cutting edge.
I share principles in this newsletter for a reason.
The tech changes.
A shiny thing will fade.
Fast tactics are trumped by sound strategy.
I work with tech startups and have done this for a long time.
So it’s only natural people want to know what is happening on the edge.
Today I’m going to indulge that curiosity - and let you know that the principles are more important.
Things like the below only work if you understand the principles.
So here’s what I’m seeing:
AI-trained video content.
Do you remember SEO? Where if you got in early you reaped the rewards for decades to come? I feel like that’s where YouTube is right now.
Here’s a video of a guy called Nick Sararev - he runs multiple things, has a lovely Canadian accent, and shared a blueprint for scraping and transcribing the best Instagram Reels out there.
He does this because it then goes into an LLM, and produces a script (in his voice/tone) that he can adapt and shoot a better video with.
Why is this useful?
Because if you’re grabbing high-performing videos (that already have high engagement) and then remixing the formula and producing your own video - it’s going to perform.
Repeatably with an automation running in the background to keep improving as you gather more video “formulas” and index it against your own.
Using it to power ads
Cody Schneider is a growth bro who pumps out content and clearly loves “the game” of finding ways to exploit systems.
On Greg Eisenberg’s podcast he breaks down how to grow in B2B.
Find and/or create popular meme content
Recreate these memes with scripts that weave in
Run these as TikTok/LinkedIn/Reels ads with AI-generated avatars on a green screen, with demo footage of your product in the background
Set budget to 80% for the best performing ad, and continue testing new meme ads on the other 20%
Remarketing pixel anyone visiting your website
Blow it up further with data
It’s now possible to grab the Total Addressable Market (TAM) for your B2B product with <$2,000.
Enrich that data to grab phone numbers and personal email addresses and feed that into the ad platforms as a matched/custom audience - you’ll get about 20-30% fo the audience matching.
So with a TAM of 50,000 people you have 10,000 you can target with the ads.
Now run the ads from above against them - and…
Outbound email the hell of them
The tools for outbound email are also more affordable than ever.
Cold email reply rates (on average) are down.
But if you’ve warmed up the audience using the above, you can email anyone that’s visited your website, clicked on an ad, watched a video, or shown a “signal” that they’re interested.
Training the signal
A growth engineer can take “signals” and blow it up even more running increasingly more sophisticated machine learning with a cloud data management platform like Snowflake.
All of the tactics above are fed into a model that becomes increasingly better at serving the right message at the right time.
I know this because I met one of them - William worked at LinkedIn on their growth, and later was part of the team that blew up Rippling.
He’s now out there consulting.
–
Where does it go?
I don’t think anyone knows.
But if you’re looking to grow the principles remain the same.
Focus on the small stuff first - and once you’ve cracked it manually I can point you in the direction of ways to make it happen faster.
-Marshy
p.s. I was at Victorian Parliament House this week through some work I do with SANE - here's me looking fancy in front of a very talented artist named Patricia
