Over and over I have seen two key reasons why analytics projects fail
#1Business Context – AI strategy needs to align with Business Strategy.
Not as a hammer in search of a nail.
A long time ago, we attempted building out NLP models with text data – we tried it all, from Word2Vec to Google BERT.
Business SME inputs oft helped us look at the right data and ask the right questions. Don’t build in silos.
If you want to increase the success of your applied data science problems, work in tandem with a business SME
Insights don’t drive impact – actions do.
#2Focus as much on the consumption of AI as the production of AI
We oft think about all the fancy ways in which we can produce predictions – rarely about how the insights will be consumed or if it aligns with what success means to the business
When you scope out a project, remember to ask, ‘so what’.
So what if you provide this – what happens then? Is it integrated in the last mile, do you have alignment from users on how it will be adopted?
Actions drive business impact – not just insights.
And if you are a data scientist who can think outcomes instead of output, you are immensely valuable.
Do you have experiences on winning with AI?
**************************
#reviewswithranjani #analytics
#Technology | #Books | #BeingBetter