We regularly hear people say things like “I think augmented reality has a role to play in our business” or “We think we need to adopt machine learning”… and the problem with these statements are not in the content. The content is great.
The problem is that thinking technology-first encourages product & service development to shoe-horn in the technology, rather than incorporating the relevant technology where the business-need requires it.
Let’s explore some examples and the potential pitfalls of thinking technology-first.
You end up with too many digital products
We find it astounding that when you search for a particular business in the app store, just how many businesses have multiple apps.
Now, this can happen due to organisational silos, but in many instances, this can be exacerbated by technology-first thinking. You want an application to leverage smart sensors inside users phones, build an app… you want to have an AR solution, build an app… you want to leverage distributed ledger technology, build an app.
By leading with the customer’s behaviours and motivations, then mapping them to your business-needs. These multiple, discrete solutions may have just become a feature set inside one point of contact.
You build something that is shiny, but not that useful
An example that we saw a few years ago was a mobile banking application, which when you scanned your debit card it visualised spending data by category, alongside a range of other balance and overdraft data on top of the card.
It is a nice idea, and it looked lovely.
The problem is… You can’t do that anywhere but in private, otherwise everyone else is going to look to see what you are doing and observe sensitive financial data.
And if you are at home, how likely are you to go and find your wallet, choose one of many cards, place it on a table and scan it? When you could just open your banking app.
It is a classic case of ‘find a technology and shoe-horn it into your business’, rather than finding a problem or need that the technology solves.
You build something that isn’t that adoptable
Last year, we saw a prototype of a machine learning chat bot, which could complete image recognition tasks. It was pretty nice and worked seamlessly.
The use case was however a sales chat bot for an American insurance company and the person showcasing the application said “Just imagine, the insurance provider asks you for your car details… make, model, year, number plate. And all you have to do is upload an image.”
It worked, he uploaded an image and the application came back with the details for you to verify…
But, how adoptable is that technology?
Firstly, how many customers want to apply for their insurance via a chat bot? Secondly, how many people have quick access to a photo of their car from an angle where the full number plate is visible? And thirdly, isn’t it just easier to type in the details you already know about your car?
Without testing and learning, we wouldn’t know what does and doesn’t work, so we commend all attempts to find new applications for emerging technologies. However, we believe that the value of the right technologies can be teased out in isolating the business need or customer desire.
If you want to find out more about our process, or have a question about extended reality or machine learning in your business, get in touch.
Related insights
