Spreadsheet Assassins – A short history of "software as a service"
thebaffler.comI subscribed to the Baffler because they write well, and to break out of my own echo chambers a bit.
Glad to see them on HN, but a bit disappointed in the response. They article is mostly correct in it’s lampooning.
Even if you disagree with points or find it ironic that they offer a subscription (btw equating producing a publication with SaaS is quite erroneous) — I think its useful to take this as a datapoint on how people outside tech see tech. That matters.
I think a lot of folks in software would agree with most of the article as well. It has a hyperbolic edge but that's a stylistic concern rather than a material one IMO.
The fact that both offer a subscription is not ironic - it simply highlights the central point. In the end, those paying a subscription for thebaffler.com or for salesforce.com are paying for something very similar: they are paying for words and ideas.
People think that building a software program is like building a house. And, I grant, in some ways it is. But in the end, the customer is not paying for anything tangible. He's paying for words and ideas.
That’s interesting. The thing is, the Baffler continues to create new words and ideas, and I pay for those new words and ideas every month if I subscribe.
No new words or ideas show up next month on your software, so why should I keep paying you? I know you’ll say maintenance and updates, but from the user’s perspective there is quite the difference.
To me, Obsidian is a perfect example of how to get SaaS-revenue without making the user captive: All data is stored in an accessible form on the user hardware, and what you pay for is sync.
X-user kepano (who is a/the creator of Obsidian) has some nice insights on this _file over app_ philosophy.
> (who is a/the creator of Obsidian)
No. He is the current CEO, an excellent communicator and definitely the right person for the job, but he wasn't there in the beginning and nothing meaningfully changed about Obsidian's philosophy since for it to be attributed solely to him.
I don't disagree with him on anything in particular, it just feels bad to see the two equated. Original devs wanted to spend their time on actually making the thing, so they hired a CEO. They even picked from the community, kepano was the developer of the most popular Obsidian theme well before he got hired, but he is absolutely not the creator of Obsidian. When he joined, Obsidian already existed.
He's also not the only community hire either, there's at least one or two more people that started as plugin developers before they got hired.
Sounds utopian and inspiring. I hope it never gets enshitified.
This is an effective, simple solution if the application doesn't have transactional constraints and authorization. The user is in full control of the data - and that's a good thing.
It becomes a bit more complicated if the data has to be in a specified shape and can only partially be authored by any individual due to authorization requirements. Having a single source of truth on a server is quite a bit easier to reason about and design in this case.
I don't know where Obsidian stands on this, but I assume that if it has collaborative features or complex data constraints it has to do some quite sophisticated stuff to make it work reliably.
> if it has collaborative features or complex data constraints
It has neither of these things - it is aimed at individual (non-collaborative) use, and is intentionally very open and structureless. It is uniquely positioned to succeed with their chosen business model.
I wouldn’t be surprised if a large part of Obsidian’s use are free accounts illegally using it for professional purposes.
> what you pay for is sync.
You also have to pay for commercial usage. And may choose to pay for early access to updates.
And the sync is locked behind their offering, if you use other software to handle the sync, you're out of luck if you want the data to be available on mobile since you can't dynamically choose the location of the vault.
It's still miles ahead of other offerings when it comes to having the data be "open", but I found this limitation frustrating.
> you're out of luck if you want the data to be available on mobile since you can't dynamically choose the location of the vault
Is this possibly an iPhone-only limitation?
On Android, I can add as many vaults as I like, and choose any directory for that. I'm happily using it with a directory that is synced via Syncthing.
I've been using Obsidian for months now. I couldn't justify the high sync cost even though the offering is nice. So I hacked a solution with the community plugin "remotely save", syncing to WebDAV. I thought it would be shoddy and after 2 days I would just go with the official solution. But no. Obsidian is a "too good to be true" kind of thing.
Also, the fact that it's just files allows me to shamelessly use Restic to back it up. No "intergrations" or "export workflows" - my data is my own (although it might be snooped by the app), I get to decide what to do with it, how and where to put it.
For me, this is the future. Imagine your data is yours. You get to enjoy all the possible ways of making value out of it, not just in some crippled ways that make other rich people richer.
Last time I checked 3rd party plugin 'Remotely Sync' worked fine on mobile. You just create a new vault with any name you want, install the plugin, fill-in the settings and the sync will copy content of your existing vault there. After that sync will work in both ways. Personally, I use S3 backend and didn't encounter any issues so far.
That’s genius. I’m glad their one primary paid feature can’t be trivially circumvented.
This is only partially true. Their business model is based on business use, not personal use. They do allow free community based solutions. Like Notion - use it for free for personal stuff. But better UX and you get to take your notes home.
> The result? Some SaaS companies achieve gross profit margins of 75 to 90 percent, rivaling Windows in its monopolistic heyday.
Gross profit is a misleading statistic in the software industry, where almost all of the effort comes from building the thing.
I wouldn’t say it’s misleading, it’s still a very important financial metric. You just have to be aware of its limitations
I would say it's misleading when it's in a paragraph which is broadly implying that a high gross profit % is a sign of profiteering. See the actual paragraph where the article compares the gross margin of SaaS against the NET margin of Visa/Mastercard.
> Some SaaS companies achieve gross profit margins of 75 to 90 percent, rivaling Windows in its monopolistic heyday. Even the reigning credit-card duopoly of Visa and Mastercard wield a mere 51 percent net margin.
I don't see how anyone can say the above statement or comparison isn't a bit misleading, as the Gross Margin of Visa and Mastercard will be closer to 100% than to 50%.
In reality it's unsurprising that software looks really profitable when you remove all the costs of developing the software.
In this article using it is very MISLEADING, exactly as Closi writes above.
Exercise: compare net profit of Visa/Master card to net profit of those "highly profitable" saas marvels.
Exercise two: compare gross profits of them.
Result: whole paragraph turns not being true while not aligned with the article opinion and showing actually opposite :-D
p.s. I do agree with the overall article :)
These saas do innovate, what will I do without my confetti animation when I drag the issue to done.
Next: animation packs for your JIRA ticket events. Only $5.99/mth/pack/user
The author shows notorious negative cases to invalidate a whole industry. Talks about the beauty of Excel more with a nostalgic look. If a company can replace a SaaS subscription with a process in which some employees share an Excel file, they don't need other software at all. Any software product (SaaS included) what tries to sell is a methodology or a way to do the things, condensing the know-how of an industry; whereas an Excel file most of the time means a clumsy attempt by someone to bring some order to the process. The data dependency is something governments should regulate, forcing easy portability. Of course, after years of using a specific tool it is natural to create a dependency, but it could be considered a tax from modern life as the author himself explains, it's something you cannot escape from. Or at least, the change will not involve just cancelling all your company SaaS subscriptions
There is a subscribe (in exchange for monthly payments) button under the article, an unaddressed irony. Or perhaps news outlets need money and software shops don't?
Is this an irony? Paying a subscription fee to a magazine, for people to actually write such sharp and to the point articles (no shameless "the industry is broken! Enter my my startup. It will solve it, buy it now" articles) is really different for paying a subscription for maintenance mode 2010 software.
The margin in the former case is much lower since people need to perpetually invest time to create good content. Which is the whole point of this article - this isn't correct for the latter.
Thoma Bravo the “bridge trolls of enterprise software,”
they are domestic terrorists if you ask me... because for the majority of people effected by them, its terrorifying. the way they operate is not in our countries/market best interest, but only the interests of a top few, who frankly include real global terrorists as members of the fund. The US should start to investigate PE, PE is crushing the economy under its own weight of consolidation and it hurts every day people.
I've been blaming MBAs for ruining large companies for about 15 years. More and more of them were minted, and true to their word, they really have gotten good at extracting every last dollar out of whatever a company sells.
Now that small business owners are watching/listening constantly to small business gurus and hustle gurus, those businesses are screwing customers in ways I didnt see 20 years ago.
I keep in contact with many small business owners in construction and repair, and almost everyone now thinks that its better to get less work and just charge more. Sure they could pour 3 driveways a week and make $2k profit per driveway, but honestly they prefer to do and make $6k profit on 1 driveway a week and call it a day. I see this constantly in small low employee count construction companies.
> I keep in contact with many small business owners in construction and repair, and almost everyone now thinks that its better to get less work and just charge more.
I don’t see what MBAs have to do with that, it’s a tale as old as time: given the choice to have the same pay and more leisure time, most people will take more leisure time. The exceptions are probably mostly people doing intellectual labor, not physical labor like pouring driveways.
> but honestly they prefer to do and make $6k profit on 1 driveway a week and call it a day.
If like many I see, they may also spend the spare time creating TikTok and Instagram videos how $12k is the absolute minimum a drive should cost
There must be sufficient amount of wealthy people needing jobs done who won't question quotes that the rest of us are just screwed and have to resort to DIY
MBAs?! Forget those fools, they don't do shit, an MBA is adult day are for rich kids.
Its PE, nothing else to say and its Thoma Bravo most of all.
Enterprise SaaS will charge hundreds of dollars per user per month, with a viscous lock in. The software mafia.
I think the next thing is adding the ability to download all your data, so you could switch to a different service. There are many advantages in having companies manage your data and software over time. The pernicious part is when they make it "sticky" by preventing you from accessing the data you have created with their service. If you're able to access your data, you at least have the option of downloading it at any time and canceling the service. Whether you can use it or not in another service is a big unsolved problem. However, it is not unsolvable. Another set of companies could easily evolve that provide transfer services.
I think the article fails to address the appeal of SaaS. The industry isn't massive for no reason and ignoring the lasting popularity makes the subject difficult to understand.
SaaS exists because managing software is hard. The easiest way to get money from a company is to make problems go away. SaaS means your data is stored remotely (you don't have to worry about maintaining your own servers, Databases, synchronization, liability goes to the SaaS company), updates come when they come (your internal support couldn't help, it's someone else's problem anyway), the spread sheets they are replacing were hated anyway and caused constant problems.
This is what I think people don't get about SaaS: they think it's about the first S. The technology. It's not. It's about the second S: the people who keep the technology from collapsing into a black hole of complexity. I've worked at a place with an AWS bill that could have paid for multiple employees. We weren't paying for the software, we were paying for it always to be running. Competent people who are willing to carry pagers are expensive and hard to find. The genius of AWS is that Amazon hires a lot of those people and obscures their existence by putting an API and a bunch of crooked salesmen in front of them.
Although I have questions about certain points, I appreciate the perspective. I wanted to give 5 bucks. There were options for 10, 20 and “any” which goes to a select box that starts at 25. Now I can’t give anything.
I agree with the author that working on files is preferable, if possible. But SaaS often also solve the hard problem of dissemination, and good luck syncing an excel file with 100 people via Dropbox.
Networked drives have been a thing for ages.
As someone actively pushing a company away from that model to the ubiquitous M365/SharePoint/OneDrive approach, there are so many hidden costs to 'a network share'.
How are you managing VPN access to file servers? How about backups/versioning? Any searching? Access control that can be easily managed by end users? Data governance, RBAC, compliance audits? BYOD/mobile device access?
> How are you managing VPN access to file servers? How about backups/versioning? Any searching? Access control that can be easily managed by end users? Data governance, RBAC, compliance audits? BYOD/mobile device access?
This seems like a great idea for a SaaS product.
Yeah it's called Google Workspace.
Maybe some aspects, but I don't know about the VPN side of it, and my point was directly to the idea of managing shared drives.
That's fair.. Google does have BeyondCorp but that's not standard in Workspace.. So it's not a packaged deal.
you can't manage permissions very well or precisely. You can give a user "read only" or "edit" but you cannot manage individual sheets or cells. Spreadsheets are great for lots of things, but sharing them with lots of people is not one of them.
Brilliant. Bravo. Well said. Look at the companies in any recent YC batch and it’s mostly just a long list of ridiculously simple SaaS ideas that anyone who knows how to use a spreadsheet doesn’t need. It’s amazing that so much money and talent in the West is dedicated to expanding rent-seeking and financing, instead of actual innovation or economic productivity. No wonder we have a housing crisis and a stock market decoupled from reality.
And yet many of them succeed? So unless you think all their subscribers are stupid, they must see a reasonable value exchange.
In the large enterprise where I work, I've often argued against building our own in-house XYZ with the position "do we want to be in the business of building and maintaining XYZ or do we want to focus on our core business?"
In some cases it definitely makes sense to build something yourself, but in plenty of cases, paying an external company a recurring fee (that is probably less than the ongoing capital you'd spend doing it yourself) is absolutely the right business decision.
Plus, I've rarely seen one-off non-core internal "products" at a company actually be better than the product that a specialized company can offer.
User experience matters, the tools and software your employees use should be held to the same quality bar as what you offer your customers.
And let me tell you, internally built software rarely comes close to that bar.
In many cases, I would not say that the subscribers are stupid, but I would say that the company's decision to use the software was stupid. Often (usually?) the decision is made entirely by people who aren't involved in the day-to-day of the work that the tool will be used for.
The people who will be the ones using the software the most would oppose the decision if they were involved, but they weren't. They begin using the software, and it becomes clear that there are numerous problems - either bugs or, more common and insidious, incompatibility and inflexibility. But now there are high-status people in the org that have reputation staked on the success of the adoption, so they continue to shove a square peg (or just a bunch of shit) through a round hole. The software becomes embedded in the org's operations, and the costs to disentangle it become high. This is how software that does not improve things can be "successful".
I have seen this happen over and over. It is of course the fault of the org for allowing this to happen.
The internal products I support may not be as fancy looking as the specialized commercial products, but it will be designed for our exact workflow and internal weirdness. IME users will generally prefer the product that has only the 7 fields they need, and already "knows" that on alternate Tuesdays the TPS reports get printed in landscape, because of that silly local council rule.
The counter-point here is that 10 years passes and processes have changed and needs have evolved and everyone's still stuck using that old internal app that was built by people who are no longer at the company and no one wants to update it because it's toil and there's no budget, and so processes are cumbersome and artificially complicated because everything has to go through that app.
Case in point at my company there's an internally-built ITSM app that requires a user to select "Approved" no less than 8 times across 4 screens to approve an internal request.
My point here is that there will always be edge cases that disprove the rule, so yes you're right, but are you right most of the time? I doubt it.
In my 30 years working in this field (and 10 of those years were spent helping companies deal with poor internally built apps), the majority of internally-built applications at companies that aren't modern software companies (i.e. most companies) age very poorly, have poor internal support and little to no technical health budget, have terrible user experiences, and are generally disliked by most employees who have to use them.
Particularly in these times when everyone can compare the tools they work with to the stuff on their phones or laptops that they use every day. It's an unfair comparison but it's still there. The expectations for interfaces and software usability have increased faster than most companies can manage for their internal products.
> unless you think all their subscribers are stupid
Ah, there’s the capitalist’s circular religious belief, if it’s profitable it must be good, and it’s good because it’s profitable.
Except some of the best things in life are not profitable, and some of the most profitable models are not good (such as Facebook/Google’s privacy invasions and the ad-driven model).
To think that subscribers pay money because they see a “reasonable value exchange” is to be completely blind to the incentives behind many big ticket software purchases. Managers pick IBM/Microsoft/Slack just so they can’t get fired for failed boondoggles or because they don’t care about anything except maintaining the status quo. Startups choose certain SaaS tools because they’re part of the same VC-funded slosh-it-around party. There’s just so many millions to be made by selling to fly-by-night startups or clueless bureaucratic enterprises or government that you don’t actually need to be innovative or useful to “make it” in todays tech world. You just need to know how to play the game. There are exceptions and true innovators of course, like OpenAI or Apple or AirBnB but those are rare.
> > unless you think all their subscribers are stupid
> Ah, there’s the capitalist’s circular religious belief, if it’s profitable it must be good, and it’s good because it’s profitable.
Not sure how you got to that conclusion from my words, unless you're projecting a metric ton of your own bias into my post.
Sure you can turn this into a big complicated discussion, but I still say that for most companies the decisions is "is it a better choice for our company" and that can absolutely involve spending money but also focus, experience, employee happiness, future flexibility and more.
> To think that subscribers pay money because they see a “reasonable value exchange” is to be completely blind to the incentives behind many big ticket software purchases.
Can you be more condescending? Honest question.
The value exchange is not the only reason but it is the primary reason most companies (and people) spend their money (at least in my experience and observations).
It doesn't actually need to be more complicated than that in a lot of cases.
I don't disagree that what you're saying certainly happens as well, but you make it out like that's always what happens and I don't believe that's the case.
After years of dismantling Solarwinds, a ubiquitous network management software, Thoma Bravo and Silver Lake dumped $286 million of shares in December 2020, the day before the company disclosed a staggering malware breach. The “Sunburst” cyberattack exposed data from California hospitals, the Department of the Treasury, the U.S. nuclear weapons agency, and dozens of other agencies and major companies. While the root cause is still under investigation, multiple lawsuits and reports describe a culture of cost-cutting and frequent security lapses. “Typically Thoma Bravo raises prices and cuts quality, but the affected constituency group—corporate IT managers—don’t have a lot of power or agency,” Stoller writes of this precarious stalemate, with potential geopolitical implications. “Their superiors don’t want to think about a high-cost but low-probability event, especially if every other big institution would be hit as well.”
There's some irony to an article about wedging into our monthly bank statements hosted by a site that asks the same...
Yeah because content and services are the same thing.
If you think it’s meaningfully different to ask for a subscription for content than for a service it would add a lot if you could share why rather than just a sneering hot take. It’s certainly not obvious to me that one is good and the other is not.
Sorry about that, you're right that it was kinda low. This is from a previous comment. It's still not the nicest way to put it, but it gets the point across.
- Does reading The Baffler trap you in its ecosystem? If you cancel The Baffler do you lose functionality?
- If The Baffler closes up shop because it was bought by a larger newsletter does it brick any of your devices?
- I turned uBlock off for this page (it was blocking 6 things) and there are no ads. Can you say the same thing for Netflix or fricking Windows?
- Does The Baffler take what you give it and resell it because its 5,000 word privacy policy which you didn't read but had to agree to says it can?
- Will The Baffler's third rate security practices release your identity and those of your customers to private auctions on the dark web?
This is how the newsletter supports itself. Looks like it's about $5 per article. In a period where everybody expects web content for free, we're paying literally trillions to subscribe to software products. As somebody who works for a content organization I'm now asking myself what aspects of the SaSS model can we take up to get people to pay our writers and editors to continue producing local news? I guess we just need to figure out what content people are keeping in spreadsheets.
Thanks for doing that. Honestly I think some of the criticism of Saas behaviours seem fair. On the other hand I don't necessarily think the criticism of the saas model itself is completely fair.
Making a website or newsletter requires an ongoing investment in people to produce content. So it's ideal to have an ongoing stream of subscription revenue to match that ongoing cost. As a user I happily subscribe to certain news outlets because I recognise the value they provide to me on an ongoing basis and think it's a fair quid pro quo.
Lots of things that people expect from software (ongoing updates, support, online features etc) incur an ongoing cost from the team so a subscription-type model is also better for the software company than a one-off payment.
So you can see why saas businesses and their investors like it. I think what they often fail to do, is show users that this model represents a good deal for them also. In other words they want to charge for software as a service but actually only provide software as a subscription (without real service behind it).
Ie if you're charging a subscription the user should get great support. They should get ongoing updates etc etc - things which cost money on an ongoing basis.
It’s interesting to see my industry described from a different point of view.
It misses the key point of the economics of subscriptions: a single up front fee is a disincentive to the constant maintenance that delivers the single biggest piece of value to the customer, that being the continued existence of the software. Instead it creates an incentive to add lines to the marketing BS on this year’s release and prioritises new customers over existing.
A subscription fee required for receiving updates for an application that you own seems fine.
A subscription fee required to avoid that the application you are using will be disabled remotely seems unacceptable.
The latter certainly does not create any incentive for the software vendor to improve the quality of their product and to fix its bugs.
On the contrary, the latter business model provides a stream of revenue for no work, so the vendor is encouraged to stop any maintenance work for their product, much more than when the customer makes a one-time payment. This has been amply demonstrated by the behavior of Adobe, Broadcom and the like.
The second point just strands parts of customers on outdated versions - remember that time when loads of companies were running Excel 1997 or 2003 to their deaths to save costs?
Then anyone on Excel 2013 couldn't use any new features because they weren't backwards-compatible 10 years to some public sector organizations that refused to update.
Office 365 is so much better, I can actually send someone a file that uses LAMBDAs or uses PowerQuery and expect them to be able to open it.
> The second point just strands parts of customers on outdated versions - remember that time when loads of companies were running Excel 1997 or 2003 to their deaths to save costs?
I do fondly remember keeping a floppy (!) of Word 5.1 in a safe place over the summer in college because Word 6.0 was such a steaming pile of crap.
You can’t expect that. A significant number of people are still using Office 2016/2019/2021 (and there is a 2024 upcoming). Personally I don’t plan to switch to Microsoft 365.
People in general don’t want their software to change continuously.
I agree that people probably don't want to pay continuously, but i'm not convinced that people in general don't want automatic/ongoing product updates.
Done at "...another CEO-turned-MAGA ghoul". Snark-as-a-Service seems oxymoronic.
He's not wrong though. David Sacks isn't exactly shy about being pro-Trump for all the worst/selfish reasons.
I find "MAGA ghoul" offensive.
Honest question: why?
MAGA's brand is being unapologetically offensive. Surely they can take what they give.
You:
- assert without evidence
- rationalize
This does not refute my point in the slightest: calling someone a 'ghoul' is pejorative, even if one finds their platform unattractive.
I wasn't actually trying to refute anything, sorry if it came across that way.
It was two separate things.
First I wanted to know what you found offensive, and you answered.
My second comment was more of a broad statement that a movement like MAGA that is itself intentionally pejorative and bigoted towards so many others in society (evidence of this is widely available) can certainly take some name-calling without crying foul.
It had nothing to do with your sensibilities or your right to be offended by said name calling. But I wasn't clear about it I guess.
"Make America Great Again" is self-explanatory.
As is the source of any "bigotry" on offer in this thread.
One hopes that you have an excellent weekend.
Lovely; a monthly sub to an article talking about monthly subs. You have to realise I will never, ever, going to pay ‘thebaffler’ and especially an article explaining that everyone wants subscriptions in a site that gives me one article and the rest is a subscription.
More than "one free article"; I checked and TheBaffler seems to allow you read 4 articles(/mth?) before subscription-walling.
It literally says it at the bottom for me; from the start; maybe elsewhere it’s another number or it lies. Point stands no? Article writing about subscriptions wants a subscription.
Interesting; I'm in the US; where are you? Also, mobile/desktop?
(It literally doesn't say it on the webpage served to me; don't assume the same version is served to all countries)
EU and mobile; it pops up right away and says 1 article. I tried to view another one and that indeed got me to the payments page.
Must be due to the GDPR then.
Doubt it but people seem to misinterpret gdpr rules all the time.
Americans are especially paranoid about the letter of the law while here we only use the intention. It’s all only about big offenders not some blog no one reads.
I didn't say "GDPR rules", more like a business decision about how many free articles to allow non-subscribers to view in GDPR vs non-GDPR region.
So ... software and content are the same thing? SaaS is the same as an article?
- Does reading The Baffler trap you in its ecosystem? If you cancel The Baffler do you lose functionality?
- If The Baffler closes up shop because it was bought by a larger newsletter does it brick any of your devices?
- I turned uBlock off for this page (it was blocking 6 things) and there are no ads. Can you say the same thing for Netflix or fricking Windows?
- Does The Baffler take what you give it and resell it because its 5,000 word privacy policy which you didn't read but had to agree to says it can?
- Will The Baffler's third rate security practices release your identity and those of your customers to private auctions on the dark web?
This is how the article supports itself. Looks like it's about $5 per article.
Still though it's interesting how you put it. In a period where everybody expects web content for free, we're paying literally trillions to subscribe to software products. As somebody who works for a content organization I'm now asking myself what aspects of the SaSS model can we take up to get people to pay our writers and editors to continue producing local news? I guess we just need to figure out what content people are keeping in spreadsheets.
ungated link: https://archive.md/BQqNk