Dirk Sahlmer

Acquiring SaaS companies @ saas.group | saas.wtf Newsletter

Engineer turned M&A professional. After 20+ transactions on the buy-side, I'm now helping tech founders run a structured exit process.

Posting regularly here on LI and in my saas.wtf newsletter on SaaS and  M&A topics.

Head of Origination

Head of Growth @ Default

I am an ex-professional Olympic athlete who sold a dozen vintage Louis Vuitton bags every month to support my journey towards a shot at the Olympics.

When no one wanted to hire me, I launched my first outbound prospecting agency, but deep down, I wanted to build software.

After sending many cold emails, Jason Calacanis gave me a $100k check to build the next generation of sales productivity software.

Ultimately, I failed, even though we were able to grow to $5k MRR in just four months. I learned so much about how to build a great product, make sure people want it, and how to grow (fast).

I decided to look for a marketing role, but deep in my heart, I wanted to move to the US. After years of trying, I can finally call myself a proud immigrant.

Stan Rymkiewicz

The first time I heard “I thought Default was bigger,” I just assumed our marketing was working. But after hearing it a few more times, I started looking for the reason behind that impression.

3 years ago, I made a bet that we needed to be “everywhere.”

I knew I couldn’t do that by myself.
But with a good product and a solid story to tell, all you have to do is be known.

Happy to report that this approach paid off!
And now, in hindsight, here is what I did right:

BOFU on all channels → grow only TOFU

For each channel I built, I first focused on capturing and converting the demand we already had. After that, I shifted 100% of my attention to top of funnel.

It took me 3 years.
It wasn’t perfect, and I learned a lot...

I’ll be sharing my exact framework, lessons, and insights in a webinar with Factors.ai on Wednesday.

Link below.

I built a system that automates all of the non-creative work of creating ads.

Over the past two years, I’ve spent $500k on LinkedIn ads, which typically drove about 40% of our pipeline.

One of the reasons, they worked is that I try to make "funny" ad.
It's time-consuming. And every month, I’d spend a week making a few.

This time, I spent my entire week building out my “ad creative system”.


I think too many people are trying to let AI handle every aspect of the task. 

My assumption is that I can probably automate 90% of the work,
while focusing on the things I “still” do better than AI:

Creativity.

Here is the process I built: (results below)


STEP 1: DATABASE

I’d done the same process multiple times, so I knew I needed to build a system I could leverage rather & scale. I chose a Google Sheet as my backend because it can easily be updated with a webhook, but I’ll probably move it to Supabase.

Here is the structure I used:

- Signals (insights worth telling in the ad)
- Emotions (what emotions I want to evoke)
- Hooks (what makes ads funny)
- Ad concepts (my previous ads)


STEP 2: SIGNALS

I needed a way to get insights from sales calls, communities, and prospects that could be used as a baseline for the ad. I built a flow that pushes all our Gong calls to a spreadsheet. I also run a weekly scrape of Reddit and other forums. My initial run found over 100 signals.


STEP 3: HOOKS

The best ads are funny, without trying to hard. They're effortless. They reuse things we say, do, and feel, and make a connection to the product. I had Claude find 100s of these in four categories:

- everyday analogies
- cultural references
- memes
- quotes

This gave me over 400 hooks I could use for ads. Then I manually ranked them and removed the cringe ones.


STEP 4: AD CONCEPTS

Once I got my signals and my hooks, I ran a Claude Code script that went signal by signal to see which hook made sense and then spit out combinations. Some were amazing (results below). Some were not great, but it was a start.

RESULTS

As for performance, I can’t say yet, but I’ll report back. The flow saved me a few days’ worth of work. The ad design was still done by me. 

Next up: automating the ad design process!

AI can make anything. It just can’t tell you if it’s good.

That’s taste. And right now it’s the most under leveraged.


Taste is matchmaking of preexisting patterns. 

AI doesn’t have patterns. It has probabilities. It picks the most statistically average output. 

That’s not how you stand out. Here are 3Cs of taste making:

Curate — what you consume. Your references are your ceiling. I stopped looking at my own industry and started pulling from film, architecture, packaging, editorial. Anywhere that made me stop and feel something.

Copy — how you practice. I take things I admire and rebuild them from scratch. Not to publish. To train my hands to understand what my eyes already like.

Create — where you diverge. Now I make my own thing. If it passes my taste filter, I ship it. If it doesn’t, I loop back to curating.

AI made execution free. 
Taste is the last expensive thing you own.

taplio

The Battle

Dirk Sahlmer

Stan Rymkiewicz

Followers

Posts

Engagements

Likes

Performance ScoreStan Rymkiewicz's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Dirk Sahlmer

Stan Rymkiewicz

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);}Stan Rymkiewicz's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreStan Rymkiewicz's LinkedIn Performance