Why IFTTT is ahead of the game

First a bit of history

When I started out, unix pipes were (and still are) superpowerful to automate things:

$ echo "mail -s "$USER: please do a server-update" john@mycompany.com" | at 1145 jan 31

It looks quite humanreadable..kindof :)

Then Yahoo pipes happened

Interactive websites were mostly limited to data-entry & form-processing. Yahoo made great efforts to create an eventful web:

For that time it was very impressive, but somehow this doesn’t look simple :)

SOAP & REST became a thing

The downside however is: it needs code, therefore a developer specialist, database, server infrastructure etc. Automation still doesn’t look simple.

Jeff Lindsay, a friend at Gliderlabs (who coined the term ‘webhooks’) did a great job promoting the idea of ‘the evented web’. Back then, it gave me these revelations:

• ‘polling-data-syndrome’ is a thing
• “Developers like to develop (too much)”
• This pushes a ‘web as a newspaperstand’-illusion to people

IFTTT

IFTTT was already around back then, and demonstrated an ‘evented’ philosophy:

Now lets look at the unix oneliner again:

$ echo "mail -s "$USER: please do a server-update" john@mycompany.com" | at 1145 jan 31

Do you notice the simplicity in both approaches?

• it’s very highlevel: it abstracts code and complexity away
• emphasis on on human language & perception
• we don’t need specialists anymore :)

Now let’s look at the competition

Over the years, lots of competitors have arised (Microsoft Flow, Zapier), ESB’s (mulesoft,JBoss), and opensource solutions like node-red, datafire etc. I’ve even designed and implemented some solutions for customers myself :)

The blind spot of the competitors?

“Abstracting away logic by introducing highlevel scripting to the user, is a solution which creates a new problem”

Mo code, mo problems. Mo yahoo-pipe-ish solutions, mo deutchs limit

Offering automations with a ‘conditional/chainable character unlike IFTTT’ creates new challenges. It’s no longer an adapter-platform (IFTTT), but a programming-silo which introduces all kinds of new risks.

IFTTT seems to be the only one which promotes the eventful web, and a “Don’t build silo’s”-attitude. This might explain why IFTTT is attractive by nondevelopers and developers.

The secret of IFTTT?

I have no idea, but here’s my gut-feeling :)

In order to pull something off like IFTTT, I think 3 things are important:

• understand the human needs
• understand the technical challenges
• create a universal solution for developers and nondevelopers

Imho IFTTT got it right from the beginning, by selling simplicity instead of complexity. It makes something incredibly complex very simple.

Why I think IFTTT is ahead of the game

Where others offer more complex solutions, IFTTT somewhat follows the rule of least power.

IFTTT seems to be layered in a smart, interconnected way:

NOTE: I’ve only noticed layer 3 by accidentally googling it. Ideally, the distinction between layer 2-3 could be more obvious in the IFTTT Platform docs,

I was also impressed by this REST endpoint validator:

This turns a very hairy human process (usually emailing back’n’forth with support) into a bunch of green lights.

In case you’re curious about IFTTT Platform’s onboarding process for applet services: I’ve made this node-ifttt-express repository to setup your own IFTTT Platform applet.

Conclusion

The trigger-action-paradigm offers a more flexible & eventful future for SaaS. IFTTT demonstrates that limiting webintegrations to an adapter-platform (unlike zapier) doesn’t really gets in the way of developers. I hope it will become a standard, and will wake up other companies as well.