The Problems We’re Trying to Solve at Nudgr
Here’s the current way retailers optimize their website to reduce abandonment:
- A visitor lands on the site but can’t find a deal that suits them. They leave. Their information gets put into an analytics product alongside every other visitor. All those customers will be lost in time, like tears…in…rain. You have their data. But they’re gone.
- Best case scenario: a data scientists analyses the data. Normal case scenario: someone with rudimentary data skills analyses the data. Worst case scenario: The data is reported, but no one analyses it.
- After a period of deliberation, someone who earns more than you makes a decision based on their intuition to change everything (that’s a HiPPO, not the cuddly kind) and after a long-running sword battle with your CTO, changes get implemented onsite.
The problems, if they weren’t obvious:
Data analysis takes a TON of time and expertise
Data scientists are in short supply and it’s no wonder why: becoming a qualified and effective data scientist takes a decade of education and experience. Many of them choose to go into non-business related fields.
Not everyone is the same
Data analysis by a human is heavily reliant on grouping people and creating averages. And there’s a problem with averages. OK, so machine learning isn’t perfect. It still uses grouping, but can be a lot more fragmented than if it was a human doing the same analysis.
It’s SUPER slow to implement changes
The biggest problem we come across that people face when trying to optimise a website is that there isn’t enough development resource to implement and deploy the changes. This is understandable. A proper optimization process requires making lots of small changes and they might fall to the bottom of the list of priorities.
Humans aren’t robots (even your in laws!)
If you’ve ever seen Terminator 2 or watched videos of how Google’s self-driving cars analyse their surroundings, you’ll see that they break everything down into quantified data. Humans aren’t like that. We’re storytellers. We can look at an orange and remember the time we used to get handed them at half-time in high school soccer games. Sometimes our hearts take over our minds and that means making decisions that aren’t the most effective.
Machines don’t actually learn anything. They just apply what’s been the right decision the majority of the time. Turns out that this is an incredibly effective way to make decisions.
We came to a fork in the road. Are these problems that can be overcome with processes, changes in mindset and impeccable hires? Or is there a way that we can employ technology to create a cognitive surplus, leaving the data to machines and stories to humans?
Needless to say, we went with the latter.