Worried that AirTable is going downhill

I, too, was quite thrilled with Evernote, a distant memory of being quite delighted with the product. And then it changed in almost imperceptible ways at first. Over time, I wondered if I wasn’t the frog in the experiment to see if I’d jump out of the pot before it got hot enough to kill me. That’s a myth BTW - the frog will jump out once the water becomes too hot. And like the frog, I bailed on Evernote, but not fast enough.

As to Airtable, I also see signs that are similar to long-lost software apps. If you live long, you’ll start to see the signals of decline, and while I’m sensitized to what you’re suggesting, there is one signal I’m not seeing – what looked and felt like some ideal features that hooked me on Airtable, still seem to be there. There’s an undeniable consistency in the core UI and functionality that was present in 2017 and remains today.

All successful apps and companies must adapt and change from time to time. However, my fear is more centred on evolving the underlying architecture fast enough to reflect the requirements of scale and enterprise suitability. How many decades will we wait for a Split() formula, or the obvious need to open the platform to support user-defined functions? These are examples of change that cannot come fast enough and will be absolute certainties for platform longevity.

Since you kicked open this door, you’ll now have to be very specific about what troubles you. :winking_face:

Hi Bill,

Thank you so much for this response. I’m glad to be able to talk about this frankly.

The things I have seen are very small and perhaps I’m being oversensitive, but they do reminder of me of other apps that declined. In no order, then are:

• The add button doesn’t fit quite into the design, and then you get a window of options that are half implemented IMO
• The dropdown menu is styled differently on some elements
• The background color on Kanban view was recently changed to a shade of grey, and it’s now harder to distinguish cards. A small but important UI change that is very noticable
• Interfaces has two entries - fist the interface base, then the interface view - that’s two redundant clicks
• Going back to the base from the interface uses a subtle difference in design languagr]
• The frequent “suggested for you” is like Evernote offering me old features I’ve never wanted
• I see more and more “Contact Sales” or options to upgrade to Enterprise, even though I’m on Pro level and a solopremneur using it just for myself
• the back button in Mange fields (and formerly in automations) is on the left side of the screen, utterly the opposite side of where it should be

I feel very pedantic. I’m not talking here about features I want that they don’t have (although, I am desperate for a better page writing thing, so that I can make better notes, like Notion). I’m just mentioning the little things that seem to indicate a slight sloppiness on a platform that for years I always held as being very tight.

And so it worries me, because Evernote was tight. And then it just became an ungoldly mess, one little UI slip after another

Yep - it would be nice, but Airtable has carved its niche without giving deference to content creation or management which is why my team at Stream It uses Coda, not Airtable. And the entire premise of Airtable is about tables with data, so we can’t indict them for remaining true to that conviction. Whether that’s a smart strategy or not remains to be seen. Lack of it certainly didn’t help in Stream It’s decision process.

It’s the “kitchen sink” indicator; once they feel compelled to toss in everything including the kitchen sink, you best be looking for the EXIT sign. Back in 2017 the ungodly mess was represented by the vast array of Goldbergian workarounds and hastily crafted Zapier recipes. This allowed Airtable to remain relatively simple. Then the other shoe dropped; they realized that (for many reasons) a number of workarounds had to be made less workaroundy. Ergo - complexities representing the true requirements that users had in mind needed to make their debut in the UI. Now, the ungodly mess has come home to roost, and it will likely get worse before it gets better.

Can they overcome the mess that is brewing?

Your points are all good indicators of the slippery slope to product obscurity, and @ScottWorld will be quick to add another 200+ items to your list. :winking_face:

I can’t go down the rabbit hole with you here, because it will eat up my entire day, but your instincts are 1,000% spot-on. That’s 10x more than 100%.

Thanks Scott, good to know, or is it, not good to know?!?!

Thanks for this Bill. I like your Goldbergian point :slightly_smiling_face:

I’ve come across @ScottWorld’s stuff before and used some of his formulas

I have lovely AirTable personal CRM setup. It runs so much for me in a very automated way (I have it hooked up to monitor my email, update my contacts, manage calendars, automate my accounts, email me reminders, etc etc etc)

But, I of course get a bit worried it might all go south!

FileMaker is amazingly still around 40 years later.

But yeah, I like Bill’s idea of having a grab-n-go bag, so you can flee at a moment’s notice. You may even want to use On2Air: Backups so your data is already waiting for you somewhere else, in case we don’t even get a moment’s notice.

I’ve been in the software industry since I was 12 years old, and I am not personally thrilled with what I see with Airtable’s internal operations.

On the other hand, Airtable has over $1.4 billion in funding, so that could sustain any downhill decline for a while.

It depends on what your needs are. Bespoke is very expensive and takes a lengthy time to build & maintain. Airtable is quick, inexpensive, fun, simple, and can easily be extended to communicate with most apps on the web using its API and a tool like Make.com.

I love Make. Yeah, if you already have a lovely setup, I would just keep everything going as-is. But just be cautiously aware, keep your defenses up, and be ready to flee if necessary.

How many don’t go south? Can you look at your business (or personal) process and find evidence of long-standing apps or services that have been sustained for at least a decade? I can see only one and despite having coined the phrase, Email is Where Knowledge Goes to Die, it is GMail.

I switched to GMail in 2005, but I should have switched April Fools Day 2004 when it was launched.

There are other apps that have sustained well; Excel and Outlook for starters. But when it comes to database apps, they don’t really stay around long. From VisiCalc, to VisiPlan, Quattro Pro, dBASE, Access, FoxPro, and Clipper; they all see their days dwindle relatively quickly. One Internet year, of course, is like 12 lifespan years.

This leaves only one known - change. Build your systems in ways that make it possible to grab a go-bag and move on. Since Airtable formulas and scripts are locked into the platform, the go-bag will probably have only your data. Ergo - best to think carefully about building deep dependencies on a specific architecture. Easier said than done.

Build your systems in ways that make it possible to grab a go-bag and move on. Since Airtable formulas and scripts are locked into the platform, the go-bag will probably have only your data. Ergo - best to think carefully about building deep dependencies on a specific architecture. Easier said than done.

Certainly ensuring your data can be preserved and exported is essential. I only agree to a certain extent with a hesitancy to adopt an architecture though. Absolutely, thought should be put into the initial choice, to ensure you are selecting an application that meets your needs better than the alternatives. But if you always keep every application you use at the end of a 6-foot pole, to prevent “building deep dependencies”, you will never get out of the application the thing you adopted it for in the first place – solutions to your architecture needs.

You have to build dependencies on something. You do the best you can to make a good choice based on research. Once you make your choice though, it makes every bit of sense to squeeze every bit of juice out of the application (and adjacent applications) as possible to make it work for you.

As you say, @Bill.French, limiting your dependency is easier said than done. You can’t sit around waiting for the perfect piece of software to come along – you’ll die, and your business along with you, before you get that software. You’re going to have to commit to something, and it’s going to hurt if you have to switch down the road. This is just a reality we have to live with. But we should still count it a blessing that the same environment that brings us this pain is the environment that also brings us the joy and drive of innovation that makes all of these incredible options we have a reality.

@Scott_Gould, I’d imagine that Airtable is probably still in the top 5 options available to you for meeting your business needs, regardless of cost, and when factoring in cost, is easily in the top 3. That being the case, I’d agree with @ScottWorld’s general advice to keep on at it, and make data backups. But perhaps to add just a little to that, maybe also document your processes and workflows outside of Airtable as well, and only in general terms of what those workflows accomplish, not how they accomplish it with your specific tools right now. That way, should you have to make a switch in the future, you are able to remember in what ways you had your data connected to processes, and are able to rebuild those workflows in another architecture (don’t know about you, but I’d forget all the workflows I had in place).

Build your systems in ways that make it possible to grab a go-bag and move on. Since Airtable formulas and scripts are locked into the platform, the go-bag will probably have only your data. Ergo - best to think carefully about building deep dependencies on a specific architecture. Easier said than done.

Certainly ensuring your data can be preserved and exported is essential. I only agree to a certain extent with a hesitancy to adopt an architecture though. Absolutely, thought should be put into the initial choice, to ensure you are selecting an application that meets your needs better than the alternatives. But if you always keep every application you use at the end of a 6-foot pole, to prevent “building deep dependencies”, you will never get out of the application the thing you adopted it for in the first place – solutions to your architecture needs.

You have to build dependencies on something. You do the best you can to make a good choice based on research. Once you make your choice though, it makes every bit of sense to squeeze every bit of juice out of the application (and adjacent applications) as possible to make it work for you.

As you say, @Bill.French, limiting your dependency is easier said than done. You can’t sit around waiting for the perfect piece of software to come along – you’ll die, and your business along with you, before you get that software. You’re going to have to commit to something, and it’s going to hurt if you have to switch down the road. This is just a reality we have to live with. But we should still count it a blessing that the same environment that brings us this pain is the environment that also brings us the joy and drive of innovation that makes all of these incredible options we have a reality.

@Scott_Gould, I’d imagine that Airtable is probably still in the top 5 options available to you for meeting your business needs, regardless of cost, and when factoring in cost, is easily in the top 3. That being the case, I’d agree with @ScottWorld’s general advice to keep on at it, and make data backups. But perhaps to add just a little to that, maybe also document your processes and workflows outside of Airtable as well, and only in general terms of what those workflows accomplish, not how they accomplish it with your specific tools right now. That way, should you have to make a switch in the future, you are able to remember in what ways you had your data connected to processes, and are able to rebuild those workflows in another architecture (don’t know about you, but I’d forget all the workflows I had in place).

Build your systems in ways that make it possible to grab a go-bag and move on. Since Airtable formulas and scripts are locked into the platform, the go-bag will probably have only your data. Ergo - best to think carefully about building deep dependencies on a specific architecture. Easier said than done.

Certainly ensuring your data can be preserved and exported is essential. I only agree to a certain extent with a hesitancy to adopt an architecture though. Absolutely, thought should be put into the initial choice, to ensure you are selecting an application that meets your needs better than the alternatives. But if you always keep every application you use at the end of a 6-foot pole, to prevent “building deep dependencies”, you will never get out of the application the thing you adopted it for in the first place – solutions to your architecture needs.

You have to build dependencies on something. You do the best you can to make a good choice based on research. Once you make your choice though, it makes every bit of sense to squeeze every bit of juice out of the application (and adjacent applications) as possible to make it work for you.

As you say, @Bill.French, limiting your dependency is easier said than done. You can’t sit around waiting for the perfect piece of software to come along – you’ll die, and your business along with you, before you get that software. You’re going to have to commit to something, and it’s going to hurt if you have to switch down the road. This is just a reality we have to live with. But we should still count it a blessing that the same environment that brings us this pain is the environment that also brings us the joy and drive of innovation that makes all of these incredible options we have a reality.

@Scott_Gould, I’d imagine that Airtable is probably still in the top 5 options available to you for meeting your business needs, regardless of cost, and when factoring in cost, is easily in the top 3. That being the case, I’d agree with @ScottWorld’s general advice to keep on at it, and make data backups. But perhaps to add just a little to that, maybe also document your processes and workflows outside of Airtable as well, and only in general terms of what those workflows accomplish, not how they accomplish it with your specific tools right now. That way, should you have to make a switch in the future, you are able to remember in what ways you had your data connected to processes, and are able to rebuild those workflows in another architecture (don’t know about you, but I’d forget all the workflows I had in place).

To clarify, there are a few subtleties worth a finer point. I’m not suggesting anyone not commit or create deep dependencies; far from it. I am saying that switching costs can be predicted and weighed against the benefits of putting down deep roots in Airtable. For small firms and systems without lots of complexities, the switching costs could be mitigated with little effort despite how deep the roots go. For larger teams and processes with mission-critical requirements, these costs can be huge - indeed, large enough to make a big financial dent in the bank account and your resume. :winking_face: I’ve seen career-ending moves at the end of a technology cul de sac. So with that, it’s wise to be cautious and transparent about worst-case outcomes.

To be clear, I said neither of these things. I’m not suggesting always avoiding building deep dependencies or waiting for the perfect solution. What I said was:

… think carefully about building deep dependencies on a specific architecture.

This is simply a suggestion; if you spend some time carefully designing your [deeper] Airtable dependencies, there are many ways to insulate them from the impact of a wholesale change to another platform. Here are a few examples …

1. You need to convey data from Airtable to another system. Do you build a bunch of Airtable SDK scripts? Or, do you establish a simple outbound webhook? The SDK script is not likely portable, but the webhook approach is very likely to be portable.

2. You need to run an algorithm that is unique and core to your business process. Do you create some Airtable-bound code to perform this computation? Or, should you consider building your own external API that is able to float on any vendor’s no-code platform?

Again, just saying - these are the considerations that I recommend, not that I necessarily know should always be chosen in every case.

You’re right — I mis-characterized what you said by placing it in the context of my reaction (which was not its original or intended context). I’m sorry for that. I wasn’t disagreeing with you so much as reacting to what I saw as an idea that could potentially drive an already despairing Airtable user to despair even more.

Even though you neither said, nor intimated, that building deep dependencies in any given software tool is a bad idea, my mind saw that as a potential interpretation by readers feeling the pressure of their decision to invest in Airtable’s architecture. I know what that pressure feels like, and so my protective half wanted to lend some assurance on the other side that a bet on Airtable was not a bad bet.

I really don’t have a whole lot of stake in Airtable other than enjoying using it for personal projects, so it’s not that I’m trying to defend Airtable, right or wrong, my Airtable. It’s just that I think using Airtable as a business platform was, and still is, a great decision for most people, despite its shortcomings.

Again, I’m sorry for having taken your words out of context with those scare-quotes. That was uncharitable.