After 5 years of waiting for the ability to link to multiple databases, Airtable now allows users to to link to one table from another base! Wait. What…One table from another base? Seriously? It’s very disappointing.
I agree with MIMs that the pro-plan cost wise does not scale well and becomes cost prohibitive after a few users are added compared to other solutions. While I love the simplicity of Airtable, trying to maintain a normalized database becomes very challenging once more than a handful of tables are added. For this reason, I’ve been slowing moving everything over to Notion since they allow linking between databases and because they offer a very customizable “front end” and organizational system.
Then there’s the elephant in the room, Microsoft. With the release of Lists for Teams, Airtable will lose a bunch of enterprise users.
You can tell a lot about a companies culture and leadership when they makes excuses for not doing something because it is hard. I have no ill will towards Airtable and wish them all the best but after 5 years of waiting, its time to move on to a better solution.
And security-minded folks like @ScottWorld might feel that this is a good thing.
You recall this one - one of the recurring surprises by new users:
The number one issue that new users (i.e., those posting comments for the first time) is the lack of fields that also support integrated formulas. This one catches almost every new user by surprise. They expect that a given field should also be able to be set with a formula like Excel and Google Sheets,
My name is Michaela and I work as a Solutions Specialist at Integromat. We use Airtable internally, so we know the product really well and we are absolutely in love with it. However - just like all of you in this thread - we were missing the feature of linking bases in Airtable. We looked into the issue and came up with a solution using Integromat - our very own integration platform. Since this is such a huge feature request, we published a detailed blog post about it, and here I am sharing it with you in case you want to check it out
It’s disappointing really, how these kind of wide asked functionalities have to be solved by external parties like Zapier, Integromat, miniExtensions, … And if only the people at Airtable would be more involved in their own community.
Indeed, but let’s be clear, these are not really solutions.
Unfortunately, this would mean that the place I edit the date would be different than the place I view the date (said here)
Like the comment above, the user has instinctively realized that copies of data are typically a bad thing because they require synching and synching invokes latency as well as incongruous data sets. Ultimately, it ends in misery for users and solution-makers.
Show of hands - who has experienced a synching process that was ever flawless?
I’ll take that as zero hands up.
An approach that sustains data truth - a single, immutable reference to data values - is always preferred and most database platforms support this goal. In every business, we seek to establish authoritative data - ground truths that are reliable and consistently maintained.
Show of hands - how many of you have created an Excel spreadsheet, shared it to three workers and then had to merge the data to create a unified version of the truth?
I’ll take that as 100% hands up.
Airtable, however, created bases to be functionally like apps - little islands of great UI functionality and data. The designers envisioned that connecting these “apps” would be no different than connecting other apps - the API would be ideal.
Boy, were they wrong. Not about the architecture; the concept is fine and the API is perfect. They under estimated the market which immediately assumed that bases were not like apps; they were simply like other spreadsheets, and what do we do with spreadsheets that need data from other spreadsheets? We copy the data of course.
Airtable is not the problem; users are the problem. Users make a lot of assumptions about their own narcissistic ways they would like to see Airtable work and then they proceed to use any manner of Goldbergian buffoonery to enforce their will. This is not necessarily irrational thinking; we all do this with technology and we’re disappointed when copies of data just can’t magically appear in a different “app”.
If it would only do this one thing - it would be perfect.
In some ways, I believe Airtable’s resistance to caving on some of these asks is both intentional and actually doing many of you favours. It’s forcing you to consider your data architecture and information interchange requirements, an investment in solutions that largely never existed when y’all were managing your data in Excel.
Send the records that require design to Base 2
The Integromat approach is certainly interesting, but it does require copies of records. And to be clear this approach doesn’t really “send” the records from one base to another; they are being replicated and any change in Base 2 will not be reflected in the version in Base 1 unless you endeavour to create another reverse replication process. This is added cost and complexity and likely to end in misery.
The remedy, of course, is to try to avoid copying data. Logical references should always trump physical ones. There are certainly some exceptions; not many though.
In 1981 I worked on a simple tool that NASA needed that was designed to move (i.e., copy) data from a desktop PC to a Grid Compass ruggedized laptop (yes, one of the very early ones - I am that ancient). The data was destined for low-earth-orbit and connectivity at scale wasn’t so good back then. The data needed to exist in two physical locations; no one could debate this. BTW, that tool went on to become LapLink.
I think this is just a UX problem, that can be solved with low effort. Just add a tag into the table attributes, which means “Category of tables”. Then see tables grouped into these tags, looking like sets. Because, the problem is mainy UX, in the sense that if you have a numerous tables in your base, you get confused by so many tables, if you can just gorup them it would solve the problem.
Seeing the level of “trainings”, video’s and, well, their marketing, it seems that Airtable is aiming for that kind of customer base. John Doe who is tired of Excel, but doesn’t necessarily know anything about database management/scripting/… Don’t be surprised then when they have basic expectations that maybe aren’t good practice.
Possible solution: train/educate/inform your customers instead of just not responding to their -in their minds- valid questions and expectations.
Perhaps, but it’s difficult to say what [exactly] Airtable is aiming for; they’re not very transparent about their roadmap or ideal hopes. And it’s a free market; they can pursue whatever strategy makes sense to them and their shareholders. If it’s customers who repeatedly get a bad taste in their mouths when they hit dead ends causing the company to rely on customer churn for growth, that’s one business model. I don’t like it, but I don’t own the company either.
In any case, the no/low-code realm is difficult at the outset so we must be understanding at times, and yet also be purveyors of tough love. The amount of tough love is directly proportional to the risk of helping users develop extremely brittle and unsustainable solutions.
As professionals, if we’re going to bloviate in this forum, we probably have a duty to guide users with sometimes harsh assessments; equal parts bloviation/harshness I guess.
I like this idea, but it doesn’t solve some of the key reasons that users seek out the use of bases as connected islands. Sometimes they do this simply to overcome weaknesses in the security model - i.e., a subset copy of data can be shared with certain personas without exposing all of the data.
Not long ago, Airtable presented itself a no-code spreadsheet replacement. Now, Airtable presents itself as a platform to create “solutions”, and the no-code and spreadsheet-like aspects of Airtable are no longer front and center. Does this indicate a change in Airtable’s goals or are these two stepping stone in a larger plan? I’m inclined to think that latter.