If I had to boil my view of the world down to one thing – it’s this quote from Warren Buffett.
“If something’s not worth doing at all, it’s not worth doing well.”
Too many products are optimised to do something that doesn’t need doing at all.
A classic example in our business is extracting data from pdf invoices.
Here’s what happens. A computer somewhere creates a dataset. That dataset is used to produce an invoice. The invoice gets sent to your business. You scan it and put it into a financial system. Along the way, the scan loses quality, the image includes that stamp someone put on it to show when it was received. And you’ve got a few thousand of these to process.
So, you go out and buy a best in class image processor that can use AI to extract numbers from these messy scans.
But why?
Why not just dissolve the problem. Go to your supplier and get the data directly from them in a machine readable format – a csv?
This cuts out all the messy middle data wrangling. The closer you get to the source the easier it is to get what you want.
Now this can be more or less difficult depending on where you are.
But this is where thinking in systems comes in.
The more complex you make something the more expensive it gets and the more likely it is that something will go wrong.
Complexity creates value for a while but after a certain point you start diverting resources from doing your task to managing complexity.
At which point value drops.
The trick, then, when building something is knowing when to stop.
Handcrafted insight 