Just getting to this article now. Interesting. Lots to chew on. I’ll show my age here a bit with some pushbacks on a very good piece. We have been talking about mass customization since 1987 when Stanley Davis published a seminal book about business and society called Future Perfect. It is still prescient. I’m not sure I would be so quick at relegating the same ad shouting into the void to be a relic so quickly. A set of ads from Pepsodent, for example, literally created toothpaste. Our media fragmentation gives the illusion that one SET of ads don’t matter and that this stuff is passe but mammoth numbers of people now look at the MrBeast enterprise which is surely an example of the sameness being shouted into the void. Recommendation engines I think fall into the same category (whose name I have forgotten; research by Kahneman I believe) as horoscopes; things people generally agree upon but aren’t really that specific. I find that for things that I do a lot and have a wide set of interests and values for - say music in my case - that the recommendations are quite trite or just more and more of the same. They don’t really have any idea what I like. What Ted Gioia calls out Spotify for by creating, through hidden artists and / or AI, bland music that it then recommends to you.
But your essential theme is valid and true. People want to be know as themselves to vendors, with their preferences well understood. Across the entire technical spectrum of devices and access points. People want digital means to be like your favourite Friday pizza place where they remember you love capicolla and buffata, but know enough about you to sometimes recommend that day’s special. I think that IRL experience is hard to replicate using technology
Thanks for these tips, Mack. I notice it's possible to segment Substack subscribers - is this just for analysis or for sending tailored posts to just one group?
Either. I will sometimes send out a new email just to my 5-star subs (Substack ranks them based on engagement with my articles). I will occasionally send out a personal email to a sub thanking them for engagement, I need to do more of that as well.
I do love the recommendations on things like movies and other various topics, but what I struggle with is that I have a very wide range of interests.
My music is eclectic, and my movie taste runs the gamut—from sappy love stories on Hallmark to series like Bosch, Reacher, and Jack Ryan, with a lot of things in between.
So, unfortunately, I probably confuse the hell out of algorithms.
Hey Bette, the recommendations on like Netflix or Spotify is really still very basic, although I do think Spotify does a good job of recommending ‘artists that sound like’ a particular artist. It will be interesting to see how AI evolves and can it spot trends in picks that even the user misses. Like Netflix recommending movies that don’t sound like a movie you would normally watch, but you watch it and love it cause Netflix picked up a commonality in your viewing history that triggered the recommendation.
Yes, I'm sure it will evolve even more. What I notice is that they tend to recommend movies based on the last one I watched. I can’t speak for Netflix, but with other services I’ve used, I find it irritating. I have so many different interests and don’t stick to one type of movie. It’s frustrating when it keeps recommending stuff based only on my last watch. Sometimes, it feels like you have to dig around to find things that align more with your broader interests. Maybe I’m the only one with that issue, though.
Just getting to this article now. Interesting. Lots to chew on. I’ll show my age here a bit with some pushbacks on a very good piece. We have been talking about mass customization since 1987 when Stanley Davis published a seminal book about business and society called Future Perfect. It is still prescient. I’m not sure I would be so quick at relegating the same ad shouting into the void to be a relic so quickly. A set of ads from Pepsodent, for example, literally created toothpaste. Our media fragmentation gives the illusion that one SET of ads don’t matter and that this stuff is passe but mammoth numbers of people now look at the MrBeast enterprise which is surely an example of the sameness being shouted into the void. Recommendation engines I think fall into the same category (whose name I have forgotten; research by Kahneman I believe) as horoscopes; things people generally agree upon but aren’t really that specific. I find that for things that I do a lot and have a wide set of interests and values for - say music in my case - that the recommendations are quite trite or just more and more of the same. They don’t really have any idea what I like. What Ted Gioia calls out Spotify for by creating, through hidden artists and / or AI, bland music that it then recommends to you.
But your essential theme is valid and true. People want to be know as themselves to vendors, with their preferences well understood. Across the entire technical spectrum of devices and access points. People want digital means to be like your favourite Friday pizza place where they remember you love capicolla and buffata, but know enough about you to sometimes recommend that day’s special. I think that IRL experience is hard to replicate using technology
Thank you, Mack! This was great, concrete, and applicable! Very useful.
Thank you, my wonderful friend! Have a great weekend!
Have a great weekend, dear @Mack Collier!
Thanks for these tips, Mack. I notice it's possible to segment Substack subscribers - is this just for analysis or for sending tailored posts to just one group?
Either. I will sometimes send out a new email just to my 5-star subs (Substack ranks them based on engagement with my articles). I will occasionally send out a personal email to a sub thanking them for engagement, I need to do more of that as well.
I do love the recommendations on things like movies and other various topics, but what I struggle with is that I have a very wide range of interests.
My music is eclectic, and my movie taste runs the gamut—from sappy love stories on Hallmark to series like Bosch, Reacher, and Jack Ryan, with a lot of things in between.
So, unfortunately, I probably confuse the hell out of algorithms.
Hey Bette, the recommendations on like Netflix or Spotify is really still very basic, although I do think Spotify does a good job of recommending ‘artists that sound like’ a particular artist. It will be interesting to see how AI evolves and can it spot trends in picks that even the user misses. Like Netflix recommending movies that don’t sound like a movie you would normally watch, but you watch it and love it cause Netflix picked up a commonality in your viewing history that triggered the recommendation.
Yes, I'm sure it will evolve even more. What I notice is that they tend to recommend movies based on the last one I watched. I can’t speak for Netflix, but with other services I’ve used, I find it irritating. I have so many different interests and don’t stick to one type of movie. It’s frustrating when it keeps recommending stuff based only on my last watch. Sometimes, it feels like you have to dig around to find things that align more with your broader interests. Maybe I’m the only one with that issue, though.
Netflix nails it because their AI makes you forget it's AI.
It just gets you.
I have a term for this "applied empathy."
And we use it to build products.
CX collects data points.
Then used these DPs to connect in ways that make customers feel like they are understood.
It also helps when the engineers are emotionally self aware themselves lol
As always thank you bro :)
“It also helps when the engineers are emotionally self aware themselves lol”
It’s still early, but I suspect this will be the truest statement I read all day! Lol. Thanks sis for reading and sharing 🤗
no problem at all. 🤗