Business Models for New Products with Potential Automation


Oftentimes when a new product that wants to take advantage of the AI/ML/Automation boom comes along, UX can be the last thing to be considered. This is short sighted. Automation based new products often don’t take into account a 

  • Collection mechanism for data or
  • The time and value that must be given to customers during the time in which there is no automation

In encouraging customers and early adopters to products, the UX of how the data is collected is key. Automation, utilising Machine Learning/Artificial Intelligence, is often seen to be at the heart of the value proposition of new products. What is not always considered is the stepping stones that are required to get there. Often the product will be requiring data, lots of it, and if possible, well formatted and labelled correctly.

Automating in a space where there isn’t a lot of machine readable data is always tricky for industries that seem impenetrable to disruption. I honestly don’t think disruption should be a creator’s ultimate goal. If anything is to be learned from Superpumped the book on Uber by Mike Isaac, aiming to be a crusader and “disruptor” can backfire easily. But valuing and respecting your users is definitely the right way to go in beginning as they will largely be those who will tolerate the use of the “dumb” system. 

Scalable Business Models

If we are making products that have a profit, or will potentially be non-for-profits, the value proposition doesn’t really change a lot. Whatever we create before automation must be small enough and easy enough to do by a human prior to automation. We are not looking to exploit people’s labour and replicate their work, we are seeking to simplify their processes. We must help them both when our product is “dumb” and when it is “smart”.

The Potential Life Cycle

A diagram in a clockwise direction.  1. A picture of very little data -  Automation can make things easier for customers but requires huge amounts of data you may not have yet Then: new products get few customers 2. But if it is hard to scale the product we won't get the data - Customer data is valuable. 3. Our product has to reward them for taking a chance on us 4. Making the initial user case specific, clear and with a direct value proposition is helpful to them The customer then finds it easier to get on board. 5. We cut down on time and get close to getting a better scalable product before branching out to another customer Customer sees the reward for our services plus it is easier for others to get on board 6. A picture with lots of data - Getting good data means having a useful, scalable and intuitive business model

The key to creating services where the customer’s data is learned from is to make the experience as smooth as possible. You must show your customers (or users if they are the same) that you respect their time and data. This is where UX must do most of the heavy lifting in getting traction for a product. So if your new ML/AI/Automation product hasn’t considered UX, then they haven’t considered how their customers are contributing their knowledge, time and energy to improving your product. The experience is just as important as the programming behind the product.