2021
    Posts:
    489
    Year vs year, for the first six months of the year.

    Revenue up $150 million,

    -- Losses up $200 million
     
  6. Gekigengar

    Gekigengar

    Joined:
    Jan 20, 2013
    Posts:
    654
    2021 has double RnD budget, does that mean we're finally getting a true real-time GI solution like Lumen?
    Are we finally getting a stable multiplayer solution 2x faster with double RnD?
     
  7. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    12,147
    No. It means they've wasted twice the money. We won't know what they used it for.
     
  8. Lagermeister

    Lagermeister

    Joined:
    May 13, 2017
    Posts:
    18
  9. Joe-Censored

    Joe-Censored

    Joined:
    Mar 26, 2013
    Posts:
    11,825
    It takes a very long time to get new developers up to speed with a large complex product like Unity. If they are in the middle of a large round of hiring, I would expect development to actually slow down this year as more experienced developers spend time helping new ones, and new ones don't actually produce much. You'd then see benefits from the new hires 9 months to a year later, longer than that from the customer perspective.
     
  10. EternalAmbiguity

    EternalAmbiguity

    Joined:
    Dec 27, 2014
    Posts:
    3,134
    angrypenguin likes this.
  11. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    18,245
    You had better hope that their doubling the budget this year doesn't mean they decided to focus on these because the actual time to develop them is not a single year. Realistically technologies like Lumen and Nanite are at least five year projects. Just to put that into perspective the SRPs have not been out five years.
     
    MothDoctor likes this.
  12. april_4_short

    april_4_short

    Joined:
    Jul 19, 2021
    Posts:
    489
    It's much more likely that this is the development costs of the extended range of bought software products increasing the total development costs.

    Buy a company, you gotta pay its developers.

    Unity's been buying a few companies.
     
  13. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    18,245
    Small companies. Parsec is practically indie. At most you're looking at a couple million.
     
  14. april_4_short

    april_4_short

    Joined:
    Jul 19, 2021
    Posts:
    489
    Here's one, that's adding to that development budget:

    It won't be hard to find others that are growing after being bought up by that Unity cash.
     
  15. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
  16. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
    Third quarter results in 2 days.
    2nd quarter 2022:


    I think I know how to read it basically Comprehensive loss is the most important info, it's Gross profit - Total operating expenses.

    So yeah, compared to 2021 in 3 months period Unity's loss increased by $50 mln, 6 months period $130 mln.
     
  17. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    8,914
    Those R&D costs are only ever going to keep going up dramatically if they keep going with their completely nonsense metaverse push and desperate chasing of new features.
     
  18. Gekigengar

    Gekigengar

    Joined:
    Jan 20, 2013
    Posts:
    654
    At least now we know from Unite 2022 roadmap that those R&D cost were well spent and worth it.
     
  19. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    8,914
    No, we don't. We won't know they're "worth it" until they've made it into the engine, prove themselves to be better than existing alternatives, and prove themselves to be actually used.
     
  20. Andy-Touch

    Andy-Touch

    A Moon Shaped Bool Unity Legend

    Joined:
    May 5, 2014
    Posts:
    1,287
    I think they were being sarcastic.
     
  21. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    8,914
    With the outright sycophancy I've seen about Unity's business decisions lately it's really hard to tell sometimes
     
  22. Gekigengar

    Gekigengar

    Joined:
    Jan 20, 2013
    Posts:
    654
    Actually not quite. Its a mix of both!

    Things I am genuinely excited about:
    - Memory Profiler 1.0 (This looks extremely useful!)
    - Multiplayer & Services 1.0 (I've been wanting non-third party MP Services from Unity since ages!)
    - Plastic SCM (Just tried this, genuinely very impressed!)
    - DOTS 1.0 (Opens up a ton of possibilities for the future!).
    - Full-screen master node for Shader Graph! (This been done in hacky ways before, now its fully featured!)
    - Actual Volume & Volume VFX (I never thought I would live to see this in a real-time renderer!)

    Things I personally am disappointed about:
    Disappointed about how Unity will be prioritizing resources on SRP-coexistence. This is betraying their own philosophy on why URP/HDRP are separate to begin with. And they will be spending years of precious resource on this very niche feature that nobody will use. If this is what it leads to in the end, they could've started URP/HDRP as one and spend less resources developing both to begin with.

    What happened to URP deferred rendering? It's very buggy and last I tested, it does not even support receiving shadow casters with Shader Graph shaders.

    True RT GI? Nothing?

    UI Toolkit taking so long to develop. Why does it take almost half a decade to develop this? And how is UI Builder this buggy? The preview just breaks if we have a slightly complex hierarchy. Today, it does not support basic feature such as % text scaling (Feature requested since 2020). Why?

    At this rate I wouldn't be surprised that we will get world-space UI, and custom UI shaders by the end of 2030. And text scaling will be on the next Unite 2031 roadmap.

    DOTS 1.0, both excited and disappointed. It was initially predicted and planned to be released at the end of 2018. Its 4 years behind schedule right now. (Will soon turn 5) How did this happen?

    The roadmap has been showcasing the same features since 2017. Terrain Tools, UI Toolkit, DOTS, and SRP. If you submit the slides to SO, it will definitely be reported as duplicate.

    Its as if doubling their investment directly halves their productivity in relation.
     
    Last edited: Nov 8, 2022
    Rewaken, JoNax97, shikhrr and 3 others like this.
  23. AcidArrow

    AcidArrow

    Joined:
    May 20, 2010
    Posts:
    9,607
    And it was useful, 5+ years ago, when it was a Unity maintained repo on bitbucket and it was working fine. Then it became a package and stopped working properly.

    If it starts working well again, sure, it would be cool, but this isn't a Unity success story. It's another "Unity started working on something cool, then they abandoned it for years".

    Also I can't imagine a feature whose bulk of work was done 5 years ago contributed much to R&D costs.
    Plastic is cool, mostly because Unity themselves have almost nothing to do with it, so I don't see how they get props. They just bought a company, this wasn't an R&D cost. Also this is too little too late to wash the bad taste Collaborate left in my mouth.
     
  24. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
    Same story with:
    Multiplayer hosting is Multiplay and services are old ChilliConnet.
     
  25. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    8,914
    As mentioned, this was something that existed for a while, got broken, and now works again.

    If you've been waiting since ages you should probably know this is like their third attempt and it hasn't exactly gone well the other two times.

    Plastic SCM has been around for fifteen years, it's just that Unity bought the company three years ago, so not really R&D.

    DOTS has been in the works for like half a decade at this point and I'm not holding my breath for a stable, production ready deployment for a while yet.

    This should have been a day one feature, honestly. Every time I was using either HDRP or URP I was using Amplify Shader Editor because of this oversight.

    Yeah, this one I'll concede to, but the fact is that the overwhelming amount of stuff you listed here either existed before, has been in production for ages, or was a glaring oversight in its omission.
     
  26. Antypodish

    Antypodish

    Joined:
    Apr 29, 2014
    Posts:
    9,912
    Maybe you are confusing individual packages? DOTS didn't exist per say back then. Its name has been cobbled around 2019 or so, if I recall correctly.

    But burst and jobs has been released since then. ECS is probably what you referring to? Indeed, ECS been worked for years. And as Unity explained, if following DOTS forum, it did required lots of internal changes in Unity code, to be be compatible.

    The massive delay around 2021-2022, was caused by earlier lack of communication between DOTS and Unity engine teams. They had to mitigate that first, before they could move toward release of DOTS 1.0.
     
    DragonCoder likes this.
  27. AcidArrow

    AcidArrow

    Joined:
    May 20, 2010
    Posts:
    9,607
    Does it matter when they coined the name? Or does it matter when they actually started working on what eventually became DOTS?

    I believe the name DOTS came in 2018, but they've been showing DOTS from since 2016.

    For example:

    "So for the last couple of Unites, I've been showing you different demos, all surrounding our new Data Oriented Design Technology Stack"

    And this what the screen is showing while he is saying that:
    upload_2022-11-8_15-0-24.png



    With that said, I don't believe they ever promised a release in 2018.

    If the community thought Unity could have completed their DOTS work by the end of 2018, then it's the community's fault for being gullible, starry eyed idiots, that hold Unity is way too high regard, a lot more than they are worth, that should know better than to think Unity needs less than 5 years to complete even the most basic of features, much less a new tech stack.

    After all ~5 years is how long it took them to implement zoom in / out in the animator window.
     
  28. Antypodish

    Antypodish

    Joined:
    Apr 29, 2014
    Posts:
    9,912
    Indeed.

    Myself personally I don't remember they had at any point firm release time. Unity always stated, they work on it, and there was something in road map ahead at least for 1 or 2 years ahead in time. Maybe someone did captured otherwise.

    Also there were clear indications while working with DOTS packages, that work is still long way to go.
     
  29. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
  30. Gekigengar

    Gekigengar

    Joined:
    Jan 20, 2013
    Posts:
    654
    Yeah my bad, I take back what I said. I remember seeing something about dots multiplayer will be released around end of 2018 - mid 2019, with 3 multiplayer template project. (FPS, RTS, and I don't remember the last one?) I thought that includes the whole DOTS stack. But hey I might be dreaming, because I can't find the source after 5 mins of googling.
     
    Last edited: Nov 8, 2022
  31. MartinTilo

    MartinTilo

    Unity Technologies

    Joined:
    Aug 16, 2017
    Posts:
    1,870
    Lol, I'm sorry but, no. No it didn't work. Especially not fine. And the "maintenance" was bare-bones.

    It was a hackweek project that got shipped, it was riddled with issues that were causing crashes and reported misleading information more often than it would work. I've worked on this thing for 4 years now, and I still don't fully understand how people managed to get anything useful out of it despite the odds. One thing tilting the odds was that Unity's allocators favores the lower value address space but God forbid your app reached into the 4 GB address range and every capture attempt could crash your game. It also read out of range memory and mostly didn't crash because the allocators weren't usually full to the brim or had a follow on allocation right after them. It didn't report the right sizes for a bunch of asset types, didn't report some sizes at all, duplicated the managed heap on capture on the device (often crashing OOM due to this), didn't report GPU sizes correctly on all platforms, had lots of other platform specific issues, didn't report VM memory for IL2CPP, needed re-implementation with the update to .Net 4.x, didn't correctly report connections between objects (it e.g. didn't properly consider the differences of these between Editor and Player builds), also reporting that in a convoluted way that was super slow and also liable to crash the app OOM (incidentally we made Resources.UnloadUnusedAssets faster in the process of fixing that up), didn't report the type information for all managed types that had objects in memory, ...

    I could go on but, suffice to say, after 4 years of trying to make this thing work, I'd say the only reason why it was helpful at all in those anecdotal instances were it miraculously didn't crash and burn, was because before the tool existed, Unity's Memory system was a complete black box so any visibility into it was better than none, even if it was at times completely missleading.

    As outlined above, I've worked on it for just a bit over 4 years now and I can see how it'd be hard to imagine from the outside looking in. Honestly, I wouldn't be able to imagine it if I hadn't seen it first hand, and I've seen stuff in it's underbelly that you probably rather wouldn't want to know we're there. My main point is, it has been actively worked on for the last 5 years, by multiple people. We released the package as experimental on 2018.3, which was possibly too early, but see my point above about the black box and, honestly, half of the horrors we discovered came through bug reports of people using the package, so releasing this early definitely helped during the development. We're not kidding when we say we need every input we can get to build things the right way.

    And we've also worked on it (alongside work to the general profiler tools as well as with the entire profiler team exclusively for the last year) since that experimental release, and yes, we dropped support for the bitbucket one in the process, but honestly, once the connections between objects had been fixed it was beyond saving. (Caveat: 2021 saw less development there as there was a company wide effort to address product stability and bugs and the team focused on stability across the profiling tools and had, by then, fixed most stability impacting issues on the Memory Profiler side.)

    I won't comment on anything else in here in detail but if you ever worked on something that took unexpectedly longer than you would have ever anticipated, that might provide a point of empathy for you. Maybe try to consider Unity's platform breadth, the sheer infinite amount of setups that Unity allows you to build, and multiply that with the general R&D estimation error margine. If I do that then suddenly, 5 years off mark doesn't sound half bad.

    E.g. I've picked up a bug to fix selection for profiler samples not carrying over from Hierarchy to Timeline view in March 2021 with the comment "let's see if I can find a minimal fix for that". Let's just say that for the rest of that year I wanted to eat those words and didn't do much else than that and related bugs I discovered while adding tests. (Most depressing is that no one notices all of these fixes because now it just works as you'd assume it should.)
    Estimates are hard, and a lot of work happens under the surface so it's like the iceberg principle ^2 if you look at it from the outside.

    Thank you :oops::)
     
    Last edited: Nov 9, 2022
  32. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
    Tbh I don't care that much about R&D costs but it's very obvious that Unreal is doing much better job at announcing and implementing new features without the
    What's worrying is that Unity keep spending money buying random companies to cheat growth (like a lot of public companies). 25% of expenses are also Sales and Marketing ($100 mln for 3 months)


    Back to the topic, copy pasted data from first report and 2019 vs 2022 looks absolutely crazy. Net loss is x6 higher, 67mln vs 381mln.



    Unity's probability of bankruptcy is 63% according to https://www.macroaxis.com/invest/ratio/U/Probability-Of-Bankruptcy . Won't be surprised to see more lay-offs at Unity in the future or a steep price increase.

    Third Quarter 2022 Financial Results Webcast 3 p.m. PT/6 p.m. ET: https://unity3d.zoom.us/j/95182853619?pwd=cFVKbTQ3aXg4cjhUYmloa1JwR2Q5Zz09

    BTW I prefer using Unity over Unreal bcs of C# but this looks ridiculous.
     
    pm007, Cranom and DungDajHjep like this.
  33. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
    Last edited: Nov 9, 2022
  34. Kamyker

    Kamyker

    Joined:
    May 14, 2013
    Posts:
    927
    Found the webcast saved stream:
     
    shikhrr likes this.
  35. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    8,914
    You won't see price increases on the engine because they already tried that and started getting major pushback even though that was just enterprise, but also because the places you'll see price increases are going to be the provided services since that's where the real money is. Services and ad payments are going to see the biggest changes if they want to bring in dollars.
     
  36. AcidArrow

    AcidArrow

    Joined:
    May 20, 2010
    Posts:
    9,607
    It was working enough to help me notice that some textures weren't unloaded properly. Sure, it wasn't taking into account stuff like my rendertextures being set to memoryless, but never the less, I got use out of the feature, which is actually more than I can say for most finished Unity features.

    In comparison the packaged version of the memory profiler, would straight up not work or just crash everything, at least last time I tried it.

    Here is me saying I found the bitbucket version useful years ago:
    https://forum.unity.com/threads/wev...ew-memory-profiler.358963/page-2#post-3396401
    and
    https://forum.unity.com/threads/wev...ew-memory-profiler.358963/page-2#post-3513339
    I mean, I always suspected that Unity doesn't understand how people use their features and what they find useful, but it's nice to get confirmation, thanks.
    I'm working on a game for over a decade, but I'll afford you the same empathy you afforded me with my next reply:
    "Lol, I'm sorry, but, no."
     
    neoshaman, Murgilod and neginfinity like this.
  37. MartinTilo

    MartinTilo

    Unity Technologies

    Joined:
    Aug 16, 2017
    Posts:
    1,870
    I see that my bemused sarcasm didn't carry across in text. I did read that you found it useful and heard the same from a lot of angles before. I've heard and read feedback and use cases for the old and new versions of the tool from probably well over hundreds of customers over the years, including yours here on the forums.

    My "lack of understanding" is akin to seeing a rusty patchwork of a car that falls apart while driving cross the finish line in a Grand Prix and going "How the heck did that thing drive all that distance without imploding".
    Hearing such points of feedback just helped reenforce how important the tool is and that we needed it in a shape that was stable and solid enough so we could build on it and actually maintain it.

    My maybe a bit barsh rebuttal to your comment was more aimed at the way you phrased your comment about the Memory Profiler 1.0 not being new because it already existed, making it sound like it was basically good to go and would've just needed maintaining and didn't see any work done to it since it was released to bitbucket after said hackweek. Because none of that is true.

    The fact that it fell apart left and right while we were trying to get it ready for a proper release probably speaks more to my point about it not having been in any maintainable shape to begin with.
     
  38. Sebioff

    Sebioff

    Joined:
    Dec 22, 2013
    Posts:
    203
    The price for Unity Pro was increased by 23% just last month.

    On the topic of the Memory Profiler: I think it's an incredibly useful tool that has received significant improvements to performance, accuracy and usability with every Unity version.
    We had some memory issues on our Unity 2019 project but tracking them down on that version was incredibly slow and we couldn't make much sense of a lot of the things the profiler reported.
    We upgraded the project to Unity 2022 just for the updated Memory Profiler and were finally able to figure out what was happening.
     
    Last edited: Nov 11, 2022
    MartinTilo and MadeFromPolygons like this.
  39. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    8,914
    You didn't even read the first sentence of a two sentence post. Astounding.

    EDIT: WAIT S*** I WAS THE WHOLE WHO DIDN'T READ

    I OWNED MYSELF

    CRUD
     
    hippocoder, Ryiah, JoNax97 and 3 others like this.
  40. stonstad

    stonstad

    Joined:
    Jan 19, 2018
    Posts:
    390
    Just curious, have you encountered any meaningful examples of C# being used in Unreal? I have seen third party tooling but nothing that convinces me I could leverage a full C# code base and/or libraries.
     
  41. bluescrn

    bluescrn

    Joined:
    Feb 25, 2013
    Posts:
    608
    I'd hope they'd start selling off some of the numerous companies that they recently acquired before it comes to that...

    But if Unity did go bankrupt, and one morning it was gone, the staff locked out of the offices, the license servers offline*, think of how many Unity-dependent studios and projects could be wiped out overnight. It'd be games industry armageddon, especially for indies and mobile devs.

    It'd be an incredibly brutal lesson in the risks of relying on closed-source engines and subscription-based tools that you don't own in any way.

    (*Ok, it's not going to happen like that with something on the scale of Unity. But I've personally experienced something similar when a dev studio went bankrupt...)
     
    Last edited: Nov 24, 2022 at 7:58 PM
  42. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    8,914
    That's really not how bankruptcy works.
     
    DragonCoder and shikhrr like this.
  43. TheOtherMonarch

    TheOtherMonarch

    Joined:
    Jul 28, 2012
    Posts:
    392
    Unity needs to stop growing and better organize its human resources Developers. It can take years to get a development team up to speed and adding more developers does not necessarily help.

    Earnings before interest, taxes, depreciation, and amortization (EBITDA) Trailing 12-month is -600,253. according to Yahoo finance. https://finance.yahoo.com/quote/U/financials?p=U On the other hand Unity is a tech company so P/E ratios don't matter.

    Sadly, I expect more layoffs and considering how disorganized Unity is I expect the best will be let go first. The cheaper less experienced and more productive shovelware developers will stay.


    .
     
    Last edited: Nov 25, 2022 at 1:13 AM
unityunity