Everyone that I talk to in retail and retail real estate is obsessed with not missing the boat on AI, and being ready for the tsunami that is coming.
Retail is the proverbial Michael Myers (of “Halloween” fame).
Perennially left for dead — this time is REALLY the end — somehow he/it keeps getting back up. Bricks and mortar is pretty resilient.
I thought about this reading the article, “AI Agents and The End of Shopping as We Know It,” by Zack Kiss
The author’s core assertion is that optionality has crippled modern shopping.
Specifically, he asserts that consumer choice, once hailed as a cornerstone of freedom and progress, has passed the point of diminishing returns.
Putting aside the nebulous touch and feel “stats” that the author puts forth to support his argument, I’ve not seen any actual, sourced data that suggests sales, profits, markets or even addressable markets are compromised by excess choice.
That’s where the rubber meets the road, after all.
How people Feel and what they actually Do, often diverge greatly.
In the macro view of our economy, the max optionality companies — Amazon, Facebook, Google, TikTok — are crushing it.
Consumers want better choices, fair prices and sufficient information to make educated decisions.
But, the author’s focus on the consumer in terms of WHO the optionality is optimized for gets it backwards. Optionality is optimized for the Operator, NOT the Consumer.
Google Search has gotten worse over the years not because it’s too hard to deliver integral search, but because Google has increasingly optimized on engagement, clicks, and price per unit over quality per unit that yields lower engagement.
Facebook and TikTok work the same way, which is why consumers see so much divisive content. It engages the best.
Amazon is emblematic. It’s gotten 10X noisier in terms of decisioning, again, not because it’s better for the consumer, but because there are so many businesses and business models that Amazon has to care and feed.
Amazon is optimizing on an ever-shifting field of a “plausible needs” based on things like:
- Whether a product is Amazon white label or third party
- Advertising pay for placement (advertising)
- Participation in Amazon prime
- Aggregate services that manufacturer pays Amazon for
- Spiffs, discounts, or margin sweeteners to Amazon
This gets to Kass’ primary argument as it pertains to retail; namely, that the next leap isn’t another recommendation algorithm—it’s delegation.
Right…the goods and services that are so undifferentiated that the customer would delegate the shopping, because the consumer sees it as a tax on their time and attention.
While this bucket may exist, it is not, in itself, viable retail.
As we see time and again in this business, undifferentiated retail is dead.
If it can be commoditized, de-localized or digitized, it will be.
A final point. There is a tendency to conflate:
- Algorithms, which have been around forever (Think: The thousands of credit card offers we’ve all gotten in the mail for ~40 years); with
- Artificial intelligence (AI), which builds on algorithms but extends it with a large language model (Think: library of all public, known information) fed into unimaginably large & powerful computing fabric; with
- Agentic AI, which is very nascent, but more akin to autonomous decision making and parallelized activity (Think: scaling a telemarketing team or call center, one agent instance at a time)
Case in point, Kass asserts that via Amazon’s Subscribe & Save, millions of households hand off their toothpaste, detergent, or pet food decisions to an ‘algorithm,’ the algorithm in this case being “send me a new order every X weeks.”
Pretty thin, as proof points go.
Mark Sigal, CEO – Datex Property Solutions