A Salesforce founder recently said something that should terrify every software company:
You will never have to touch the Salesforce interface ever again.
Think about that for a moment. The founder of a multi-billion dollar company just told his customers they do not need to use his product’s interface.
This is not a bug in the system. It is the entire future of software.
And it has a plain consequence for anybody paying a monthly fee for a tool: the part you actually look at is becoming the least important part of what you bought.
This piece follows the argument made by Riccardo Vandra in Why Software Interfaces Are Becoming Obsolete, rewritten for owners who do not work in software.
The interface was always the bottleneck
Say you want to add a job to whatever system you use to track work.
You open the app. You click New. You type a name for the job. You click into a description box and type what needs doing. You pick a date. You assign it to somebody. You hit Save.
That is seven things. Seven clicks and keystrokes to write down one job you already had entirely in your head before you opened the laptop.
Now ask why it takes seven.
Not because the work is complicated. Underneath, that job is one instruction: make a record, give it these details, attach it to this person. The seven steps exist because of you. You are a person. People cannot talk to a database directly, so somebody built a room full of buttons for you to stand in and press things.
That was the right call for forty years. But it is worth naming what it is. The interface is not a feature of the software. It is a workaround for the fact that a human being is slow.
An assistant doing that same job does not open the app. It does not hunt for the New button, because it has no eyes and it is not looking at a screen. It goes underneath, to where the information actually lives, and does the whole thing in one instruction.
Seven clicks become one sentence.
That is not a small speed improvement. It is a different shape of work. And once you see it, the monthly fee starts to look strange, because a good part of what you have been paying for is a room that nothing needs to walk into anymore.
Why you still need something to look at
Here is the part most people get wrong when they first hear this. They conclude screens are finished. They are not, and the reason matters.
There is a trap on the other side. Hand the whole job over and you gain the speed, but you lose sight of what happened. The work gets done somewhere you cannot see, and you find out how it went when a customer tells you.
These systems are also not perfect. Vandra’s figure from building them is that they start out around 90% reliable and reach roughly 99% with time and correction. 99% is very good. It is not 100%, and it never becomes 100%.
That last 1% sounds like a rounding error until you put volume behind it. Something that runs 200 times a month and gets it wrong 1 time in 100 is 2 wrong invoices, or 2 wrong appointments, or 2 wrong messages sent to real customers, every month, forever. At 10 jobs a month you would catch those by accident. At 200 you will not.
So the screen survives. It just stops being the place you do the work and becomes the place you check it.
Think about building a house. You do not stand over the crew watching every nail go in. But you would absolutely walk the place afterward with the plans in your hand, making sure what got built matches what you asked for. That is what a blueprint is for once the house is up. Not instructions. Verification.
The interface is not going away. It is turning from a tool you use to create things into a dashboard you use to confirm them.
The two layers
That split is the thing worth actually sitting with, because it explains everything else.
Software is separating into two layers that used to be welded together.
The infrastructure layer. Where the work happens. Records get created, appointments get booked, invoices go out, numbers move from one place to another. There are no screens down here and there never were. This half has always spoken in plain data, and it always worked fine with nobody watching.
The verification layer. Where you look in and check. Dashboards, lists, a calendar, a record of what happened and when. This half exists for one reason: so a person can oversee something they are no longer doing by hand.
For your entire working life those two arrived bolted together in one app. You used the same screen to do the work and to see the work. Nobody chose that on purpose. It was the only option, because the only way to make something happen was for a person to click it.
Now they come apart, and that is a rebuild rather than a feature. Software companies have to start designing as though the main user is a machine and the human being is the supervisor. Very little existing software is built that way.
Here is what it looks like from your side. You say: book me an appointment at three.
Underneath, a lot happens. It checks whether three is actually free. It creates the appointment. It attaches the right customer. It sends the confirmation. None of that involves a screen and none of it involves you. You never open the calendar app.
Then, later, you open your calendar and see three o’clock filled in.
You did not book it. You confirmed it. The doing and the seeing came apart, and the screen you looked at was only ever the second half.
What actually matters now
If software is reached directly rather than clicked through, then how good the screen looks stops being the thing you are choosing between. Which raises an uncomfortable question for every business, including yours: when the interface stops being the product, what is left that a competitor cannot simply copy?
Three things.
Your data. Years of customer history, job records, what you charged and why, which jobs went badly and what you learned from them. Nobody else has any of that. It is the one asset that gets more valuable while everything else gets cheaper, and most small businesses keep it in somebody’s head or a phone full of text messages.
Your audience. The customers who already know you and the ones who send you referrals. A competitor can copy your service in a week. They cannot copy fifteen years of somebody trusting you.
Your brand. What people say about you around town. That was always worth something. It is worth more now, because when everybody can do roughly the same thing, being the one people have actually heard of is most of the game.
When building things gets dramatically cheaper, features stop being protection. Whatever clever thing a competitor has, somebody can now rebuild it in a fraction of the time it first took. Those three cannot be rebuilt, and none of them is software. All three are things software either strengthens or wastes, depending on whether you ever built anywhere for them to live.
What people actually ask for
This is not a prediction. It is what already happens when businesses are given the choice.
Vandra’s consultancy has now built this for over 40 businesses, and reports the same pattern every time: people want to execute actions naturally and verify the results visually. Say what you want in plain words. Look at a screen to confirm it happened.
What they do not ask about is which tools are connected underneath. That turns out to be a question owners never cared about, only tolerated.
Notice where that leaves traditional software. It sits awkwardly in the middle of those two things. Too slow and too fiddly to be the way you say what you want. Too cluttered and too busy to be a clean place to check a result. It is competing against a sentence on one side and a dashboard on the other, and it is not obviously better than either.
Nobody wants to learn twelve tools. Nobody ever did. Owners put up with it because that was the price of the work getting done. Given one way in that speaks plain English, the pile of separate logins on your desk starts to look like a temporary condition rather than a permanent one.
What this means for your business
Two sides to this, and both are worth ten minutes.
As somebody who buys software. The shopping list has changed, and roughly inverted:
- Can other things reach it? Not can you log in. Can other software get in and out without a person copying and pasting. A tool nothing else can reach is a dead end, however good the screens are.
- Can you get your information out? Your customer list, your job history, your invoices. If the only way to see it is to log in and look at it, you do not own it. You are renting a window onto your own business.
- Does it show you what it did? Not what you could build with it. What it actually did, and when. That record is the part that gains value, because it is the only thing standing between a wrong result and finding out from a customer.
- What does it connect to? Feature lists sold software for twenty years and they matter much less now, because a feature nothing can reach is a feature nobody uses.
That last one is roughly the opposite of how software still gets sold to you. Longer list, higher price. Ask what it connects to instead.
As somebody who sells something. The same question points back at you. If your prices, your services and your availability only exist inside a form a human has to fill in, you are built for a visitor who is slowly stopping being the only kind you get.
Which leaves one question worth answering honestly, and it is the one the whole argument comes down to:
Are you designing for humans who click, or for agents that execute?
Where to start
None of this means cancel your subscriptions on Monday. Most of your tools are earning their keep today, doing exactly what they do. Your accounting software is not evaporating this year. This is a direction, not an emergency.
So do two small things instead.
Next renewal that comes up, before you click yes, ask the two questions that now decide a tool’s shelf life. Can my information get out, and can other software get in. If the answer to both is no, you do not have a crisis. You have a tool with an expiry date, and you now know which one it is. That is a far better position than finding out the day you try to leave and discover five years of customer history only exists on somebody else’s screen.
Then, separately and quietly, start writing down what you know. Where your customers came from, what you charged, what worked and what did not. That is the part nobody can rebuild for you, and it is what makes every tool you buy after this worth more.
The interface is not dead. It is evolving into something completely different, and the businesses that understand that distinction will be the ones still standing when the dust settles.