Making Machine Learning / AI Practical
March 21st, 2017One thing we’ve (at Nudge) been toying with and discussing for the last two years is – what’s the best play for machine learning for marketers?
There’s a lot of hype.
And of course, being in the marketing world, only amplified this.
Our lens has been, how does it make you super human? What can it do, practically, today, that makes you better at your job?
Till we found that, we weren’t gonna touch it.
But now I’m excited to share, we’ve got the beginnings of it. The team built a CommonSenseBot, which scans brands content, to find what is or isn’t working. Then lets you know.
Easy.
We flipped the switch on it and just had it running in parallel. Within a few days it caught a major problem on a campaign, which had it not been identified could have cost the client a lot of money. More money than they pay us! (Yes we need to charge more ha).
Each week we are adding new components as we operationalize our knowledge and experience into it.
As we expand, we’re able to tap machine learning to identify potential problems and wins FASTER through predictions, thus minimizing that time to respond.
To me this passes our test, it’s smart, it’s consumable, it makes your better at your job – and it doesn’t add anything more to your plate. i.e. it does make you superhuman!
The future of this space is not AI as a sector, it’s of smart products which use it to simply enhance and do a better job.
Part of our product inspiration has been Marsbot, a subproduct of FourSquare, which makes recommendations based on where you’ve checked in before, passively in the background. If you haven’t tried it, do check it out.