Zero-Waste GTM

AI will result in cutting 50% of GTM cost by simply eliminating wasted time of people updating tools and each other. Teams that are successful will be able to pour more resources into Pipeline building activites while reaching sustainable CAC Ratios.

Customer Acquisition Costs (CAC) are steadily going up. For everyone, not just public SaaS. But for them, we have the actual data:

What you are seeing here is that for 71 public SaaS, the vast majority is spending $2 or more cash to acquire $1 of additional revenue. And quite a few are in the $4:1 and $8:1 bracket.

The math here is not straight CAC Payback. It also takes into account Churn, so it's really "Net New ARR". If you had high churn, which ate most of your New Business number, then comparing your CAC against that will look very ugly.

But ultimately, doing Net New ARR-based CAC Payback is a very healthy way of looking at your growth.

So why are all these companies out of whack with what we know is a sustainable model?

PipeGen is getting more expensive.

"I don't know it for a fact, but I just know it's true:" all the money is going into top-of-funnel expenses. How do I get to that conclusion?

Every GTM Org (Marketing, Sales, CS) is really just a function of how many leads, deals, and customers it needs to work on at any given time.

If you don't have enough deals for your AE team, you will reduce your AE team eventually. And this purge has already happened in the last 2-3 years.

If you see a lot of churn of customers and don't get new ones, you will lay off your CSM staff.

So the only real bottleneck for every single company seeking growth is Pipeline Generation. Call it DemandGen or Top-of-Funnel or whatever. Getting stuff into the funnel is the big issue.

And everyone can be better and more efficient in processing it through the funnel, but it's not like people aren't trying for years and decades. So this really also isn't the solution.

So, where is all the CAC going? It's being blasted into the top funnel.

Another example of this truth

Another example of this truth and how really the top-funnel impacts a company is monday.com:

Why did it drop last week? Because they said that Organic Search isn't working anymore. Whoops, their major source of cheap pipeline disappeared.

Some people think this is no big deal, but investors didn't. Stock dropped 30-40% in a day.

Why? Well, only two bad options:

• You grow slower without that top-funnel (BAD)
• You invest in more expensive channels to close the gap (BAD)

Stock drops.

The "Cool" Ideas That Won't Work

And there are some "cool" ideas on how to solve this puzzle going forward. I wrote about it in the new playbook that I discussed with Jacco.

While really exciting to read about, I would say that for 99% of companies, this new strategy just won't work.

Really, the idea is to encourage more Word of Mouth from your happy customers.

See? When I put it like this, the whole newness and excitement of it crumbles.

The Growth AI Messiah disappoints

When something systemic stops working, a new Messiah needs to be found. Someone who walks down the hill with some Clay Tables (lol) and tells everyone how to use AI to drive more cheap top-funnel.

Well, the polls are kind of in: It doesn't work.

AI SDRs don't party on Ibiza while "working" remotely, and they don't call in sick. And I would say that most AI SDRs are actually not worse than real ones. Sorry. But I have been working with 100s of SDRs for years, and they don't have a secret sauce - aside from Cold Calling, which most of them actively avoid like the plague.

The problem with AI SDRs is that Emails and LinkedIn inboxes were crowded before, and it's even more so now. I think those AI solutions really work just like SDRs: barely.

And the whole idea of just 10x-ing your SDR team with 1 Agent doesn't work for a simple reason: even if SDRs would cost nothing, adding 10 SDRs into an already saturated market would make no sense - which is why you didn't do it before.

So where do we go from here?

The CFO AI Messiah comes next

You know what AI is really, really good at? Doing stuff you are already doing now, but for 5% of the cost.

Not everything. But for some roles, it can do a lot and replace them - SDRs & Support. For other roles, it can do a lot but won't replace them for a while: AEs, CSMs, Managers, Coaches. For those roles, AI will simply supercharge those employees. One AE can (easily) do the work of 2-3 AEs if supported properly with AI.

But do you still need all of those employees? Assuming that you can't just sell 2-3x more since PipeGen still limits you.

Well, that's why it's called the CFO AI Messiah: you will likely reduce your team by half but keep the same revenue throughput.

And that reduction will really lead to 2 things:

• Companies get back to healthier metrics
• Companies take that additional cash and pour it into PipeGen

As I was wondering what is more likely, I looked into what happened in the last 40 years in terms of automation in manufacturing.

Very similar to what we are seeing. Automation did not increase their top-line. But it reduced their costs heavily. Leading to more competition on price but also more investments in additional automation improvements.

What they did not do was to take that cash and send it as dividends to their investors.

In other words, the automation gains were reinvested. From Salaries for workers to other areas (don't ask me, I am not an expert in manufacturing).

And I think the exact same thing will happen here.

Some teams in the $8:1 ratio will, of course, scale down both overall spending on top-funnel and cost savings from AI.

But the vast majority of teams will use the AI cost savings and simply redistribute them in the organisation.

My bet is on additional PipeGen. But some will pour it into R&D or Product - seeing those investments as "strategic" PipeGen.

Why haven't we seen this yet?

"Hey, we are 2 years into this now, why can't we see it?"

Well, seeing how Microsoft, Meta, and Google are all laying off people despite record revenue growth and massive valuations, I would say we are seeing this already.

In those cases, the main gains come from fewer Developers - I'd argue. AI is already great at writing code.

But we haven't seen it in GTM - I would agree with that.

And here is my logic why not:

1. AI needs clean data. But data is a mess in every GTM.

Messy data is a feature, not a bug. We need to stop pretending that we can just "clean it up". Won't happen. The reason is people. You can't control what they are inputting into the system, and that will lead to bad data.

2. Messy data means you can only do low-impact 1-1 automations.

Real automation on top of messy data is not possible. What people are doing is mostly cute non-critical automations - where if they go wrong, nothing really happens. But that also means the automation itself is hardly valuable.

3. A bunch of low-impact 1-1 automations will be its own messy problem.

Have you ever tried to de-bug a flow error in a heavily automated Salesforce that has been inherited 3-5 times?

I have. The problem is, you pull on one thread and it links in so many ways to other areas that suddenly you don't dare pull more.

And many times the solution was to actually burn the whole thing to the ground and start new.

The same thing will happen to the slew of 1-1 automation tools. I use n8n myself, and it's great. But the feeling I get is closer to an Excel Spreadsheet than anything: it's super powerful and flexible, but it's just a matter of time until the whole thing breaks.

There is a simple solution.

I have previously talked about the intelligence, execution, and orchestration layer.

The team and I went deeper and deeper into this problem, and based on customer feedback, I think we are emerging with a stupid but brilliant solution:

I previously thought you needed to solve messy data before you could start orchestrating the whole thing.

This is why we built the intelligence layer first.

But we have learnt that this is actually a co-dependency.

You need Orchestration for Intelligence to work.

And you need Intelligence for Orchestration to work.

Let me explain:

Orchestration → Intelligence

We solve two super simple and super high ROI problems for new customers right now:

• Run through past transcripts and emails, extract key insights like competitor contract timing, feature requests (might be delivered now), or similar, and add those insights as fields in the CRM.
  → Based on that data, you will get a meeting-booked surge resulting in ROI
• Take every new transcript and extract key fields, autopopulating the CRM.
  → saving 2-3 hours per day per rep, resulting in sustained long-term ROI

We are solving infinitely more use cases, but this solves messy data. Resulting in better intelligence.

We seem to be the only vendor that does the back-filing. And we are recorder agnostic. No need to switch to a 2nd-rate and 5x the price recorder for a Conversational Intel tool.

Intelligence → Orchestration

Intelligence for us means that the AI understands your business. We achieve that with the GTM Graph.

Because the AI understands you, you can simply write a prompt on what automation you want to have. The AI builds it in 1 minute. You hit publish, and it's live.

No need to define the workflow or where to take data from and where to push it. The Intelligence Layer already knows.

Zapier & Co. are cool, but at least I get stuck every time despite their AI.

But the real kicker is this: The same AI that understands your Data is also the AI that understands your Automations.

Which is really the main strategic argument for why you need Data and Automation under one roof.

This is a self-reinforcing system of intelligence with arms and legs to orchestrate what is happening outside.

One example use case a few customers are working on:

• Extract Pitched Use-Cases from past calls
• Analyse which Use-Cases have the highest win-rate for what ICP
• Take upcoming prospect meetings and scan for lookalike ICPs
• Send Rep a Slack proposing what Use-Cases to Pitch based on lookalike
• Keep improving with new Calls made and deals closed, won/lost

How does this solve the CAC Problem?

It's pretty simple.

Once you can escape the messy data and messy automation conundrum, you suddenly see that you have a System that you can build upon and improve.

You start with the simple stuff: reduce admin for Reps, give them 2-3 hours more per day, increase quota expectation, and potentially reduce the team.

And you do some MEDDIC scoring so your managers can give better, more direct feedback.

This will lead to smaller, highly trained teams with higher win-rates, ACVs, and NRR.

But once you conquer the simple stuff, you realise now what 2nd-order impacts you can achieve:

• A Self-learning system that improves and tailors every pitch for every SDR, AE, and CSM
• A System that detects inconsistencies between what the website is saying, what the reps say, and what the CSMs on board with. And fixes this to smoothen conversions in the funnel
• A System that selects the right accounts for the right reps and gives them the right reachout-hook
• A System that can think and act - right in the middle of your GTM.

As a result, costs will go down, and funnels will get optimised. It will be a Zero-Waste GTM.

And the "saved" Cash will get redistributed in the organisation. Likely short and long-term Pipeline Generation.

But with healthy unit economics.