Most of these apps are useless garbage nobody wants and few people are proud of working on. Even the biggest flagship apps are pretty thin - organized hitchhiking, a replacement for television, a new modern version of the Sears catalog.
This "article" is just an ad for Airtable which looks like a reasonably competent low-code system. Commercial low-code systems (e.g. ServiceNow as a more entrenched established comparison) exist so as to lock you into a platform for which you'll pay exorbitant per-user pricing that really starts to hurt as you scale up.
Maybe there's value to be realized by letting individual teams have influence over their business processes while still using a centrally organized system; but there are many significant costs to locking into something like this, and when you hit its limits you're still going to need coders and custom middleware, all of which will again be a sunk cost into a proprietary system you can never have real control over.
Having built a lot of glue apps with Airtable it’s definitely the case that coders and middleware get involved fairly quickly beyond spreadsheet use cases. I’m quite certain that Airtable > Spreadsheets, if for nothing else that field types, views, comments, and permissions are far more ergonomically implemented.
The 50K row limit is quickly met for some use cases. Airtable proved out visual SQL and now is going full no-code with their Apps system. I think this makes for a big opportunity for a novel no-code OSS tool. Node-Red is fantastic, but having that central data layer at hand is a killer feature.
> Maybe there's value to be realized by letting individual teams have influence over their business processes while still using a centrally organized system
This is absolutely one of the core reasons cited by large orgs that purchase these solutions. And it's part of the sales motion from these vendors.
When Bob in Finance needs to simplify one of his monthly close processes, he's not likely to get any traction asking for a new greenfield internal business app. But if his company has an investment in of of these platforms (like ServiceNow, SalesForce, Mendix, etc), it's possible that his team can take it upon themselves to solve their problem without having to get "real" dev teams involved.
This brings issues along with it for sure: governance, quality, and as you mention: lock-in. But quite a few of these departmental apps become quite useful/popular, and enable teams to innovate on their own.
I do worry about the future (~5-10 years down the road) when the proliferation of these apps becomes a problem of its own, at which point I fully expect to see a slew of new startups claiming to solve this problem.
The better vendors in this space are already baking "governance, scalability, CICD..." into their products for this reason.
Goddamn it's hard to build stuff with Airtable. I tried building a spreadsheet like thing with a couple of other web services for things like images and locations along with user access controls and it was such a pain. Ended up push it out to a custom service in a Droplet which did most of the work.
I had a thing that was outgrowing Excel, and a bunch of people said "Airtable! Airtable! Airtable!" and holy cow do I ever now downgrade any technical rec those people give me.
Maybe I do not understand Airtable, but given its limitations I'm 100% not sure why it exists in a world that includes Google Sheets.
The good thing about Airtable for me is how it is very convenient to set up connected data and navigate that.
At the company I work for we use it to encode the scripts (as in screenwriting) of interactive fiction, which allows us to have a nice schema to define information about characters, dialogues and other game elements.
I also use it as the backing database for my personal website, sort of as a CMS. I'm also making an e-commerce site for my dad that will use it to store product, client and order information.
Wintergatan (of Marble Machine fame) has a team of people collaborating on making a digital CAD version of their physical machine, and they use Airtable for task management and as a structured wiki.
I do wish that Airtable supported some other things, but it does fill a good niche where you need a more organized and connected spreadsheet with a convenient way of visualizing data beyond charting.
Google Sheets works fine, but there is no alternative for Airtable's record links, which is how one expresses relations between tables. You can get data from other tables, yes, but more often than not range syntax breaks, or someone puts the wrong data type somewhere and it's tricky to debug. For me, the main selling point of Airtable is record links and typed columns.
I don't think Airtable will be replacing general purpose spreadsheets anytime soon, nor can it handle very math-y sheets well either.
Thank goodness I bumped in to this comment. I run a software company and we use a spreadsheet for a CRM and I thought I'd try AirTable as an upgrade without taking the big step towards SalesForce.
I just could not make it work with AirTable. I was seriously starting to doubt my IT abilities!
If your answer to a problem is "I'll just write the solution myself", this isn't for you.
But if you don't have the time or programming skills to create something from scratch, Airtable is a very compelling alternative to "I'll just throw it in a spreadsheet."
I could not agree more. I fit very much in the "I'll just write the solution myself" camp, but as a product manager at a company that sells enterprise low-code software, you've hit the nail on the head for why people buy these solutions.
Orgs don't buy these solutions because they believe the end result is somehow better than a pro dev team and a proper software project. They buy these solutions because they can't afford to put their pro devs on some use cases, or just simply don't have that kind of staffing on hand.
As an organization grows, the more "spreadsheet apps" there are, the messier things get. It's harder for teams to collaborate with each other. It's harder to replicate success across teams. Pretty soon, everything is SharePoint+Spreadsheet hell.
Even if these low code platforms only provide one thing: a central data store - they've helped immensely. But interestingly (and again, I'm a bit biased due to my role/employer), people can actually make some pretty useful/functional apps.
Often these are apps that would have never been turned into an app otherwise. But the underlying business process benefits greatly.
I've also seen (a small percentage) of these simple apps become full-blown apps managed by dedicated dev teams over time. The problem was validated, and the value realized was so great, that the problem finally got the time of day and a "proper" solution was implemented.
These things can't/won't replace "real" developers any time soon. But they do provide a path to a very under-served segment of knowledge workers.
Personally, I think Microsoft could KILL if they made Excel more robust. I'm not talking MS Access here - I'm talking about a modern, cloud-native app dev platform that is centered on Excel. They already enjoy ubiquity, and they could change the game.
There's a reason that the onboarding experience for most low-code platforms is "Choose a spreadsheet to get started".
The low code platforms give raise the so-called citizen developers for biz analies, PM alike. There are so much value because they don't need to squat additional IT resource for simple internal apps to boost dept productivities, when they grow beyond Excel.
Why, yes, I too wish the world saw the value in modularity and that the Unixes and Hypercards took off in the way they were intended, but it's going to take a heck of a lot more than one company.
I can see in the future when hiring for a software developer position using an actual programming language, someone putting "Airtable" on their resume like it means something.
Given how many kids and youth are growing up with only cell phones and limited (or no) experience with real computers, these days Word and Powerpoint experience is a differentiating skill in certain circumstances. It's obviously not relevant for IT workers but it's absolutely something for employers to pay attention to if they're hiring young people for unskilled positions.
It's hard to say if Airtable will rise to a level of adoption/prevalence to result in (or deserve) this kind of thing.
But just to provide a counter-point (your comment is a bit catty): If it does, then people should put this on their resume, and that's perfectly legitimate.
Large vendors like SalesForce, ServiceNow and even smaller players like Mendix, OutSystems, etc. do create new job markets.
Go look up salaries for SalesForce Developer or ServiceNow Consultant. These are roles that one can learn without a full-stack background, but often command extremely high salaries because for better or worse, it's still a niche skillset, and someone has to manage those apps that someone built.
If you can show something you've built using Airtable at a job interview, that could still showcase your ability to document requirements, deliver a solution and create some kind of value for your employer. Considering that developers sometimes don't have the best communication skills, showing > telling.
Developers have to start somewhere, and most aren't going into their first job interviews having previously written full-stack apps that weren't Twitter clones or thrown together for a hackathon.
>A half billion apps will be created. We need more software builders
If "a half billion apps will be created", isn't the conclusion that we need fewer software builders? A half billion apps sound way more than needed....
Indeed. I mean, just look at product hunt. Nearly every failure is due to lack of demand because there's already way too many competing products built. And there are a lot of failures. What is the percentage of apps that end up getting a sizeable userbase, less than 1%? There's already such a vast oversupply of software products out there without any additional platforms.
It's far more accurate to say, we need fewer software builders.
Software, unlike restaurants, has trivial economies of scale, and colossal start-up costs.
It's much cheaper to open a restaurant, then build a word processor that someone can get value out of. It's much harder to build your restaurant chain to serve 100,000 people, then it is to scale your word processor, that can serve 100 people, to serve 100,000 people.
The company where I work has about a hundred engineers, and a team who makes tools for those engineers. There are uncountably many small niches of software customers, whether they're external or not.
This is completely misunderstanding what the article is talking about though - this isn't about launching saas products, it's about lowering the bar to creating bespoke software internal to businesses.
We need far more builders working on compatible open source platforms, apps, and tools, so we stop rebuilding dead end proprietary apps or, even worse, successful walled gardens.
> A half billion apps sound way more than needed....
'Half a billion' figuratively speaking is not much, if you account for all internal apps in enterprise (ex approval flow for weekly cafeteria menu). If there are millions of small and medium enterprise then there would be millions of cafeteria menu app.
There's a lot of negativity on HN for tools like AirTable, but as someone that uses Coda regularly, I find it invaluable as a prototyping tool, if not a "production" internal app. There's nothing quite like having modeled and populated tables in such a flexible tool while everyone is trying to reach consensus.
73 comments
[ 7.3 ms ] story [ 618 ms ] threadThis "article" is just an ad for Airtable which looks like a reasonably competent low-code system. Commercial low-code systems (e.g. ServiceNow as a more entrenched established comparison) exist so as to lock you into a platform for which you'll pay exorbitant per-user pricing that really starts to hurt as you scale up.
Maybe there's value to be realized by letting individual teams have influence over their business processes while still using a centrally organized system; but there are many significant costs to locking into something like this, and when you hit its limits you're still going to need coders and custom middleware, all of which will again be a sunk cost into a proprietary system you can never have real control over.
This is absolutely one of the core reasons cited by large orgs that purchase these solutions. And it's part of the sales motion from these vendors.
When Bob in Finance needs to simplify one of his monthly close processes, he's not likely to get any traction asking for a new greenfield internal business app. But if his company has an investment in of of these platforms (like ServiceNow, SalesForce, Mendix, etc), it's possible that his team can take it upon themselves to solve their problem without having to get "real" dev teams involved.
This brings issues along with it for sure: governance, quality, and as you mention: lock-in. But quite a few of these departmental apps become quite useful/popular, and enable teams to innovate on their own.
I do worry about the future (~5-10 years down the road) when the proliferation of these apps becomes a problem of its own, at which point I fully expect to see a slew of new startups claiming to solve this problem.
The better vendors in this space are already baking "governance, scalability, CICD..." into their products for this reason.
Maybe I do not understand Airtable, but given its limitations I'm 100% not sure why it exists in a world that includes Google Sheets.
I'm not sayin Airtable is a silver bullet though.
There are external services like Sheetsu or Sheety to make Google Sheets API usable.
If the tool actually makes it easier to access the api, I would prefer Google sheets over airtable.
At the company I work for we use it to encode the scripts (as in screenwriting) of interactive fiction, which allows us to have a nice schema to define information about characters, dialogues and other game elements.
I also use it as the backing database for my personal website, sort of as a CMS. I'm also making an e-commerce site for my dad that will use it to store product, client and order information.
Wintergatan (of Marble Machine fame) has a team of people collaborating on making a digital CAD version of their physical machine, and they use Airtable for task management and as a structured wiki.
I do wish that Airtable supported some other things, but it does fill a good niche where you need a more organized and connected spreadsheet with a convenient way of visualizing data beyond charting.
Google Sheets works fine, but there is no alternative for Airtable's record links, which is how one expresses relations between tables. You can get data from other tables, yes, but more often than not range syntax breaks, or someone puts the wrong data type somewhere and it's tricky to debug. For me, the main selling point of Airtable is record links and typed columns.
I don't think Airtable will be replacing general purpose spreadsheets anytime soon, nor can it handle very math-y sheets well either.
Does this include random public JSON endpoints? I might have to take a look at it for a project I just started.
They mention it on the linked timestamp, but I recommend watching the whole video to understand what is there completely.
I just could not make it work with AirTable. I was seriously starting to doubt my IT abilities!
https://cloudtables.com/
But if you don't have the time or programming skills to create something from scratch, Airtable is a very compelling alternative to "I'll just throw it in a spreadsheet."
Orgs don't buy these solutions because they believe the end result is somehow better than a pro dev team and a proper software project. They buy these solutions because they can't afford to put their pro devs on some use cases, or just simply don't have that kind of staffing on hand.
As an organization grows, the more "spreadsheet apps" there are, the messier things get. It's harder for teams to collaborate with each other. It's harder to replicate success across teams. Pretty soon, everything is SharePoint+Spreadsheet hell.
Even if these low code platforms only provide one thing: a central data store - they've helped immensely. But interestingly (and again, I'm a bit biased due to my role/employer), people can actually make some pretty useful/functional apps.
Often these are apps that would have never been turned into an app otherwise. But the underlying business process benefits greatly.
I've also seen (a small percentage) of these simple apps become full-blown apps managed by dedicated dev teams over time. The problem was validated, and the value realized was so great, that the problem finally got the time of day and a "proper" solution was implemented.
These things can't/won't replace "real" developers any time soon. But they do provide a path to a very under-served segment of knowledge workers.
Personally, I think Microsoft could KILL if they made Excel more robust. I'm not talking MS Access here - I'm talking about a modern, cloud-native app dev platform that is centered on Excel. They already enjoy ubiquity, and they could change the game.
There's a reason that the onboarding experience for most low-code platforms is "Choose a spreadsheet to get started".
But just to provide a counter-point (your comment is a bit catty): If it does, then people should put this on their resume, and that's perfectly legitimate.
Large vendors like SalesForce, ServiceNow and even smaller players like Mendix, OutSystems, etc. do create new job markets.
Go look up salaries for SalesForce Developer or ServiceNow Consultant. These are roles that one can learn without a full-stack background, but often command extremely high salaries because for better or worse, it's still a niche skillset, and someone has to manage those apps that someone built.
Developers have to start somewhere, and most aren't going into their first job interviews having previously written full-stack apps that weren't Twitter clones or thrown together for a hackathon.
If "a half billion apps will be created", isn't the conclusion that we need fewer software builders? A half billion apps sound way more than needed....
It's far more accurate to say, we need fewer software builders.
This is normal, there will be more and new software products.
It's much cheaper to open a restaurant, then build a word processor that someone can get value out of. It's much harder to build your restaurant chain to serve 100,000 people, then it is to scale your word processor, that can serve 100 people, to serve 100,000 people.
If you have too few users, this means that you can't afford to eat, while too many users means that other app developers can't afford to eat.
I also could use a half billion units sold.
'Half a billion' figuratively speaking is not much, if you account for all internal apps in enterprise (ex approval flow for weekly cafeteria menu). If there are millions of small and medium enterprise then there would be millions of cafeteria menu app.
I'd count that into the apps (and workflows) we don't need.
I built a CMS for my wife's business and she's used it to manage all her clients accounts for the last 3 years.
Took me all of an afternoon to figure out. It's been pretty solid, and incredibly low maintenance for me.