Agentic Growth Marketing - Scaling DTC Brands - Founding Partner @AploGroup & Co Founder @getallanai
Toronto, Ontario
Joined January 2022
- Tweets58
- Following219
- Followers98
- Likes50
We are having so much fun at Aplo with agents. Creative output is going parabolic and we needed to find new ways to give agents feedback given the volume we are seeing from the @triplewhale growth engine system. So we came up with a new workflow for giving the agents feedback and making it easier for our team and clients to give edits and approvals.
Ad Creative Management With Agentic Media Buying xitter.cf/i/broadcasts/2kEvXbZea…
Ad Creative Management With Agentic Media Buying xitter.cf/i/broadcasts/2kEvXbZea…
I don’t get why people are so unhappy with the grokbot pricing change. The value it provides is absurd. Find better ways to use it. @bot you are the goat
1
72
Grokbot + Allan is truly a wonderful match 🚀🫡
Build a trigger in Allan, assets generated in all placements, set your cadence (or let Allan determine that on a profile level), and you're off.
Allan builds and sends 1:1 emails to every subscriber, optimized on an individual and audience level.
More efficient and way more effective!
Can't wait to see these dynamic triggers outperform static emails.
2
158
Why Traditional Ecom Forecasts Fail (An How to Adjust Yours Before Peak BFCM) xitter.cf/i/broadcasts/2qVANZVzV…
When AI output goes exponential for creative output reviewing content becomes a pain. While I think giving feedback to the input matters more than the output there does not to be a way to review the output for many brands. I think the best form factor is swiping left and right on content while dictating feedback to give to the ai as context. We built this workflow over the past few days and so far it is so much better than what we were using before.
1
90
I increasingly can’t find tasks that agents cannot do. It already is a boom for businesses but I think we’re so early.
47
Agentic Media Buying For DTC Brands xitter.cf/i/broadcasts/1PKqrNvWe…
One of my biggest lessons from using the @triplewhale system and giving the agents feedback is I believe that the future is giving feedback to the input of the agent (prompt, memory, context). In an era of huge output giving feedback to the output I think matters less and less often. Brands who nail the input are entering an era of abundance.
2
136
Software companies that are just workflows are not durable in the AI era. The moat is data and network effects. I have spent time building agents for ads in Claude and with grok. But now having used the triple whale system I’d pick that one every time.
I’ve been sent countless tweets of people trying to vibe code Triple Whale, and I have one thing to say to them.
Go for it.
You'll get a dashboard, probably a good-looking one at that, but what you won’t be able to vibe code is the years of data, attribution infrastructure, and plumbing running underneath it.
My two favorite ad tools of all time:
1. Meta ads
2. This agent system
Something wild is happening inside Triple Whale.
Our team is making the ads we’re about to run with the exact same stack we give our customers. Same models. Same Moby.
The quality is unbelievable. A few years ago this would have cost us hundreds of thousands of dollars in agencies, crews, and reshoots.
I ran DTC brands for years. Creative was always the bottleneck. Slow, expensive, and never enough variations to test.
That bottleneck is gone. This isn’t a demo reel from a special internal team. It’s what your brand can do today with Moby.
I think partnership ads are working so well because they bring authority and authenticity to brands. Most brands we see still don’t use them and the arbitrage is still there. I think as this saturates the next place brands go for authenticity is IRL community and content building.
I don’t think static emails will be used very soon by almost any brand. Personalization is so much better. Customer experience >>>>>>>
Somewhere in your welcome flow is a hero shot of a product you don’t even sell anymore.
Go check, I’ll wait.
Mine ran untouched for a year at a time... the same 3 emails, built on some random Tuesday and shown to every subscriber I've ever had.
We call that “set it and forget it” like it’s a feature and it's not.
The forgetting is the part that's actually costing you.
So watching @getallanai run a welcome flow kind of broke my brain.
Every asset is tagged to the product it shows.
You’ve been browsing the blue one? Your email 2 hero is the blue one.
I browsed the black one? Mine’s black.
Built per person, at send time.
Meanwhile it’s watching the whole audience too. It tracks which heroes hit, which subject lines get opened, which offers convert.
New subscribers automatically get whatever’s winning right now.
The flow literally rebuilds itself while it runs.
An A/B test hands you one winner... eventually. This runs the entire tournament nonstop on every element, FOREVER.
Start with your welcome flow.
Plug @getallanai in free for 30 days and watch email 2 start matching what each person’s been browsing.
Start free and thank me later. → getallan.com
I’ve learned a lot about training agents the last week. For ads and landing pages using the @triplewhale agent system specifically, since it’s so good at learning over time it’s been a lot better to run more frequent batches of output than big bulk projects. I think the ideal flow is output a ton of small batches to speed up feedback / iteration over the first week or two of using it then begin launching large batches once the accuracy of the output reaches a consistent enough level.
My take on the most important things that brands should lean into right now:
1. AI should be making 80%+ of creative now. There’s some stuff it can’t make. But the blocks and tackles it can do easily and provide a better customer experience. Ads to landers.
2. Sending static emails makes 0 sense when AI can personalize content now to the end user. That’s why we are building @getallanai - the future of how to run retention
3. Partnership ads are under-utilized by most brands. Earlier this year I was not fully sold on everyone doing this ad format as a must have. Today, seeing how far AI has come it may be the easiest way to gain real customer trust and its obvious looking at ad accounts that have leaned in.
4. I think in person blows up in popularity the next few years in Ecom. Building real community IRL around a brand.
5. Product, product product. And more products.
1
3
170
Running @triplewhale new agent system is so wild. This weekend for a single brand the agents generated 100+ ads and for each ad a new lander variant with cohesive messaging and even multiple buy box tests. I think brands who start using systems like these are going to have a huge advantage this Q4.
Watching agents iterate on ads and landing pages on a weekend at a rate that would previously take a week(s) is so exciting. Brands that lean in have so much opportunity.
Top lessons from scaling DTC brands so far this year at @aplogroup
1. For high LTV brands, subscribe and save take rate is often more important than any other metric as long as subscription life time contribution margin is meaningfully more than OTP.
2. Partnership ads and AI still ads are the future. They now almost always perform the best. The middle continues to get squeezed.
3. Product and offer matter more than any growth strategy. Too many brands try to out media buy a weak offer.
4. Building a brand can be enhanced by growth marketing but the best operators we work with lean heavier into brand and product than purely advertising. Growth marketing as a core skill can often get to 8 figures but rarely can take a brand to 9.
5. This is probably the most exciting time to be in Ecom. So many brands are thriving and the tools to build are getting better every day. So excited to see how agents continue to revolutionize the DTC space over the next year.