Jake Abrams

I help 7-9 figure DTC brands with revenue & CAC | I write The Process, a newsletter about modern growth marketing | Sold 100K+ personal cooling devices

Consumer Marketing & Growth Leader.  

Brands I've worked with:

Caraway
Jones Road Beauty
Skylight
Canopy
Insurify
Naked Wines
Rugiet
Sundays For Dogs
Momofuku
Sunday Lawn Care

100+ others

Book a discovery call in the featured section below 👇

Founder

The lean DTC growth marketing partner that I wish existed when I was an in-house marketing lead. 

At Odyssey, we help 7 & 8-figure DTC brands increase revenue & ROAS by finding the right combination of:

- Messaging
- Audience targeting
- Media channel allocation
- Creative

https://www.growodyssey.com/

Comfrt Clothing, Frost Buddy, and Underbrush are all using the same strategy to consistently go viral on TikTok while building massive creative pipeline.

Some marketers call it "nichemaxxing." I've created a 19-page playbook to help you deploy the exact same strategy for your brand.

Here's what's inside the Brand Nichemaxxing playbook:

- 4-tier content matrix from AI to hero content
- PDA framework for building micropersonas
- Full tech stack: Higgsfield, Genviral, Octo, Snowball, Discord, etc...
- Creator outreach templates that actually get replies
- Cost structures for every creator tier
- Winner selection framework for scaling organic to paid

The idea is simple: own the interest graph first, then amplify with paid.

Most brands do it backwards. They spend on ads before they understand what resonates.

This playbook walks you through 6 phases: niche definition, account setup, content architecture, creator recruitment, production rhythm, and measurement.

Get the 19-page playbook for free. No email gate. Just take it and use it.

1. Follow me
2. Like this post
3. Comment "playbook" and I'll send you the link.

Nichemaxxing across paid & organic social for 8-9 figure consumer brands | Writing about AI & advertising | Sold 100k+ personal cooling devices

This AI agent analyzed 12,500+ customer reviews and support tickets for an 8-figure DTC brand, created its own documentation for overcoming objections across hundreds of SKUS, then just started doing it...directly on the PDP.

Using AI like this can mean the difference between a 2.8% CVR and a 1.7% CVR. 

Essentially an agent that lives on your site, is trained on your catalog and every meaningful customer touchpoint, that can answer every question in 20 seconds or less.

These agents can handle conversion killers like:

Product fit: "Will this luggage fit in Delta's overhead bin?"
Comparisons: "Hard shell vs soft-shell for business travel?"
Technical specs: "Does it have 360° spinning wheels?"

I'm giving away the an 8-minute Loom video showing use cases for this type of agent, plus a guide to setting this up on your Shopify site in 15 minutes.

To get the video and guide:

1. Connect with & follow me
2. Comment "agent" below
3. I'll DM you the link

#DialogPartner

undefined undefined

The smartest people at Meta will tell you to build ads around motivators, not just micropersonas. 

I like this shift to motivators, because it forces you to start the creative process from a better spot.

Personas make you think about demographics. You end up with a bulleted list about a fictional human named Sarah who likes yoga and shops at Whole Foods. Nothing about that tells you what makes Sarah pull our her credit card and spend $150. 

Motivators do, though. 

When I do creative strategy now, I start by mapping out the actual motivators behind someone would buy or be interested in a brand. 

So for Sarah, maybe it's "I need an outfit that works for the December holiday party."

Then for each motivator, you map one core benefit that delivers the solution. That way, you build the creative around that feeling, not around "Sarah."

Then I do it for another motivator. And another....

By the time you've worked through 4-5 of them, you've got a full portfolio of creative assets, each built around a different reason to buy. Different themes, different visuals, different emotional angles. All for the same product.

That's the part people miss, I think. You're not trying to reach different types of people. You're trying to speak to different reasons to take action. A single product can have five completely different ads that all make total sense.

I feel like that's how you scale creative without running out of ideas or making yourself crazy with format multiplication.

Turn 35 today. Sheesh.

Here are 35 takeaways I have from the last 6 years as a "non-business" person building a business with 30+ people:

1. Expect daily disappointment
2. It's ok to be demanding if that's the clear standard
3. The ideas in your head are smart
4. Respect the audience's attention
5. Show undeniable proof
6. Work with challengers
7. Build systems for 50x output
8. Stay on the forefront of technology
9. Extreme bias for action
10. Creative generalists win now
11. Get over fear of speaking to your phone
12. AI + bad taste = epic fail
13. Rip structures, use original thoughts
14. Incentives drive everything
15. Pretty much everyone is winging it
16. Overcome initial content creation friction
17. Grow your own accounts first
18. Authenticity on Zoom is a superpower
19. Daily follow-ups with blockers, Mr Beast style
20. Work should feel like video game
21. Non-business people build differently
22. Ideas die without proper execution
23. Attention is the most sacred currency
24. Charge more than feels comfortable
25. Comfortable teams ship mediocre work
26. Systems unlock exponential output
27. Tech resistors always lose
28. Speed + direction = unfair advantage
29. Generalists are more important than specialists
30. Loom videos beat meetings
31. Tools don't fix shitty taste
32. You can only get so far ripping other's ideas
33. It's ok to start ugly
34. Practice what you preach, or else you're a fraud
35. Build leverage before you need it

---

Feel free to follow along as I document the path to building a $10M+ agency in 2026 by helping 50+ brands manifest the next era of paid social.

https://lnkd.in/epcpxd4V

I built a 250-video visual script library application in a day with Replit, Airtable, and Gemini. 

So I recorded an 8-minute Loom video outlining the entire process. 

I'm not an engineer. I did run marketing for a coding bootcamp, and maybe that experience helped me overcome the fear of "being technical."

Here's what I cover in the Loom video:

- Why my Airtable database was useless until I made this
- How I used Replit to put a visual frontend over raw data
- The hover-to-play feature that made this actually usable 
- How it auto-analyzes any TikTok or Dropbox link in ~60 seconds
- The prompt I use to extract hooks, transcripts, and full scripts
- How we use it to rewrite winning ads for new brands
- The N8N nightmare I replaced with this

I've been in marketing since 2013 and run a creative agency working with 50+ brands. This is the internal tool I wish I had years ago. 

Want the full 8-minute Loom walkthrough?

Follow or connect with me
Like this post
Comment "Build" 

And I'll send it over.

---

Repost for priority access.

I know an AI agent that sat on a DTC brand's PDP for 2 weeks, captured every question from shoppers, then showed the brand exactly why visitors weren't converting.

Using AI like this can mean the difference between a 2.8% CVR and a 1.7% CVR.

Essentially, this is an agent that lives on your product pages, answers questions in real time, and logs every conversation as a data point. After 2 weeks, you have a clear picture of what's blocking purchases within an easy-to-understand dashboard. 

These agents can handle conversion killers like:

- Compatibility: "Is this compatible with my device?"
- Confusion: "What's the difference between X and Y?"
- Fit: "Will this actually work for me?"

85% of this feedback normally dies in chat logs. But agents like this can turn that data into an actionable CRO roadmap.

I'm giving away a step-by-step Notion document that explains exactly how to set up this system on your PDP’s… 

So you have AI handling conversations, running an analysis on why people didn't buy, and delivering that to you in a dashboard with visualizations.

Works for any Shopify brand doing $1M+. Free to install, takes ~15 minutes.

To get the Notion doc:

1. Connect with and follow me
2. Comment "insights" below
3. I'll DM you the link

#DialogPartner

One of our creators just hit 50k views on a single TikTok.

She has 6,500 followers. She's paid ~$500/month for ~25 posted TikToks plus bonuses based on views her content drives.

This creator is an actual customer. She had a real transformation with this product. The proof is undeniable in the video she posted, and that's why it worked.

You can't script something like this. You can't hire a generic UGC creator to fake it. You can't recreate it in Higgsfield.

She's in our Discord community, where we're sharing tips for going viral and building a community around helping people earn by sharing their authentic stories.

It's all real. No generic content or fake testimonials.

This is the nichemaxxing strategy for 2026.

Massive communities of real people within a specific niche who you coach to go viral on TikTok. Everybody wins.

You have to be doing this as a brand in 2026.

The creative bottleneck doesn’t exist in 2026.

Learn:

Claude
Octo
Wispr Flow
Snowball

Combine with:

Exceptional creative taste
Operational psycho-ness
Next-level communication skills

Might sound harsh, but you gotta figure it out or you'll get left behind.

It takes you way too long to write scripts, and it's not your fault. 

It's because you keep having to re-feed the context to your creative strategy copilot every single time, and that's extremely annoying to do. 

Instead, run use this 3-part architecture that I custom built out for my agency:

First, work backwards from a format you already know worked. Have Gemini analyze the entire video. 

Second, you pull that context into Claude, which already has project files about everything you need to know about the brand you're writing for.

Third, your Claude needs a set of overarching skills that sit within your Claude organization that tell it how to write like your agency or you. 

When you do this, the output is actually good. It's not a shitty first draft. 

Because Claude already has the direction...what made the original format work, how to write like the brand, and how to be good at creative strategy. 

That's the architecture.

I recorded a quick video walking through exactly how I set this up. Watch it if you're tired of re-explaining your Claude how to write like your brand or write like a real creative strategist.

Here's the embarrassing truth that a lot of "AI is replacing everything" bros haven't talked about honestly...

Making good AI video content at scale is actually kinda hard.

Getting prompts right is a skill. It takes time to develop. You have 8-12 seconds per scene. Iteration is slow. A lot of the output you get early on is going to be unusable and you're going to waste a lot of time. 

Compare that to certain types of content that are just genuinely faster to shoot on your phone.

For example: I've been making 7 talking head "yap" videos per week for my own TikTok that I've grown to 27k+ followers. I just pick up my phone and speak for 60-90 seconds. And then it's done.

Why would I spend the time to clone myself and try to recreate something with AI tools that will always feel slightly off, when it doesn't save me any time or cost? Plus people will hate me for AI-ing myself. 

How about if you're Comfrt Clothing, and your top performing TikTok affiliate videos are just Gen Z-ers in their bedrooms taking selfie videos? They just show off their hoodie, show off a few features, and add a caption. That's a 15-minute shoot on an iPhone.

There is no AI workflow that beats that for speed, cost, and authenticity.

So before you burn 20 hours crafting an AI workflow for videos that are easier to just shoot with your phone... ask yourself if it's actually worth it.

A Matthew McConaughey & Timothée Chalamet interview recently went viral, in which McConaughey said that the moral plea against AI is not strong enough and will lose out to the economic incentives.

No lies told there. 

But here's what might actually slow it down:

Everybody hates "AI passed off as real" content.

No major brand I know would unleash an army of AI influencers right now who are pretending to be real. That nuance is everything. The brand risk is enormous.

Maybe that changes in 10-15 years when Gen Z and Gen Alpha are the primary consumers and the cultural relationship with AI-generated content has shifted.

Impossible to actually know.

But in the next year? 

AI takes over certain categories, but humans retain control of most categories.

Because authenticity and emotional connection is still what builds brands. 

Faking it has massive consequences.

DTC brands, I would STOP making $300 UGC videos and immediately firing them into your ad account.

This used to be a logical strategy. But in 2026, you're missing the bigger opportunity.

I've worked with 100+ consumer brands, and so many are spending 100% of their time "creating content for paid", and 0% of their time deploying it organically first.

If your ad account is full of unspent creatives, do this instead:

- Scrape your customer reviews for content ideas using Motion (Creative Analytics) or Claude
- Scrape organic TikTok for winning angles using Snowball
- Talk to your ad account data using Octo
- Identify good creators with Meta creator marketplace
- Set up a Discord to manage and motivate them

Then:
- Get creators on performance deals where they post to their own accounts
- Build your own network of slideshow & faceless accounts
- Deploy every creative test organically first
- Prioritize putting paid spend behind content that earned organic views

Test this for 30-60 days.

You'll see massive improvements in creative volume & winning creatives while reducing wasted ad testing spend. Plus you'll build a massive halo that lifts your entire business. 

Let that drive which content deserves your paid budget.

You're stuck at $250k/month in Meta spend because you don't understand exactly what's happening inside the system you're feeding money into. 

Let me break down the 4-model architecture running your ads:

ANDROMEDA
This is the retrieval engine that scans tens of millions of ads to find candidates for each impression.

GEM
This is the ranking superbrain that takes Andromeda's pool and decides which ad wins the impression.

LATTICE
This is the consolidation layer that connects multiple ranking models and shares learnings across Feed, Reels, and objectives.

SEQUENCE LEARNING
This helps model user behavior over time. Not just one impression, the full sequence of actions before and after.

How they work together:
→ Andromeda retrieves 
→ GEM + Lattice rank 
→ Sequence learning personalizes.

What this means if you spend $250k+ / month as an 8-9 figure brands:

- Creative diversity is how you deliver ad candidates
- More variants = more signals = better matching
- The AI handles distribution, you handle inputs

You need operational willingness to hit quality + quantity + differentiation b/c that's what the entire system runs on.

I built my first real vibe coding app with Airtable, Replit, and Gemini. 

Process was:
- I built out this big Airtable databse
- Gemini analyzes any video and put the analysis in Airtable
- I created a front end for this using Replit

You can hover over the videos, and they auto play. You can open them up, and get the script. .You can even copy a prompt that you can preload into Claude
to do a rewrite of that video for your specific brand.

The prompt auto checks the project files in Claude that tell it how to write like that brand, and then it does a rewrite and gives you all this extra helpful context.

This was a massive game changer for my process of:

- Working backwards from a format
- Adapting it for the brands that I'm working with
- Applying my creative taste to edit the content in a human way afterward

"Content creators are toast, AI is going to replace all of you."

Here's why I believe this take is fundamentally wrong...

I run an AI-native creative agency. I use AI every single day. And I genuinely believe AI is one of the most important shifts happening in our industry.

AI will infiltrate certain categories and never touch others.

Any animation that is not Pixar-level, top 5%...is probably gone. Why pay $5,000 when you can prompt and tweak something that is relatively comparable?

Same with word walls, cinematic story ads, static creative, memes, and product renders.

But then there's content where the entire f*****g point is that it's real.

Pull up Comfrt Clothing on Kalodata. They're doing $10M+ per month on TikTok Shop. Look at their top 10 revenue-driving videos. None of them are AI. Almost all are real micro and mid-level influencers.

Because the content that works on TikTok is "I am a real person who bought this thing and here's what I think."

The second that's AI, the whole thing falls apart. 

At the end of the day, authenticity has always been the driver of building the best brands.

That didn't change just because it got cheaper to fake it.

Brands that have hired ex-Odyssey team members:

Gruns
Hexclad
AG1
Trip

I don't believe there is a better place to learn modern creative strategy, as long as you're hellbent on joining the top 1% of humans at this skill. 

We're hiring again for a critical position to help us deliver on our goal to elevate 8 & 9-figure consumer brands into the top 1% of their category for creative diversification & distribution. 

Apply here: https://lnkd.in/eYex6PPW

To apply: send a Loom video showing proof of two things:

1. A TikTok Shop you've managed...walk us through the results & what you did personally to grow it.

2. A Discord creator community you've managed...show us the server, your activity, and how you kept it engaged and growing.

"Do you get brands from posting on TikTok?"

I don't even know what to say back, cause I don't think of things that way. 

All I think about is:

- The thoughts in your head about marketing are actually smart
- If you pair them with the right hook, people will pay attention
- Just keep putting that out there, into the world
- They will find the right people
- When they're ready, they will just reach out
- I have a better chance than someone they never saw in their feed

That's literally it.

Let's say you need to build 10+ affiliate programs from scratch.

The goal: onboard 30+ micro-influencers per brand within 15 days.

Here's the setup:

Target: creators with 5k-50k followers already posting in each brand's niche (not random people)

Half the brands sell on TikTok Shop. Half don't due to business model constraints.

Compensation structures vary by creator:
• Commission-only + free product
• Retainers for high volume (30 posts/month) + view bonuses
• Retainers + sales commissions

Requirements:
• Full content buyout for ad rights across all platforms
• Dead simple process for sparking and partnership ads
• Ability to manage multiple programs simultaneously

Nice to haves:
• Payment management
• Dynamic creator payments based on CPM model

What's the 10/10 tech stack for this for < $2,500 / month?

Is it:
• Just Upfluence
• Just Aspire
• Just SARAL - The Influencer OS
• Refunnel + a CRM like Streak 
• Social Snowball ☃️ + a CRM like Streak 
• Superfiliate + a CRM like Streak 
• Some other combination
• Something else entirely

You need operational efficiency here. 

What have you used at true scale that actually works?

Upfluence

Aspire

SARAL - The Influencer OS

Refunnel

Streak

Social Snowball ☃️

Superfiliate

Jack Dorsey fired 40% of Block staff "because of AI."

Here's exactly what to do next if you're afraid of being next...

18 random thoughts from growing my TikTok from zero to 26k+ followers in a few months.  

- I kinda think you're a fraud if you make content for a living but are unable to grow your own social accounts
- Every thought-out LinkedIn post is a script for TikTok
- Unscripted TikToks become more LinkedIn Content
- Just fire off that "hot take"

- Audio quality matters more than camera quality
- Just buy the RØDE mic, it's awesome
- Making your own videos is best way to become a better CS
- Every CS should try to grow a TikTok, even it's a faceless account

- The hook is everything
- The hook should be oddly specific: use keywords and numbers
- TikTok "categorizes" you, and it becomes easiest to get views in that category
- The random thought in your head is good enough for TikTok, as long as you have a specific hook

- It's easier to get views on TikTok than LI w/o resorting to "lead magnets"
- It's possible to work out and make 3 TikToks during my kid's naptime
- People will watch and comment even if it's just a talking head video
- Your audience is on TikTok b/c everybody is on TikTok

- TikTok drives real business opportunities
- I had 3 leads comes through last week who discovered us on TikTok

Anyway, those are my thoughts.

Certified Clay Expert & Agency owner | I'll help you build a scalable outbound engine that gets meetings booked on autopilot in the next 180 days | Visit Youtube: @growth-today to see how

Growth shouldn’t feel random.

Most of the B2B companies hit the same walls:
Inbound leads that don’t convert

Outbound that burns reps out without results

Partnerships that look good on paper but never drive pipeline

A CRM full of noise instead of reliable data


I’ve seen it across founders, sales leaders, and marketing teams. The problem isn’t effort, it’s the lack of a repeatable, scalable GTM system.
That’s what we build at Growth Today.

Instead of disconnected tools or random campaigns, we design and implement repeatable GTM engines across sales, marketing, and partnerships that…

✅ Enrich and qualify data
✅ Activate inbound leads into real opportunities
✅ Automate outbound prospecting
✅ Keep CRMs clean and reliable
✅ Orchestrate multi-channel campaigns
✅ Make BDR teams more effective or fully replace them
✅ Align marketing + partnerships with sales for compounding growth

Recent wins:

 → Built a clean, automated CRM that doubled conversion rates
 → Launched a content-driven demand gen engine that 3x’d inbound lead quality
 → Systematized partner outreach → pipeline from alliances grew 40% in 6 months

Growth doesn’t come from more outreach or more tools.
It comes from building the right system once and scaling it.
If you’re a B2B founder or revenue leader serious about building predictable pipeline, let’s connect.

Book your call: https://calendly.com/growth-today/intro-meeting

Brigitta Ruha

Efficiency is part of building any company. I get it.

But the problem starts when GTM spend goes down… and revenue expectations go up.

Being disciplined about CAC makes sense. But reducing spend removes the crucial part of building, scaling GTM - all the reach, experimentation, and ultimately - staying visible on the market ✌️

all of which eventually influence the CAC, the efficiency….

Predictable pipeline usually sits on top of a lot of unglamorous work:

- Ongoing testing
- Consistent market presence
- Regular feedback from the team, customers, and prospects

When those inputs shrink because we cut down on costs on revneue team, the impact doesn’t always show up immediately. It shows up a quarter later (fewer deals, thinner pipelines, longer sales cycles, etc)

Revenue rarely grows when the effort behind it shrinks. Reducing costs can absolutely work

but only if the system itself is improved, not just scaled back

P.S. How do you balance cost discipline with maintaining momentum?

Co-Founder@Growth Today | We Build Repeatable GTM Engines Across Sales, Marketing & Partnerships 🚀

Before you spin up another outbound campaign this quarter, please:

Do not start by writing new copy.

You’ll end up polishing something that was never the real issue. I promise you.

It’s completely normal to see outbound look like this:

• Emails going out every day
• Activity numbers looking healthy
• A few replies trickling in
• But pipeline still questionable

Some weeks look promising. Others not.

If you’re about to reset outbound (or you’ve been tweaking campaigns but nothing really moved the needle), here’s a simple 7-day reset we use:

Day 1 – Fix the list
Day 2 – Fix deliverability
Day 3 – Fix the offer
Day 4 – Fix the CTA
Day 5 – Fix personalization
Day 6 – Fix sequencing
Day 7 – Fix measurement

(my design team put together this image)

Track these automatically:

Inbox rate (did your email land in the inbox?)
Reply rate + positive reply rate (was your message relevant to them?
Meetings per 1,000 sends (was your follow up timely enough to book them?)

Yes, you want more meetings. But consistency usually comes from fine-tuning the fundamentals

& outbound has a LOT of those levers so first, take a look all these levers.

if i were you: I’d update the offer first!

How can I be SO irresistibly helpful to my prospect?

→provide that value in the email : ) & that can be anything from an audit to checklist to a designer bad

hit the 🔔 on my profile if you want more actionable, no fluff, BS tips like this

We've ALL made every GTM mistake in the book

I’m sure you noticed it is never about the tech itself 

it's getting people actually to use them 

And we learned this the hard way… 

→ change management matters more than the system itself 
→ not everything needs to be automated 
→ bad data is way more expensive than you think 
→ outbound without relevance is just noise 
→ partnerships need as much system as the sales team (this team is often swept under the rug)

.. a few tips that can go a long way 👐

DM me if your team needs some expert guidance, hands-on execution

Co-Founder @Growth Today | We Build Repeatable GTM Engines Across Sales, Marketing & Partnerships 🚀

5 THINGS WE WISH WE KNEW

Woah, this is happening right now.

Clay is officially here in London 🎡 !

I remember 2023 when I first spoke to Varun Anand.

I felt !! FINALLY !! a tool built for growth.

for Growth teams.

for high-growth companies

Who are all SOO keen to use the data, but what if more business ppl could use it instead of just relying on techy guys from sales engineering backgrounds?

Suuuupperr excited to be here at Clay’s EU/London office and Europe.

& grab some ☕ with Jani Vrancsik Tyler Swanson, Ziqi Deng, Juan Ravagli, and Michael Vitak

We’ve been experimenting with AI in our GTM systems recently 👀

There are 3 different scenarios when you want to build autonomous systems:

chat w ChatGPT or Claude (e.g.: threads)
autonomous workflow (i.e Hubspot or agentic workflow)
ai agent that executes something very specific end-to-end

Let’s say you want to schedule a quarterly check-in with a client.

The first version looks something like this:

1. Assistive setup

So you open the CRM. Check your calendar. Ask ChatGPT to draft the email. Edit a line or two. Send it. AI helped you write it.

2. Automation-based

Now the system does most of it. After 90 days, the workflow runs. It checks the calendar. Writes the draft. Sends the email. Simple. Less work for the team. But it still follows a script. If something unusual happens, it doesn’t adapt. Your ops team needs to debug and fix it.

3. Autonomous Agents

Instead of telling the system, “Send the email.” You just provide the overall goal: “Alert if a customer stopped leveraging the product.”

It watches how the client is using the product using the data from your Snowflake/AWS database. Notices when something looks off. Sends a follow-up to prevent churn. Flags the account for potential churn if the Account is customer Success managed. Provides a summary for the CSM to prepare the conversation. Now the system isn’t just completing a task. It’s working toward an outcome.

& that’s a very different thing!

im gonna share our agentic agents as well, we build internally across sales and marketing ✌️
Where are you experimenting right now? 👇

I was thinking about this over coffee this morning ☕

Most GTM teams don’t struggle with ideas anymore.

The playbooks exist.

Targeting logic is solid.

Sequencing frameworks are mapped out nicely.

The weak point is the execution and/or strategy layer.

Most stacks today are deeply connected: CRM, enrichment, ads, content systems, sales tools.

Everything talks to everything. But connected doesn’t mean autonomous.

Someone still has to write an email
Someone still has to decide when a workflow should run.
Someone still has to follow up and follow through.

That’s where things get fragile.

I’ve been testing FlashLabs SuperAgent recently because I wanted to see what happens when that decision layer is automated.

What works here isn’t that it writes emails. (Plenty of tools do that.)

It monitors public and social signals, understands curiosity vs. buying intent, and then triggers predefined GTM playbooks 👀

Which means:

– detecting early evaluation signals
– launching the right outreach sequence
– generating landing pages or ad variants
– updating CRM fields
– preparing follow-ups using meeting intelligence

End-to-end. Without someone babysitting the workflow. (or the team)

& it won’t be 100% pixel perfect

But it does get the job 80% of the time

which is AWESOME

Because if you do 80% best version out of yourself/system

You can buy back SOO much time

With consistency, which will give you a * huge * compounding result over time

This is like the early versions of an autonomous GTM layer.

Still experimenting with it. ✌️

But if you’re a GTM leader thinking about how execution scales

…not just how strategy looks on slides, FlashLabs SuperAgent is worth checking

Link pinned in the comment 👇

random question ☕️

When was the last time you actually trusted EVERY (or any?) field in your CRM? 😅

Not in theory.

But like… truly trusted it.

When you go to the CRM, you start to notice these small things piling upppp 👀

- Titles slightly outdated.
- Mobile numbers missing.
- Company names that don't even exist anymore (rebranded, acquired)

& that’s enough friction to slow things down. 

So instead of planning another BIG CRM sprint, we connected prospeo.io to auto-enrich all companies and contacts directly inside the CRM.

Now, when a new contact comes in:

- Default fields get filled out
- Email address is verified through multiple providers
- Details are refreshed when the person gets promoted, moves to a new employer

(Automatically.)

we still review things occasionally

we still care about governance (LeanData is great for auto-de-duping companies/contact btw!)

but we’re not exporting CSVs back and forth anymore. 🙃

It’s just… CLEANER + EASIER + MORE MANAGEABLE by default.

& saves a lot of time w/o second-guessing the data all the time.

CC’ing Matthew Putnam you can reach out to him in DM directly if you need more 👀

I ran my first outbound campaign 6 years ago. Drove 75 demos/mo for 2 E-Commerce SaaS as their growth leader, and then my cold email agency from 0 to $2M+ ARR. Here are 7 things I wish I knew before starting:

(1) Truly nail down your lead list

Sending your outreach messages to the right person matters more than anything else:

→ Precisely figure out what problem you're solving.
→ Find companies struggling with this problem.
→ Identify relevant individuals in these orgs.

(2) Give as much value as you can

You’re stepping into someone’s inbox uninvited.
& the fastest way to earn attention is to be USEFUL.

Give give give > ask

A simple approach:

→ Identify their problem
→ Share one practical solution for free
→ Do this as many times as you can on all channels where you are.

(Think through: why are you reading this LinkedIn post, right? EXACTLY!)

(3) Pick only one prospect and write for him/her

Many people try to write for 500k+ people.
Start with one instead.

Pick one prospect you’d genuinely want to work with.
& write as if they’re the only person receiving that email.

Then do the opposite:

→ Find 20 similar personas
→ Find similarities in their challenges, org size/org challenges
→ Then automate all these at a bigger scale

(4) Take care of the deliverability

Good deliverability doesn't guarantee your campaigns will succeed. But...

Bad deliverability guarantees your campaigns will fail.

Make sure to...

→ Send a few emails per mailbox (< 20 per day)
→ Set up MX, DMARK, SPF, DKIM records.
→ Disable open tracking.
→ Warm up mailboxes (bare minimum 14 days)
→ Weekly run inbox placement tests
→ Daily run domain blacklists checks

(5) Take the extra mile!! 

Any GTM channel is somewhat hard to crack. The same goes for cold email.

Make sure to go above and beyond when reaching out to your prospects:

→ Look at triggers, signals, current events, org. charts.

→ Reference insights from the job descriptions.

→ Don't stop at 1-2 email variations. At Growth Today, we frontload 8 variations

(6) Complement cold email with other GTM channels

Cold email is my favorite channel because it's cheaper to get started, and you can find what works a lot faster than with any other channel.

But it works even better with other channels.

→ Run ads to the same people you're cold emailing. warm them up! ♨️

→ Make great, authentic, and genuinely useful content and use email and LinkedIn to distribute it

→ Use cold emails to distribute valuable lead magnets.

(7) Don’t just use technology, leverage it to SYSTEMIZE

→ Have a look at growthtoday.co/sales-tools (or my fav one: our blog posts that include copy-pastable email templates)

→ Do a few tests manually to see what is the most interesting angle on the market to pitch your prospects.

→ Before hiring more people, ask: “Could I achieve the same with technology?”

→ Before doing more, ask: “Could I build a repeatable, scalable SOP or system instead of me doing this again and again?”

7 TIPS TO GENERATE MEETINGS WITH COLD EMAIL

“I think we’re losing more deals to bad data than bad messaging.”

If that is true,

I can feel you!

The usual process of getting good and accurate data somewhat requires this:

- One tool to find them.
- Another to verify.
- Another tool to find more because the previous ones didn’t find enough
- Export. Upload. Repeat.

It works, but unnecessary very tedious.

& never really knew which source actually found the valid email.

This is what I like most about BetterContact:

Instead of: “Let’s try provider A first…”

1- its email waterfall runs automatically
2-  through 20+ providers 
3- at scale, through the API (through Clay, or your CRM, AWS/DB) so you don’t need to import/exports

If one doesn’t return a valid email → it moves to the next. If it fails verification → it keeps going. Until it finds something actually usable. & you’re only charged IF the email address is actually found and valid.

If you’d like to save time on using 5+ different tools with 3 CSVs whenever you need an email, def try out BetterContact. ✌️

It’ll save you SO MUCH time.

p.s. how are you running enrichment right now?

4 things I did because I trusted my gut (not the people around me)

1-Started building programmatic prospecting in 2020 (way before Clay was cool)

No one understood why I'd struggle with no-code tools and Python when you could "just hire BDRs

After doing it full-time at a company and through consulting, when Clay popped up on my feed, I felt like I could create something different to help more people succeed with outbound.

Way easier than before, BUT leveraging all the experience I gained over the years.

2- Posted frequently on LinkedIn when everyone said it takes forever to see results

By the end of year 1, I was consistently building a pipeline from inbound, which eventually led to building Growth Today (and still keeps us going when the market gets tough).

Now I have CSM, GTM engineers, a team who assist me with content, and with sales, and freelancers helping me build what we believe in every day.

3- Let custom RevOps projects go

The team closed a few deals with enterprise customers.

On paper, it's awesome: no churn, high expansion, upfront payment…

But in the end, it didn't fit our ICP or the service we wanted to build at the time.

Letting go of some customers is hard, but necessary in the long run, depending on what route you choose to build your company.

4- Hired people who didn't have the skills yet, but had the potential

I always get DMs, and sometimes I make exceptions.

There was one person I almost fired in her 2nd week. I told her upfront: if we're both not getting value, it's hard to bring positive outcomes, and none of us will enjoy this work relationship.

I was candid, she was too. She mentioned she should leave herself if she won't hit my expectations.

She did it.

A-players are important, but sometimes your gut tells you to take exceptions.

Eventually, some people float to the top, and some just can't get you to the destination.

What about you?

What are the things you did because you knew better?

what we changed to fix their outbound

One thing I love about outbound is how quickly the truth shows itself.

If the system is off, the results will drag. If the system is right, everything clicks.... even if nothing else changes.

Last week we worked with a services company whose outbound “felt flat.”

But once we dug in, the problem wasn’t just list or the copy
…it was that their system was fighting against them.

Here’s exactly what we changed:

1️⃣ We stopped selling the service at scale only a few people respond to

They have 3 services.

All great.
But one is a terrible conversation starter.

Cold prospects don’t wake up thinking:

“Wow I really hope someone offers me a 40-page report today.”

Their prospects literally respond ~2× better to the other 2 services, but outbound wasn’t aligned to that.

So we repositioned around their two strongest conversation starters

2️⃣ We fixed the invisible killer → deliverability

Their audience = mid-market + enterprise.

Enterprise = Microsoft.

Microsoft = “I will evaluate your entire family tree before letting your email through.” 😅

So yep… we’re have a different mix Google + Microsoft inboxes and all these inboxes are purchased to multiple resellers, so we can continously monitor which ones are performing the best and use the data from our inbox placement tests to make adjustments to keep hitting the inbox.

3️⃣ We replaced the “quick chat?” CTA with value

Their CTA used to be a meeting request. (in a soft way which is GOOD).

But (especially in the agency space), you’re likely selling something that is “easy” to execute, so the barrier to entry is low.

So we introduced CTAs that give something instead of ask for something:

- campaign ideas
- roundtable theme maps
- insight drops
- quick audits

Concrete example if you’re running an ad agency:

→ “I have analyzed your ads published in the last 90 days. Can I send over a quick audit to decrease the CPL?”

→ “Would you be open to exploring how we can help you decrease the CPL on your Meta ads?”

GIVE → GIVE → GIVE → before you ASK.

4️⃣ We added intent + timing signals instead of guessing

Before: outbound went out in usual prospecting campaigns
Now: outreach triggers when something meaningful happens.

Signals we added:

→ new senior hires
→ hiring for marketing roles
→ aggressive ad spends
→ thought leadership indicators
→ content activity

5️⃣ We refreshed the messaging (just enough)

Not a rewrite — a refresh.

We updated angles based on:

the market’s current priorities
2026 plans
current challenges in their space
This avoids message fatigue

The fun part?

We didn’t change:

❌ the ICP
❌ the TAM
❌ the channel
❌ the core offer
❌ the send volume

Takeaway:

When outbound stops performing, people jump to copy first.

But the real leverage is almost always in:

→ targeting
→ timing
→ positioning
→ offer
→ infrastructure

Once the system is right, the copy suddenly becomes “amazing.”
Even if it’s the same copy you thought was the problem. 😅

Whenever sales slow down, the first instinct is usually:

Automation won’t fix a broken GTM system:

- Wrong ICP.
- Weak market signals.
- No clear value.

Before we automate anything at Growth Today, we run one test. We call it the Manual 30.

(like Forbes U30 but Manual U30 ✌️)

Open a simple sheet. Add 30 accounts you genuinely want to work with. Not random prospects. Not scraped lists. Companies you’d be proud to call clients.

Then go deep.

Layer 1: Buyer signals. Promotions. New mandates. Ad infos, 10-K report, and their initiatives. What they’re consistently talking about. What’s top of mind for them?

Layer 2: Company signals. Hiring insights. Expanding into new markets. Fresh funding. Sales leadership changes.

Layer 3: Thought leadership context Search their name + “podcast.” Search their name + “interview.”

Long-form conversations reveal operational tension people never post about.

Layer 4: Their GTM stack. CRM. Routing. Scheduling flow. Form logic.

Just by looking at their tech stack, you’ll know so much about how they operate and what GTM context their potential challenges are.

Layer 5: Process gaps. Job descriptions. Reviews. Case studies. Tells you what they’re focused on and what they’d like to actively solve

Layer 6: Job posts. Look at the language they use:

If this doesn’t generate real conversations, doing this at scale (by leveraging automation) will just help you fail faster.

& ONE MORE thing most teams underestimate: Speed-to-lead.

The longer they need to wait for your response, the lower your chances of turning that lead into a real conversation — and a booked meeting.

Strike while the iron is hot 🔥

Every GTM system we’ve successfully scaled started with manual validation first.

Not because it’s cute.

Because it works.

How do you validate targeting before turning automation on?

AUTOMATION WON'T FIX A BROKEN GTM SYSTEM 

If your database charges you PER CONTACT 👀 Your ABM motion turns into credit math 🧞 like this:

Want the CEO? pay. 💸

Need RevOps + Marketing too? pay again 💸💸

Want the buying committee? Now you’re doing FINANCE 🪄, not outbound.

You want to use these tools to maximize the coverage on your target accounts & the whole buying committee

So why pay for each individual When you get ALL of THEM?

With saasyDB: 1 credit = you get the whole buying committee 👀👀👀

The whole org, so you can actually multi-thread all relevant exec, functional leaders, operators …whoever influences the deal.

Stuart from saasyDB realized we shouldn’t be multi-threading to require a CFO’s approval to have access to the prospects at the target accounts we want to talk to ✌️

if you’re targeting SaaS = use SassyDB alone or on top of your existing tech stack

p.s. Does your enrichment provider force you to optimize for CONTACTS… or ACCOUNTS? or you can get both? 👀

I just * literally * watched AI build my workout plan in real-time!!

Prompt 1: "I want isolated exercises to form all parts of my glutes." AI responds with: → Exercise selection (cable kickbacks, bridges, abductions) → Sets, reps, rest periods → Why each movement targets specific glute regions

Prompt 2: "Generate this in the app."
(so that I don’t need to add each exercise to my workout plan. sorry im lazy! 😄)

Done.
Workout is live and ready to use.
This is the best AI implementation I've seen in a fitness app.

I’m just SO fascinated by how easy to use AI in every part of our lives
& honestly I can’t wait to see

-AI cold callers to do the phone calls
-AI to prospects target accounts

I’m already using AI to amplify my prospecting processes, content creation, and ideation

& I can’t wait to see what else it can do
The future is SOO exciting!

Some tools with AI (agents) you can test:

FlashLabs’s super agent for list building, prospecting, and follow-ups
Roman Hipp’s crawler AI agent to give you valid data
Carmen from Scripe to write content on LinkedIn
& to make the most out of these tools

Roxana Irimia Lemlist course on how to build a personal brand
Lara Acosta’s course on how to think about content creation on social
Nathan Lippi 🥾 Lippi or 🌪️ Marcos Stu’s Clay bootcamps and courses on GTM engineering/sales prospecting

Clay’s own university
P.S. NOEX Training if you're curious ✌️

A company came to us with a problem: they couldn't fill their sponsorship slots.

They run newsletters for crypto, gaming, and startup audiences.
But their sponsor pipeline was empty.

The issue? Their ideal sponsors (crypto companies, indie game studios, funded startups) didn't exist in Apollo, ZoomInfo, or any database they could buy.

So they asked us: Can you help us find and market us to them?

Here's what we did:

1️⃣ Email Infrastructure:

We bought domains, configured inboxes, and inbox placement tests to adjust sending based on inbox/spam ratio to maintain their health to prospect programmatically at scale.
If emails don't land, nothing else matters.

2️⃣ Custom data for list building

For crypto companies, we identified crypto monitoring directory sites. Built a custom Python script to scrape all of them. Enriched the companies. Found the right contacts.

For game developers, we scraped the Google Play Store's trending games. Used Claygent to reverse-engineer which studios built them. Then pulled decision-makers from those studios.

For startups, we sourced from different directories and databases and ran look-alike modeling to expand the list.

3️⃣ Messaging

Looked at scalable & cost-effective public data points that would either add value, or prove we did our research instead of just saying "hey, nice game."

4️⃣ Execution, Testing, Maintenance

Launched 6 campaigns.

Ran our frameworks to figure out what worked for each vertical in 90 90-day timeframe. Monitored, adjusted, scaled winners, killed losers.

The results over 90 days:

→ 104 interested responses
→ 2.4-5% reply rates depending on vertical/offer
→ 30% of positive response rate
→ $270K+ in qualified pipeline ($3k ACV on average)

The best part? They now have a repeatable system to fill their sponsorship inventory without waiting for inbound.

With proprietary data that their competitors can't access. 👑

If your ICP doesn't exist in standard databases, you're not stuck.
You just have to get creative about where to find them. 💡

We can write Python scripts to scrape niche sites. Or use tools to reverse-engineer company data. Or build enrichment workflows that run on autopilot.✌️

Proof in the pudding below 🍮

Let’s just send more emails.

The CRM that hasn’t been cleaned since 2022:

Ok, jokes aside.

How to GUARANTEE an unpredictable pipeline:

- Send more emails every time numbers dip
- Ignore deliverability because “we’ll fix it later.”
- Hire more SDRs to drive the car before fixing the engine
- Blame messaging before mapping out your service market
- treat your CRM like a storage unit

& how to help build a PREDICTABLE pipeline:

- Validate ICP before scaling volume
- Test messages manually before automating
- Fix routing before adding fuel
- Treat your CRM as a revenue architecture,
- Reference job descriptions, funding, and K-10 reports in your cold emails
- Treat tools as an access to systemize, not just tools

We can’t fix the pipeline if we don’t fix the fundamentals:

- Better reply rates with more conversations held
- Reduced response time
- Scaled outreach without hiring more reps
- turned “meh” campaigns into repeatable systems
- finally trusted their CRM data

& yes, every “let’s just send more emails” moment usually came right before someone opened the CRM and went… y tho.

So next time when sales slow, and the instinct is:

“Let’s just send more emails.”

Remember to take a look at the car engine first,

not just the car itself 🚗

taplio

The Battle

Jake Abrams

Brigitta Ruha

Followers

Posts

Engagements

Likes

Performance ScoreJake Abrams's LinkedIn Performance

Performance ScoreBrigitta Ruha's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Jake Abrams

Brigitta Ruha

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Jake Abrams's LinkedIn Performance

Performance ScoreBrigitta Ruha's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreJake Abrams's LinkedIn Performance