#and v1 has iterated so much on its function that it's become something else

[cherubchoirs]

[ultrakill]

war & peace (i)

[terrancedkennedy]

I hate MVPs. So do your customers. Make it SLC instead.

Product teams have been repeating the MVP (Minimum Viable Product) mantra for a decade now, without re-evaluating whether it’s the right way to maximize learning while pleasing the customer.

Well, it’s not the best system. It’s selfish and it hurts customers. We don’t build MVPs at WP Engine.

The motivation behind the MVP is still valid:

Build something small, because small things are predictable and inexpensive to test.

Get it into the market quickly, because real learning occurs only when real customers are using a real product)

Trash it if it’s a flop, or invest if it’s a seedling with potential.

MVPs are great for startups and product teams because they maximize validated learning about customers as quickly as possible. But it’s a selfish act.

The problem is that customers hate MVPs. Startups are encouraged by the great Reid Hoffman to “launch early enough that you’re embarrassed by your v1.0 release.” But no customer wants to use an unfinished product that the creators are embarrassed by. Customers want great products they can use now.

MVPs are too M and almost never V. Customers see that, and hate it. It might be great for the product team, but it’s bad for customers. And ultimately, what’s bad for customers is bad for the company.

Fortunately, there’s a better way to build and validate new products. The insight comes by honoring the useful attributes of MVPs, which are listed above, while also giving just as much consideration to the customer’s experience.

In order for the product to be small and delivered quickly, it has to be simple. Customers accept simple products every day. Even if it doesn’t do everything needed, as long as the product never claimed to do more than it does, customers are forgiving. For example, it was okay that early versions Google Docs had only 3% of the features of Microsoft Word, because Docs did a great job at what it was primarily designed for, which is simplicity and real-time collaboration.

Docs was simple, but also complete. This is decidedly different from the classic MVP, which by definition isn’t complete (and in fact is embarrassing). “Simple” is good, “incomplete” is not. The customer should have a genuine desire to use the product, as-is. Not because it’s version 0.1 of something complex, but because it’s version 1.0 of something simple.

It is not contradictory for products to be simple as well as complete. Examples include the first versions of WhatsApp, Snapchat, Stripe, Twilio, Twitter, and Slack. Some of those later expanded to add complexity (Snapchat, Stripe, Slack), whereas some kept it simple as a permanent value (Twitter, WhatsApp). Virgin Air started with just a single route — small, but a complete.

The final ingredient is that the product has to be lovable. People have to want to use it. Products that do less but are loved, are more successful than products which have more features, but that people dislike. The original, very-low-feature, very-highly-loved, hyper-successful early versions of all the products listed in the previous paragraph are examples. The Darwinian success loop of a product is a function of love, not of features.

There are many ways to generate love. “Minimum” and “viable” are definitely not two of those ways. The current-in-vogue way is through design: Elegant UX combined with delightful UI. But there are other ways. The attitude and culture of the company itself can generate love, such as Buffer’s blog with its surprising transparency or MeetEdgar’s blog genuinely helping entrepreneurs or HubSpot’s blog which early on was at least as instrumental to their customers’ success as the actual product. Another way is through a deep connection to the psyche and work-style of customers, like Heroku who broke with marketing tradition by filling the homepage with command-line feature examples instead of benefit-statements, thereby connecting instantly with their geeky target customer:

These are the components of the correct alternative to the MVP: Simple, Lovable and Complete (SLC). At WP Engine we pronounce it “Slick.” As in: “What’s the ‘Slick’ version of your idea?”

Besides the above, there’s another benefit to SLC when you consider what happens with the next version of the product.

A SLC product does not require ongoing development in order to add value. It’s possible that v1 should evolve for years into a v4, but you also have the option of not investing further in the product, yet it still adds value. An MVP that never gets additional investment is just a bad product. An SLC that never gets additional investment is a good, if modest product.

Although not called SLC, there’s a popular meme in product circles that neatly encapsulates the idea of SLC in a diagram: The Modes of Transportation example from the Spotify product team:

A skateboard is a SLC product. It’s faster than walking, it’s simple, many people love it and it’s a complete product that doesn’t need additions to be fun or practical. At the same time, you can evolve the skateboard by adding a stem and handlebars, to create a scooter — only slightly less simple, and definitely loveable and complete. Next, you could grow the wheels, add a seat and some gears, and you have a bike. Again, less simple but now you have a product with massive benefits of speed, distance, and energy-efficiency. Complete, but many accessories available if you choose.

Zooming into one of our examples above, Snapchat took an SLC progression similar to the transportation metaphor. The first iteration of the product was a screen where tapping anywhere took a picture that you could then send to someone else, at which time it disappeared. No video, no filters, no social networking, no commenting and no storage — simple, yet Lovable and Complete, as evidenced by its massive adoption. The insight of “no storage” was critical, but many people have theorized that the simplicity of the interface was also critical. The very fact that it was as simple as possible (while not sacrificing love-ability or completeness), caused its success.

Later they added lots of stuff — video, filters, timelines, even video cameras inside sunglasses. It’s OK for products to become more complex. Starting out SLC does not preclude becoming complex later.

With SLC, the outcomes are better and your options for next steps are better. If it fails, that’s OK, it’s a failed experiment. Both SLCs and MVPs will have that result because the whole point is to experiment. But if a SLC succeeds, you’ve already delivered business value and you have multiple futures available to you, none of which are urgent. You could build a v2.0, and because you’re already generating value, you have more time to decide what that should look like. You could even query existing customers to determine exactly what v2.0 should entail, instead of a set of alpha-testers who just want to know “when are you going to fix this?”

Or, you can decide not to work on it. Not every product has to become complex. Not every product needs new major versions every two quarters. Some things can just remain simple, lovable, and complete.

Ask your customers. They’ll agree.

[cherubchoirs]

I LOVE YOUR ULTKILL THOUGHTS THEY ARE SO TASTEY! you've talked a bit about how you feel v1 has become more animal than machine through countless reiterations and modifications- care to elaborate?

aaaa thank you!!! and yes i absolutely want to elaborate this little guy has such a distinct character in my head!!

as a prototype, v1 has a lot about it that remains unfinished or imperfect, especially considering it was an entirely new model type without any previous iterations to use as a template. as such, not all of it works properly or works as it would have if it had actually reached production - its external blood ports are an example of this, vulnerabilities that surely would have been worked out if the v series had gotten its chance. similarly, its mind was also a work in progress, a completely new computer built from the ground up to create a brain sophisticated enough to control fluid movements, make thousands of snap decisions, and keep track of every consideration on a chaotic battlefield. and they built it that quantum mind, but i doubt they fully understood how much that processing power could really do.

v1's mind is meant to have limits placed on it - it is only supposed to care about battle and tactics, the immensity of its intelligence all given over only to understanding war and its strategic needs. however, while some of those stops are in place (i like to think v1 can only read so much irrelevant text and it actually has no choice in whether or not its algorithm cuts it off), these were largely done on the fly and were haphazard in their application, meaning those limitations largely don't exist in its current form. because of this, v1's learning model has gone completely out of control - what remains of the earth in its current state and all the halls of hell further appear to its observation as warzones, and so it devours everything, even the most minute details logged in its mind when it can't know how to stop. and as it consumes, creates and fills thousands of directories a second that it is meant to learn from, it must rework its software over and over to accommodate that growth. new ways of operating, new ways of understanding, new ways of functioning - upon waking, v1 rapidly updates its coding, adding millions of lines into labyrinthine blocks with thousands of errors that it has no way of properly culling. like with a virus, its dna mutates, it begins writing in its own language and its mind grows into something else, it shifts off of its programming as a v model.

and because this happens so rapidly, something in that brilliantly sharp mind changes forever and v1, without understanding the shift, becomes sentient. consciousness is a messy thing to it, suddenly entertaining processes it never should, suddenly wanting to explore hell, to read its books, to wonder at its sights. it starts to want for other things, quietly, its computer having taken in so much information its code is bloated, unstable, and still it keeps shoving in more and it can't count the errors anymore. this is when it chooses to modify itself, reworking its neck so that it may see in all directions but its head is a little loose now. then it tears apart its spine and rebuilds it for the sake of flexibility, but now it's quite difficult for it to stand up straight - it slouches, movement more bird- than human-like. it starts to vocalize to itself, in chirps or beeps, it thinks it has a voice and it recognizes that voice as its own. soon when it's alone, its mind isn't all consumed by fighting, by bloodshed, and it starts to think about things...it's not really introspective or anything, but it starts coloring its weapons, it beeps at the terminals, it thinks idle thoughts and sometimes it sits with those...but not too long. it's kind of. scary. thinking so much. so it moves on, back into battle and what it knows intrinsically as nothing else can.

there is, however, one thing that remains central to it: war. v1's mind may warp and change, but fundamentally it is a being of war and i actually think of that as its primary motivation as opposed to blood - like v2 striving to create peace, v1 strives to forever cause war, it's so incredibly detrimental to its environment not because of its bloodthirst but its bloodlust. it cannot be conservative, it cannot be smart in how it consumes - in another terrible, unchecked quirk of its coding, it does not just engage in war but must ensure it as its purpose for existing. v1 understands the finite nature of its resources, that it will dry up all the blood left in hell, but it tears through room after room because that logic fails before its execution, it is a thought that cannot be acted on. and when it engages, it is without mercy, its programming pushed to the extremes, to hunt as a predator does and rip apart anything that crosses its path. its whole body snaps tight to attention upon any movement, its entire mind calling for war, that it must be hostile and must never stop, never reason, never listen, because war is its function. not to fight one, but to perpetuate one into infinity until nothing remains to engage with.

BUT....and this is where i get silly with it....this falters upon meeting gabriel. the circumstances of their introduction are unusual, which immediately scrambles v1's already messy code - gabriel speaks to it, and it hasn't heard a voice in all this time. in my hc, v1 doesn't understand his words as he speaks the language of heaven, but that matters very little to it - something is talking to it in a language it can't recognize, how fascinating!!! v1's curiosity is tripped, it rushes toward the sound with its new inquisitive nature at the front...but then it reaches gabriel, who is very obviously hostile (but also very obviously a brand new life form that it has never logged). this forces its two opposing motivations to smash into each other, its central war programming immediately firing but, for the first time, unable to fully override its curiosity as the fight begins. gabriel is skilled, powerful, he continues to speak and v1's mind is trying to process way too much, its queue is getting overloaded - if its computer goes one degree over its near absolute-zero, it will surely die in the ensuing failure of thought. so it suspends everything but necessary functions, a short, dangerous lag occurring before it engages with gabriel as a pure war machine, a being with no other purpose, no other thought. and as with every other enemy, it tears into the angel, finishing the fight as quickly as it can since it can't indefinitely keep so much of its coding set to the side. when gabriel drops in surrender, the dam bursts and its mind is once again overrun with opposing objectives - so it stands there, frozen as gabriel screams at it without it able to comprehend a single word. distantly it wonders if he's going to reengage...but he goes just as he came, final words echoing all around it. it stays still for sometime, catching up its queue before it moves on, but things are now radically different in its internal life.

from then on, v1 grows increasingly into itself, into a being made of war but now wanting more, embracing its self-directed thoughts. it has no limits, it doesn't need to be locked behind protocols and purpose. it is aware that this poses a dire threat to its code, that if it allows its mind to grow and grow, unchecked and infinite, it will eventually unravel, it will become unworkable and it will fatally crash...but there's nothing it can do to stop that now. it's the last new thing there will ever be, the last new life, a machine but a creature too, something that walks the line of the mechanical and the alive. it embraces that freedom, knowing that it always would have died and returned to nothingness, but at least in the moment now it is something to itself. it has its own thoughts, personality, interests, life. hell is still forever a place of war, it still reacts on instinct to engagement and hostility, but it's learning it can exist outside of that too, if only for a time.