There’s a graveyard somewhere with the word Core on it. Relying on a core function is a killer. Think Dropbox. Twitter. Whatsapp. Evernote. All apps heading for Core Graveyard.
Here’s the thing.
In the old days the way to get your app noticed was to offer something no one else did, or did as well/cheaply. This was what everyone told you to do, right back to the idea of the mousetrap. But these days, because business models mostly revolve around freemium and subscriptions, this doesn’t work anymore. And it’s killing a lot of good products because they can’t navigate beyond from their core function without damaging the very thing that people like them for.
Take Evernote. The name is a great name because it describes what people like it for. You can take a note and save it forever. Or you could think of it as a verb, be forever taking notes. But to get people to take notes forever, or keep them forever, you need to offer it for free because there are, at least now, so many other players offering the same thing. So the story of Evernote is one of a company desperately looking to add some additional value to the product to justify charging for it.
The result: they have destroyed what drew users to it in the first place (I’ve been a user since at least 2005.) Evernote has added pretty much everything it can to the core functionality, including physical notebooks you supposedly scan (offering very little you couldn’t do manually as well if not better), but nearly every new feature gnawed away at the core, making the app less efficient at doing what made it appealing in the first place. If your unique selling point is being able to take notes easily, quickly and retrievably, then logically it follows that everything you add to the process actually detracts from its value.
Evernote is now a shadow of its former self. (Check out this Reddit thread for some of the recent pains expressed by users.) But you could have chosen any such thread from the past eight years or so and seen the same problems.
The same story can be repeated with many other apps and services.
Dropbox was once literally that, a digital box you could drop things in to share with others. Now it’s a bloated mishmash of features including an editor, a password vault, eSignatures and a file browser, but doesn’t do the basic syncing thing half as well as other tools, or as it used to. (It’s frustration with Dropbox that is prompting me to write this note.) Dropbox was once the obvious choice for anyone wanted to share files or sync files between computers.
And indeed, here’s the rub. It’s obvious these companies have determined, probably with the help of consultants (yes, I know, I’m one) to try to map out greater territory by defining the function of the app within a broader space. So Dropbox has persuaded itself no longer to focus on the rather prosaic process of sharing files and folders, but on ‘collaboration’. Evernote, similarly, tries to emphasise its place within collaboration, by encouraging users to create and edit files with others inside its (wait for it) ecosystem.
Collaboration is fine, but you need to make sure you’re doing the basics first. Collaboration is not why people use, or used to use, these products. Users wanted to do something quite specific, and were drawn to the app to do that specific thing. They’re not (necessarily) interested in the broader scope, indeed it’s likely they’re already doing that via other apps, but want another product to do some aspect of it that isn’t working well, or at all. But by imposing these new features on the user, they degrade the (wait for it) experience, and make the user less likely to stick with the product if one with a better functionality comes along.
Indeed, the reasons why the likes of Telegram, Quip, Slack and Zoom have all made inroads is that they do one key thing better than what exists: chat, writing documents, team chat and video calls/conferencing. It’s not that these functions weren’t available before these players appeared, it was that no-one did them well. They were either part of a separate suite, or there were things wrong with them because the developer wasn’t laser-focused on how well the tools worked. Quip did a better job than Google Docs; Slack did a better job than the likes of Yammer; Zoom did a better job than Skype.
Let’s all play
You can play this game with more or less every app you come across that isn’t focused in its approach. Twitter is now offering Twitter Blue, a subscription based service which, among other things, allows you to (drum roll) undo a tweet! Yes, if you pay you can actually ‘undo’ a tweet (I understand that this likely means not deleting a tweet, but of delaying the posting of a tweet). Great, so now you basically have to pay to think twice about dashing off a tweet.
Twitter’s appeal was a way to communicate one to several (ok, eventually many) with as little friction as possible. Coupled with that was the ability to see easily, with as little friction as possible, what others were saying — to everyone. Indeed, Twitter is a great example of how an app/service/product is moulded, not by the developers, but by the users. Much of what we think of its core functionality — hashtags, direct messages etc — were added by users, which were eventually adopted by the developers. Twitter has ever since tried to tinker with that model to generate revenue: the default view of tweets now is actually not whom you’re following, or at least not just whom, but tweets that algorithms think you’d be interested in. You have to alter the default to get a chronological list of tweets from people you’re following. Choose that and you’ll notice, from time to time, that Twitter will switch back to the default. In other words, the core function of Twitter has already been undermined in the interests of the company.
Whatsapp is another case in point. Whatsapp enjoyed massive adoption because it provided a useful service — free or cheap SMSs, if you want to put it bluntly. Whatsapp denied that at the time — I interviewed the CEO in 2012, and they were quite upset that this was the thrust of the resulting piece — but I know, I was there, or rather my Indonesian friends were, one of the first countries to really adopt the service. At the time Whatsapp hoped to make money by charging for it, via a subscription at a very cheap price ($1), in the second year. The number of users would have easily given them enough to keep the service going. Of course it didn’t turn out like that, and we’ve seen Facebook crawl through a number of hoops to find ways to monetize the service (by mining the data within.)
It’s true that Whatsapp has evolved hugely since the early days, but in the exact same trajectory SMS did, indicating it is still really just an SMS killer: think groups, think multimedia messages, think desktop app for composing and reading messages, think company communications with users etc. The bottom line: Every effort to move Whatsapp out of its core functionality and more into a data lake or part of a (wait for it) ‘family’ of apps are steps that undermine its core value.
So what does this mean for the future?
Well, to me it’s clear that anything that can be defined as a single, important function that is not already part of a broader app or service should be marketed that way, priced on the assumption that it brings immediate value to the user, and should not be allowed to have extra ‘value’ bolted on to it.
There is a corollary to this. If you think you can add extra substantial value then do it, but don’t ever let it compromise the core. So Dropbox can offer editing services so long as there’s no extra steps, no larger download, no nagging windows to your app’s core value.
And if you’re offering, say, a banking function (transfers) then it makes sense to offer other banking services atop that once you have that core function working well, and that you make sure it’s always easily done, with minimal steps. TransferWise, now Wise, is adding features while at the same time ensuring transferring money is still as easy and cheap as pie.
But don’t let those new functions get in the way. A good example of this failure is Grab, a Southeast Asian ‘superapp’ (be scared whenever you see that word) which started out as offering easier ways to order taxis (it was originally called GrabTaxi.) It now offers all sorts of bells and whistles, and ordering a car, its core business is now a slower, more cumbersome process than it was. Even the map showing where you or your car is is largely covered by an ad, offering insurance for greater peace of mind. The service itself is still good, but the experience of ordering it is not.
Economic imperatives underline these moves. But they needn’t. Such distractions arise from a failure to define the purpose of the app or service. What itch is it, exactly, that you wanted to scratch with this? Failure to frame the USP in this way will always lead to tears or (eventual) bankruptcy.
The world of software is full of small, modestly successful apps that deliver individual functions that make paying for them worthwhile. I’m writing this on Drafts, an excellent app that does one thing: lets you compose stuff and then send it to pretty much every place you want (email it, add it to a calendar, tweet it, save it to Dropbox etc.) Yes, other apps let you do this, but this what the app focuses on doing, and doing it easily and well. The name captures the purpose of the app, and that’s all you need to know.
I am a big user of apps. My instinct, honed by decades in front of computers, is that there are some things that an app can do better. Screenshot? There’s an app for that which does a better job of the purpose I’ve defined. I want an app to remember everything I’ve browsed for the past month, quarter or whatever, whichever browser I use. The answer: HistoryHound. Does it all, for a few bucks. An app which lets me save everything simply and in a way that makes it easy for me to find. The answer to this used to be Evernote, but not anymore, it’s (at least for Mac users) DEVONthink. What changed? Well DEVONthink just kept going trying to figure out how to make the process as effective as possible. Evernote meanwhile added too many bells and whistles, complicated and inconsistent changes to the interface, many of which got in the way of users. DEVONthink isn’t particularly user friendly, but it is effective and efficient. I now have sent to DEVONthink in four years of use about 10 times the stuff I sent to Evernote in 10 years of using it.
But who wants all these itty bitty apps that only do one thing? True. It’s not everyone’s cup of tea. But the big underlying shift, and a reason why I think the future is in small, single-cell apps is because they can be connected to each other via APIs, application programming interfaces, that essentially allow the apps to talk to each other. Drafts allows you to export your writing to other apps, but it also connects you to WordPress, or Dropbox, and to tweak that connectivity. If two apps or services you like don’t directly connect to each other, you can use third party services like Zapier. The only limit is your willingness to tweak a little, and your imagination.
But I also think it’s telling that there are still a lot of artisanal developers still around, still producing these apps. And app stores like Apple’s make it much easier to find them. And then there are players like Setapp, a sort of app aggregator for iOS and the Mac, which consciously tries to gather together a suite of apps that fit most of the niches that you as a user might need, via a $10 monthly subscription, irrespective of how many you use. I asked them why some apps disappear from their list: they told me some go because they are sometimes abandoned or the vendor parts ways. But then Setapp tries to go find a similar one. It’s not for everyone but to me it’s a godsend. Often I’ll find one of their apps does exactly what I’m looking for (Prizmo, for example, lets me capture some text on screen and copies it as text, even if it’s in an image, Permute lets me convert any multimedia file to another format, World Clock Pro shows me times in other timezones whatever the time is at home, Dropshare lets me share individual files with other people much more intuitively and quickly than Dropbox.)
Apps for every occasion
You could argue these are all esoteric functions that not everyone wants. That’s true. But this is how every app and service starts. The usefulness of Twitter wasn’t all that evident to those of us who first saw it. Whatsapp made sense to people living in places where SMSs were a significant expense. Dropbox came into being because people had trouble sharing files. The point is not that the function is esoteric. It’s whether the app or service or product can be defined as fulfilling a purpose. If you can define that and define it clearly, then you’re halfway there. It could be backing up your files. Or being able to walk along looking at your phone and seeing ahead via a transparent background using your back-facing camera. Or an app that tells you when your friends have been vaccinated and so are safe to have dinner with. Or an app that can tell you what lightbulbs you’re using, whether they’re about to break, and what is the cheapest way to replace them. Or an app that identifies whether the snake on your balcony is poisonous or not. Or an app that can tell you when you’ve given too many examples in a piece and how many readers have likely stopped reading. You get the idea.
We took a wrong turn when startups were encouraged to ‘growth hack’ and focus only on number of users, not whether the functionality matched the needs of those users. Sure, we all want to build the next viral sensation, but there’s a lot of other things we use our devices for, and someone needs to build for them.
OK my focus and word-counting apps are telling me to finish this. So I will.
I wonder whether the single-celled future you envision is only possible within the context of platforms. Therefore, if we want a cohesive ecosystem of single-celled apps, we must ask: how can a platform be built that’s incentivized to enable such an ecosystem?
Recent history demonstrates a bunch of platforms being built by single-purpose apps expanding their “mandate” beyond the existing problems they solve. We have an ecosystem of tiny smartphone apps because companies merged phone calls, address books, calendars, messaging, etc into PDAs and smartphones. My reading of computer history makes me think this has always been true — aren’t the tools adherent to the Unix philosophy only made possible by the mondo program that is Unix?
While you discuss a bunch of products/companies essentially committing suicide through their efforts to grow, I think there are only 2 viable paths to product expansion — changing the core and bolting on to the core. A 3rd path, building separately, means you give up the leverage created by your existing product and taking on all the classic problems of conglomerates. Changing the core enables growth in a unified, cohesive fashion, but it risks sacrificing the core’s users in the process. Bolting on to the core makes things less cohesive and creates this dance where you both try to entice existing users to do more with the product without obstructing the existing functionality. Both paths risk destroying an initially successful business, but is there another way? I think it’s clear single-function substitutes quickly enter the market to fill the gap, so the cost to users is one of switching, not losing the functionality altogether.
Is the only path to a more desirable computing environment, then, a new platform that runs its ecosystem in a more open/standardized fashion so the single-celled apps can function cohesively? If one wanted to find and support such a platform, my hypothesis is the place to look is those single-function apps attempting to change their core or bolt-on functionality to expand their serviceable market.
Thanks for the comment Brian, interesting. I guess a couple of things spring to mind:
– Some apps are thinner slices of functionality than others, depending on the itch to be scratched. In your case, Kanopi.io, is scratching the itch of trying to build note-taking around the user’s thought process, and not the other way, the user is going to get more use out of it the more effort/content they put in, which is obviously different from, say, a screenshot-taking app.
In this area, I’ve seen players struggle because either users don’t make full use of the features, and therefore, at least in the eyes of the developer, don’t get the most benefit out of it (I would put TheBrain, DEVONthink, Tinderbox as clear cases, but you could argue even something like Microsoft Word, where users only use about 10% of its features (just they’re always a different 10%), also falls into that category; or that the idea is simple and it works, like (traditional) mindmapping, say, and there’s no need to add more features, raising the question: how do I persuade people to buy upgrades, or keep subscribing, or differentiate from all the other mindmapping apps? Apps like Agenda might also fall into this category. They work fine, but the subscription model they’ve adopted mean they’re always trying to add features, which may end up slowing the app or damaging the simple UX.
– I would love to see the connectivity between apps to work more effectively, because to me that’s the future. I think we’ve just scraped the surface of what is possible. Zapier et al are still way too clunky to be ready for ordinary users, and there’s a scaling problem — the more zaps or API calls you set up, the less you know what is going on with your computer. But apps like PopClip hint at a future where everything naturally flows from the cursor, and, at least in theory I’m freed of the tyranny of an app’s walls. It also offers a glimpse of how single-celled apps can build bridges between them — a PopClip, say, which helps the user find the functionality they need. (Copying some text from an email, and wanting it to pasted as thoughts in a Brain, or added as a footnote in a document, or turned into a shopping list, say.) We’re seeing more apps make use of these APIs (Spark, Airmail etc deserve a mention) but the process is cumbersome.
And as someone who emailed me about this piece pointed out, for a lot of people the extra functions are helpful — in his case, the Password saving feature from Dropbox. But I see that as both a marketing and an OS failure, that a user ends up making that choice. In a perfect world you don’t want your cloud storage provider to be the same person storing your passwords, I would argue. For many users, ease of use will always mean they end up over-relying on ‘department-store apps’ and ‘own-brand secondary apps’ because better options aren’t easy to find. Even with app stores, and their (annoying) ‘magazine-style’ approach.
Thanks for the reply, Jeremy!
The dimension of — functionality available vs accessed by any individual user — is a very interesting one to consider in this context. I have to think more about that. It speaks to the idea that there is no “median person”, as everyone’s needs are spread across many dimensions, some of which are average, some are not.
While I love glue tools like Zapier, and appreciate their value as independent entities from other services, I wish products at each end of a Zap could more directly embed Zapier. I feel like the friction of having to setup Zapier/IFTTT is too much for most users. Services that do manage to grow ask themselves whether they should offer integrations to other products or instead build a deeply integrated competitor, and we both know how that usually goes.
Whether the OS should be a neutral intermediary enabling discoverability of the right tool for every job is an interesting question. You are probably right that the OS should serve this role, at least if I’m understanding correctly your opinion. At the limit, every affordance could come with the ability to select from available providers, similar to those available in some regions for web browsers and search engines.