Prairie Fire: AI in B2C Marketing
Prairie Fire: AI in B2C Marketing
A Lightning Strike In A Dry Grassland
Fire is as much a part of the prairie landscape as grass….Fire has many roles in preserving the prairie.
Fires remove old plant material, consuming dry, dead grass and woody shrubs and trees and returning nutrients to the soil.
The burn exposes soil to the sunlight, and new grasses grow quickly in the warmed soil.
—NASA Earth Observatory
Over many years, the desiccated remains of grasses from seasons past can compact and grow into an impenetrable mat of material.
The soft, new heads of grasses trying desperately to find sun and unfurl their leaves are no match for the thatched roof trapping them in the dark.
Already existing, larger plants pay no mind to the ground covering.
Existing plants benefit from reduced competition.
The weak plants hang on when they shouldn’t
This cycle repeats.
Year after year.
Fewer and fewer new growths.
But the pressure cooker of life doesn’t leave the lid on for long.
FLASH: A lightning bolt strikes a thick part of the thatched mat and a fire springs up.
The fire spreads across the prairie, gaining speed with the growth of each tongue of flame.
As the wall of fire quickly slides across the surface of the prairie, all of the weak, barely alive plants are burned to a crisp.
The deep-rooted trees survive, and the new grasses—cleared of the thatch—thrive, quickly filling in the new space.
This wasn’t really a story about grasslands.
AI is the lighting strike that’s lit the fire, but the destroying fire hasn’t spread yet.
Businesses and tools powered by advanced analytics are going to be the flame that rips through business and public consciousness.
How Will AI Change Things
In my previous post, “Where Does AI Take MarTech”, I broke MarTech into large areas for likelihood of disruption.
In this article, I’m breaking down the “Likely for Disruption” category into more detailed descriptions of why each section is included in this section.
If you haven’t seen the previous article, check it out here:
The categories that I called out before were as follows: AdTech, Social Media Marketing, Content Marketing, and Brand Marketing.
These 4 areas will likely require understanding and experience with AI tools within the next few years.
The Likely For Disruption category extends beyond just these categories.
The key identifiers for things ripe for modification are:
Well-Defined Variables
Large Datasets
Fast Cycle Times
Low Regulatory Burden
Well-Defined Variables
If you can specifically say if something was successful with numerical metrics, then it can be automatically measured
Defining a robust error function that is useful across whatever space your dataset contains lets an ML or advanced analytics system approach a better system
Large Datasets
To train and run advanced analytics, you need gigantic sets of data. Beyond the data just needed to train a model, you need to add an additional 30% to ensure that your parameters aren’t overfitting past data
Never forget, Garbage In: Garbage Out. More data does not equal better!
Fast Cycle Times
Creating great analytics and AI for a system that evolves very slowly isn’t worth the squeeze.
There is a hefty overhead required to any project that involves teams of technologists creating and implementing a new project. Ensure that you have a very large potential gain from the investment before taking on a big project.
XKCD #1205
Low Regulatory Burden
The pharmaceutical industry spends crazy amounts of money on marketing.
The pharmaceutical industry ALSO spends a ton of money on paying fines
Building an AI tool that has a lot of leeway is a dangerous proposition that could quickly have negative ROI from penalties
In conclusion, AI has an extreme edge in things that have good data, chances for mistakes, and requires fast responses.
Firms that are well-adapted to understanding their customers, have strong fundamentals, and high customer loyalty can withstand the fire. They still need to be wary of potential changes, but they are the trees sitting in the field of weak shrubs.