Airtable, Retool, and steady state

In the path for THE app builder and workflow automation tool, Airtable and Retool will find each other competing for the same market.

Are some things inevitable?

Airtable and Retool in the same sentence sounds odd. At first glance, you wouldn’t assume they are competitors. Most people don’t. And why would they? Today, Airtable is clearly targeted towards non-technical users, no-code developers, and people in marketing and business operations. Retool, on the other hand, is the king of internal app builders, focused squarely on developers.

For all intents and purposes, Airtable and Retool are not competitors. Nothing in the design, product, or marketing messaging indicates this. But then marketing messages can quickly change and landing pages and be quickly updated, sometimes even on a Friday afternoon at 4.38pm.

So why do I think they will compete for the same TAM?

The short answer involves both companies reaching the same steady state, converging (or colliding?) into the same market and both eventually transition to become a low-code internal app builder for data workflows and automations. They will each dominate their current market position and their primary way to grow/expand will be to move along the no-code ←→ developer spectrum in opposing directions.

The longer answer jokingly involves users, product managers, and aggressive focus (or expectation¹) on growth. It starts like this: users ask for more “capability”. Product managers in both companies will listen to their users and build them what they want, until eventually you end up with a product that does everything for everybody. That tool would be an all-purpose tool that’s easy to get started with, yet flexible enough to be extended and customized to suit ever-evolving use cases.

Airtable users will ask for more flexibility (and more rows!²). Retool users will ask for more user friendliness and for it to be more accessible for people who aren’t as technical. Airtable and Retool will give people what they want. This means Airtable will add increased flexibility, better APIs, more integrations. And Retool will add more built-in UIs, components, and app templates. And both companies will continue to build workflow, automation, and scheduling tools.

But if you look no further than at the recent product launches you’ll start to see that Airtable and Retool start colliding into the same market, use cases, and value proposition.

For example, Airtable has started focusing more on APIs, developers³, partnerships, workflows and automations. Retool is adding Retool Databases (basically Postgres) with a nice spreadsheet UI on top⁴, that also integrates with Retool’s internal tool builder out of the box, as in, you won’t even need to type in port number 5432 or any other database creds to connect it. Basically, Airtable will build more developer tools, and Retool will build better UI.⁵

Of course this is not a zero sum game (or market). It’s pretty large, growing, and there’s space for both, and more. But let’s talk about some extreme scenarios if they were to go head-to-head.

Scenario 1: Both win⁶

They both let Tiktokers build data apps and workflow automations in the metaverse.⁷

Just kidding(?).

Here it is.

So Airtable rebuilds their architecture, to cost efficiently allow for larger tables without sacrificing performance. Database tech is getting better. Maybe they can use serverless Postgres, SQLite, DuckDB under the hood. Or maybe all, or some hybrid. I don’t know. Airtable, having seeing hundreds of use cases, could re-build down the stack.⁸ Know which high-value use cases people pay to solve, and build the best backend and low-level stack to solve them elegantly. Building down the stack can sometimes be easier. It takes time, effort, and courage. Airtable becomes the best tool to solve the universe of high value use cases. Airtable cements itself in the no-code app building and workflow automation space. The new stack would better allow Airtable to take the Webflow approach, essentially no-code, but if you really need to, you can code it too.

Retool builds an awesome spreadsheet UI, powered by a proper database, that users can directly access in multiple ways. It can store more than 50k rows without going for an enterprise plan. It adds filter, sort, and grouping. Next, it builds intuitive table relations. Finally, it starts launching pre-built Retool apps or templates that sit on top of this database. You can access the database through an easily generated REST APIs or an autogenerated one, or direct database connections, because it’s just a SQL database under the hood. Retool can now target really low-code devs, and eventually no-code devs. Apps built in Retool are easy to use, but super powerful under the hood, because it’s a whole set of dev tools and frameworks.

Ok, it might also be easier if the two companies M&A each other. Maybe I should have written this one first.

Scenario 2: One or both fail

Product cluttering. The products become bloated and cluttered. Cluttered products in turn become slow, harder to learn, and harder to use for simpler use cases. Flexibility vs simplicity trade-offs.

Unclear marketing. The messaging or positioning isn’t clear about who should use which product for what. Incorrect or over-generalized positioning fails to attract the right people⁹, who may try one of them, realize it isn’t fun, and then give up going this approach altogether and hire an intern with a premium subscription to Github Copilot or some application of LLMs instead.¹⁰

Architectural limitations impacting UX. Airtable fails to build a more flexible architecture that allows for a world where datasets are default bigger. Retool fails to nail the UX and fails to “teach” non-devs to think like devs.

Newer entrants. Startups that nail the platform flexibility and the UX and take over the market. They have the benefit of hindsight, understand the limitations of both products, and can build an architecture that addresses them and a UX that just makes sense. Realistically speaking this won’t be easy, but it’s not impossible either. A startup can nail the flagship use cases for each company and create the workflows or apps out of the box, and even allow users to peek under the hood to modify/customize them.

Open source. I have to call this one out separately because there are already quite a few of them, including Baserow and NocoDB competing on Airtable’s side and Appsmith, Budibase, and ToolJet on Retool’s, just to name a few. Although I believe open source entrants are more of a threat to Retool than they are to Airtable at the moment, there’s a universe where open source is the dominant way to go. After all, companies can self-host and have full control of how and where they deploy these apps, without hitting enterprise tiers, unlike the non-open incumbents.

Microsoft. Microsoft has the magical ability to simply make the TAM of most B2B SaaS startups smaller than they appeared. Although I say Microsoft, I really just mean large companies aggressively entering the space. A Microsoft PM might come along one day and get “inspired” by everything both companies are doing, build it, and audaciously push it to all customers, with quick links and announcements directly from Microsoft Teams and all. Oh wait, I think it’s called Microsoft Power Apps.

Airtable missteps. Airtable’s architecture fails to allow for more flexibility and customization. It also fails to transition and capture the interest of more technical users who need to do more and who may perceive Airtable as a shiny toy.

Retool missteps. Retool fails to get the UX right. Retool is powerful and you can build some great apps with it, but it isn’t immediately obvious how to use it. Even as a developer, I still need to watch a few videos and tutorials to figure things out. As much as I enjoy watching Alex build a Retool app in 4 minutes, there’s a certain philosophy, a subtle framing, or a sort of “developer mental model” that you need to transition yourself into when you build using Retool. A misstep is if Retool doesn’t manage to abstract the power and complexity of its architecture into straightforward use cases that users (devs included) can build very quickly.

What does steady state look like?

In both cases, I think both Airtable and Retool become low-code builders. Airtable will go from no-code to low-code and Retool will go in the opposite direction, from code to low-code. Both will reach a steady state of becoming more of a hybrid low-code app builder, where you can easily build apps and workflows, but you can also peek under the hood to customize and even extend their capabilities beyond those out of the box.

Who has an advantage?

Currently I think Airtable has market advantage (distribution), Retool has technical advantage (product).

Strategically though, Retool may have a slight advantage at the low-code steady state. This could change as products mature, but it’s likely easier for Retool to go from high-flexibility platform towards ease-of-use, better UI, and templated workflows for low-code devs or non-devs. Retool could build simplicity on top of power and flexibility.

On the other hand, Airtable may need to layer complexity at the lower level of the stack, which may require both architectural changes on their end and a shift in mental models on their existing customers’ end. This results in a UX that deviates from their expected notions - those same ones that Airtable tirelessly taught them for years. Architectural changes for a product that has broad distribution, established positioning, and a solidified design language tends to be difficult, if not at least highly disruptive for its customers. Execution will make all the difference in the end though, and I’m confident their teams can get this right, which is why I think Scenario 1 is the likelier case.

Is steady state inevitable?

Maybe it’s not. But it will take an immense amount of effort to fight against the two opposing forces: the desire to grow and the need to go beyond their core offering. Classical business strategy would suggest that each company focus on protecting their core and maintain dominance in their respective areas, even as they continue to expand and grow. In Retool’s case, fully focusing on their core would help fend off exciting new approaches to building internal tools, such as Airplane.¹¹ In practice though, it is hard to both fend innovation and to not get distracted because the everyday decisions that lead to growth tend to also move companies beyond their current offering. One could argue their cores are large enough and will continue to grow so they might never have to move beyond it, but I rarely see this happen in real life.

I’m a big fan of what both of these companies are building and what they have accomplished. They’ve had a huge impact and made life better for lots of people. I’m excited and interested in seeing what happens next, as they start to dominate their respective markets, and look for future ways to grow.

Hope that was a fun read! I had definitely had fun writing it.

Also subscribe so I have a reason to continue experimenting with more blog posts on Substack.

1. Either or both companies going IPO will probably accelerate steady state

2. I hear this over and over again from people I talk to, in forums, and community discussions.

3. Airtable has had an API for a while, but I’ve seen a stronger push more recently. Also I don’t think it had metadata access when it launched. That seems like a more recent release.

4. It’s hard not to be inspired by Airtable.

5. Better UI is relative. Retool’s UI is not bad, but it depends on who the user is. If you are a developer, it’s good; if you are non-technical, it’s bad.

6. I think this is the most likely scenario. It’s a big market, and it’s only getting bigger.

7. They make it dead easy for a TikToker to build a workflow where they push the Polywork profile of every video commenter into a Airtable or Retool-built CRM, which then puts them in a drip campaign to push for a “collaboration” in the metaverse.

8. This is similar to Apple building their own M1 chips. They know what they need those chips to do because they know what billions of people want to do with their desktop and pocket computers.

9. Especially newer generations perhaps, currently aka the TikTok crowd. They have completely different expectations of how they want to do build things and how they want to interact with them. What may be commonly accepted as the standard way, may no longer be it.

10. It won’t be long before I can type “a component with 3 input fields, name, lastname, and email, and a submit button that when pressed maps the records to my database” and I’ll get a nice React form component, with a submit form button, and POST endpoint that writes to the connected database.

11. Which seems to be doing alright and with $32M in new funding, which is quite a few EUR and GBPs these days!