Story #1

I'm a serial note-taker. I've used dozens of apps to improve my note-taking: tag information, simplify formatting, improve navigation, etc. I wasn't limiting myself to free software only: paying for software was never an issue - as long as the software was adding some real value. Some of these apps were real eye candies, customisable, with multiple plug-ins, client apps on every device type, myriads of new functions added monthly or so & of course, the obligatory "dark mode" ;P

Recently, I ditched them all to use Roam. But no, it's not a blog post about Roam (even if it's so good that it deserves one). The whole point is that Roam ...

  • ... doesn't have client apps on any device but web browser ;P
  • ... has a crude and simple interface in one freaking color scheme (unless you want to play with CSS overriding ;P)
  • ... has a graph view that has some glitches
  • ... sometimes loads your data even for few seconds, regardless of how much you've put there
  • ... is about three times more expensive (monthly) than top note-taking apps I know
  • ... feature-wise is nowhere close to any of its well-established predecessors

But all of that doesn't matter.

Roam does just ONE thing, but does it BY FAR best - it provides ultra-intuitive, fast & symmetrical approach to linking your notes. Links are bi-directional AND Roam can also find non-referenced links (something that looks like a link but was not marked as such). Everything else in that app is just a detail built UPON THAT foundation.

Roam creators didn't follow with adding f*cking skin support or embedding funny videos from Giphy. Instead, they've created queries to search through references and graphs to visualise their mesh. That way they simply gave their users malleable thought-workspace to create their own insights in the intuitive, flexible & quick way.

And that makes a hell of a difference. That's why I've decided to pay more for slow, a bit buggy, crude, plain Roam instead of battle-proven, polished, native clones of traditional paper notebook. It has exactly the feature I need. And doesn't have exactly all those features I don't need (& I don't want to see/pay for/etc.).

It's just one example. But I have tons of them.

Story #2

Have your heard about Enterprise Architect (EA)? It's a modeling tool for software architects - one I've started using around 2001/2002 & I was still using occasionally even about 2-3 years ago. Over the years, it has accumulated all the possible stencils, notations, templates. Whatever you put in, it's kept in a metadata "database", so the elements of models you create can be re-used in different views (perspectives) or on a different level of detail. You drag elements of the ribbon (old Office style) interface wherever you want them. There are millions of property dialog windows for everything.

I've finally ditched EA around 2018 - I've moved to Miro. Miro is radically different. It has very few shapes, just post-it-like "stickers" & a bunch of basic geometric figures: boxes, ovals, etc. Miro has no metadata database. Miro supports neither UML nor BPMN.

But Miro does ONE thing exceptionally well (which makes me spend 15 dead Murrican presidents monthly on it). Essentially it's a whiteboard, devoid of all EA's features and featurettes. But it enables ultra-smooth, in-browser on-line collaboration in real-freaking-time. And design/modeling is primarily a team sport. Whilst, in almost 20 years, creators of EA couldn't figure anything better than a shared Access database ...

In Miro, it takes me 5 seconds to publish an infinite, zoomable flat workspace for anyone I want. Then we collaborate in a visual way, seeing each other's cursors, comments, changes. Fifteen minute-long collaboration sessions in Miro solve design problems that used to take days of detailed model crafting before. Literally.

Again: Miro is simple. Miro doesn't try to do a million things. It does just ONE thing but does it EXTREMELY well. I have only one comment:


There are far more "stories" like those. Another great example is ("big") Visual Studio. Microsoft keeps adding new features edition by edition, year after year. I was always checking them out carefully first thing after the release & ... never used them again, because they were so useless (or already present in ReSharper for ages) ;P

OK, enough. I think you've got the idea.

Featuresteins

How can we summarize that?

Well, product management is freaking hard. I had a pleasure to observe several products & product-oriented companies in the last few years. All flavors of them: B2B, B2C, global/local, of different business models, varying characteristics of revenue streams, etc. Majority of them suffered from the same ailment:

They were producing "featuresteins" - amalgamates of features (of questionable usability ...) that have been added once upon a time, because:

  • a local "visionary" had some sort of an epiphany (& he can't be wrong ;P)
  • "well, we had no clue what to do next, but we had to try something ..."
  • everyone else has something like that ("dark mode! gief dark mode!")
  • "it's coherent with our previous design decisions" (not validated with actual usage data)
  • one client wanted that, and their guys were more assertive ...
  • it seems that the P&L on that will be positive (which sounds like scrapping some low margin profitability ...)

Featuresteins are the harbingers of mediocrity.

Why so?

  1. Because they kill focus (of the product development teams, that are supposed to tackle the core domain ONLY).
  2. Because they threaten the clarity of product's vision (what problems does it solve, where's its niche, who's the target audience)
  3. Because clients/users don't really care about them. Did you see how much time it took me to jump to Roam/Miro? F*ck "brand loyalty", I just want USEFUL stuff.
  4. Code (features) is a liability. Increases inertia and slows you down.

NEVER offer stuff just because you CAN offer it (when it comes to digital products, customers no longer follow the saying "never look a gift horse in the mouth") - this is how you over-bloat your product. People are generally okay with paying MORE for a tool with just ONE feature, but a world-class one - one that makes a difference, solves them a real problem, changes the rules of the game for them.