Nobody Would Buy Glue Without Things to Put Together: A Story About IFTTT
Greg Leffler
Director of Developer Evangelism at Splunk. Former SRE Leader and Editor at Large at LinkedIn.
This post tells the story of exactly how weird businesses on the Internet are and how people who run them can so aggressively misunderstand exactly what their business does.
To understand my confusion, you need to know about a platform called IFTTT. IFTTT is a service that lets you automate doing things online. As an example, every tweet you favorite on Twitter could be automatically added to your Facebook account, so that all your Facebook friends could verify what great taste you have in tweets. It’s pretty cool, because it means that you can send stuff between services easily, even if the two services weren’t made to work with each other.
An email was sent yesterday to users of the online bookmarking service Pinboard telling them that IFTTT integration was being disabled in a week, without any real explanation as to why this decision was being made. Well, that’s a lie, there was some generic blame being cast upon Maciej Ceg?owski (the developer behind Pinboard) for not doing enough work to re-integrate Pinboard with IFTTT:
Recently, we've worked with our partners to migrate to the improved platform, but some have chosen not to do so. Unfortunately, the Pinboard Channel did not migrate to the new platform and will be removed on April 4th.
Keep in mind, this message was sent on March 28th, in the late evening. We (IFTTT users) got exactly a week's notice that this service was being discontinued, which some might argue is generous since we don't (and can't!) pay for IFTTT. Fair enough. However, I argue that IFTTT has made a critical error in their most recent pivot, and that they don't really understand what their business even is.
IFTTT is a 'glue company'. This means that they don't really have a product themselves. Their product is connecting your other products together in a way that is sane and easy to use. Without these other products integrating into their platform, they don't have a product to sell at all. After all, who wants to connect nothing to nothing?
IFTTT initially recognized this, and wrote a bunch of code to integrate APIs (for my nontechnical readers, APIs are basically automated ways to get data out of or into a service) that were completely disparate and to convert data between formats and to otherwise add a ton of value making it easier to connect services together. IFTTT was responsible for figuring out when one of those services' APIs changed, and for modifying their platform code to deal with that. All was right with the world -- the service owners got to have the benefits of this integration without extra work, IFTTT got to create a compelling product for people to use, and users of IFTTT were able to automate more of their life across different platforms.
All of that came crashing down with this latest change to IFTTT's business model, that's explained in a blog post by Maciej. Now, instead of IFTTT doing the work to keep up to date with changes to the APIs provided by the services they connect, they want the service operators to instead deal with changes IFTTT makes to its API. This makes no sense!
IFTTT's entire reason to exist is "dealing with unpredictable APIs". They take on the risk of dealing with the fact that sites can deprecate or remove or upgrade their APIs at any moment, and they should be responsible for doing the work to make sure their platform works with the services their users want to connect to. In exchange for taking on this risk, they offer a valuable product to their users who they could charge or advertise to or whatever else they want to do to try to make money from them.
Users of Pinboard reacted with support for Maciej's position of not wanting to be beholden to IFTTT in a post made to Hacker News, and all was going well until Linden Tibbets, the CEO of IFTTT showed up, and posted this:
At IFTTT, we've been on the receiving end of platform changes too many times to count. I want to make sure we do it better. Pinboard is a beloved service. Every service is, on IFTTT or not. We care about the people who use them and build them. The changes we are asking for are indeed more work, but we know they will lead to a better Pinboard Channel on IFTTT in the long term.
Well... yes! The entire point of your product is that you do the work to integrate the services! What benefit do you provide to the service owners that would merit suddenly shifting all the work around to them? Why should I, as the hypothetical showrunner at Pinboard, spend time doing work to integrate with your service when it's you who need me?
I legitimately don't understand this viewpoint, but I think it's representative of the viewpoint a lot of Valley businesses have. Shifting away the work to "some other guy" is always an appealing thing to suggest, but in the case of a glue company, what's left for your company to provide? How do you add value and differentiate yourself?
I don't know the answers to these questions off the top of my head, but I think the next great problem in Silicon Valley is going to revolve around answering them. Too many Valley companies are glue (or as Steve Jobs once dismissively called Dropbox, merely 'features') to ignore the conundrum of figuring out how to operate and make money from putting together other peoples' stuff.
Lead Developer at Farm Credit Mid-America
8 年I think it's possible that IFTTT is testing out the waters, flexing its muscle. As other commenters have pointed out IFTTT does add some value to the products it integrates. As a business model, showing up to the party a harmless integrator gets them in the door, their service then attracts users and suddenly they become somewhat important. What power they actually have may be the question this move ends up addressing.
Expert Architect bei AXA
8 年The IFTTT channel count could be a good indicator to see whether IFTTT will get away with this or not.
--
8 年Sir you are 100% correct! "Shifting away the work to "some other guy" is always an appealing..." and with work comes cost! It happens in all area of our lives, ex. some wireless companies send their AD's to costumers cells applying time charge. Are the consumers aware? The fact is that more Companies want to charge up front for building up their companies. The cost for Job Applications and Background Checks are part of doing Business, yet, its pass on to the applicant! So, ifttt is doing what everybody is doing, from the Top Down, passing cost down to those who can lest afford it, by shifting its R&D to others. Not Good For Its Business In The Long Run!
Strive to improve the world by your doing
8 年I understand your point: Why should I invest time for free to make your product work? This would only make sense if your product is selling in the millions and I can profit from this. But overall it is a quite common issue If I get it right: Splendid new features are more important than backward compatibility. I see it more often in the Open Source arena, but it happens everywhere.
human bean at earth?
8 年I really like the concept of IFTTT for personal use. I'm a developer, which means all the things IFTTT does, I could do with scripting, except that they take the technical issues out of the way and it is enough for what I intend to do, on my personal use. It just makes the well known concept of task automation, that you know all about if you are a technical person, available to everybody, with an interface simple enough so that every one can use it. I don't see how people can disagree with these kinds of concepts that make you waste less time on useless tasks and be more productive. It is true nowadays that all services are scattered among multiple platforms because of competition between companies and we cannot stop this from happening, but instead we can try to tie all these platforms together thanks to the glue that IFTTT provides.