A short piece on where I write cheques and, more importantly, where I deliberately don't. The honest version of any operator angel's investment thesis is a function of where their experience can plausibly help — and an admission of the spaces where it can't.
For me, the strongest fit is marketing-tech, agency-adjacent SaaS, and the broader infrastructure of how digital businesses go to market. There's a specific reason for each.
What two decades in agencies gives me
I've been operating digital agencies since 2007. That's a long enough run that I've watched several full cycles of how marketing teams buy software, what they actually use, what they abandon, and what they pay real money for vs what they reluctantly tolerate.
A few patterns that experience teaches:
Marketing teams adopt tools the way they adopt agencies — slowly, on word-of-mouth, and with a strong preference for tools that don't require them to change behaviour. The product matters, but the way a tool gets adopted matters at least as much. Founders building marketing tech who've never sold to or worked alongside a marketing team consistently underestimate the friction of adoption.
Agencies are the world's best beta testers and the world's worst paying customers. They'll trial anything. They'll give detailed feedback. They'll tell you exactly what's wrong with the product. They will also negotiate the price down to the point where the agency tier of the customer base subsidises everyone else. Founders building tools that target agencies have to understand this dynamic.
The unglamorous parts of marketing operations are where the durable software businesses get built. Reporting, attribution, workflow, data plumbing, integration. Not creative tools. Not "AI for marketers" sold on the AI angle. The boring infrastructure outlasts the categories built on hype.
This is operator pattern matching that I genuinely think helps founders. It's the value I can add, and it's why I lean into companies in this space.
What I deliberately don't invest in
The corollary of having a clear thesis is having clear non-thesis areas.
Consumer. I have no operator credibility in consumer products. I don't know the failure modes. I don't know the talent pool. I can't help with the channels. A consumer founder taking my money is taking the cheque without the operator wrap, which is fine, but they should know that's the deal.
Deep tech, biotech, hardware. Same logic. I don't know enough about the underlying domains to add value beyond capital. Better to leave the cap table slot for someone who does (more on what an operator angel actually adds beyond the cheque).
Late-stage growth. I write small cheques. The dynamics of late-stage growth — board governance, executive hiring at scale, secondary markets, IPO prep — aren't where my experience is. Founders raising series B and beyond are better served by professional growth investors who do this for a living.
Companies in direct competition with Marketing Signals or its main service categories. Conflicted by definition. The founders are usually surprised to hear this on a first call but they understand it once it's named.
The honest non-thesis is, I think, more useful than the polished thesis. It tells founders specifically when I'm worth talking to and when I'm not.
What this means for the kinds of deals I see
The thesis filters the inbound. Most of what comes in is in the right zone — marketing tech, ad tech, content tech, agency operations, vertical SaaS that touches the marketing function. Within that, I'm looking for a few specific things.
Founders who've operated on the customer side of the product they're building. Marketing tech founded by people who've never been in marketing tends to build for problems marketers don't have. Marketing tech founded by people who've been in marketing for a decade tends to build for problems marketers desperately want fixed. The signal is clear in the first five minutes of a call.
Distribution insights specific to marketing teams. How do marketers find tools? Through peers, through case studies, through the agencies they work with, through specific conferences, through community. The founders who've thought hard about which of these channels they're going to own are dramatically more likely to make it work than founders who haven't.
Pricing models that respect the customer's actual budget shape. Marketing budgets are weirdly seasonal, weirdly siloed, and weirdly opaque. Pricing models that don't account for this — annual contracts that ignore the reality that the marketing director's authority maxes out at a particular quarterly threshold, for instance — kill deals that should otherwise close.
These are domain-specific signals. They're the things I look for because I have the context to look for them. An angel investing across categories without specific domain context wouldn't know to weight any of them.
Where the thesis is currently most active
Two areas where I'm writing more cheques right now.
GEO/AI-search-visibility infrastructure. This is the specific area where Marketing Signals has been investing internally — amivisible.co is our own product in this space. The category is new enough that the early companies have a structural advantage, and old enough that the customer demand is starting to show up clearly. I'm interested in companies building anywhere in the layer between brand content and the AI engines.
Workflow automation for marketing operations. Agentic AI applied to specific marketing workflows — campaign QA, attribution reconciliation, brief generation, reporting automation. The teams that win here are the ones building for the operational layer of marketing, not the creative layer.
Vertical SaaS that's marketing-adjacent. Categories like e-commerce operations, fintech with a strong consumer marketing component, marketplaces where customer acquisition is the central problem. Areas where my agency experience overlaps with the founder's domain.
The mix shifts over time. Two years ago I was writing more into general marketing tech. Now the GEO and workflow-automation slices are more weighted because that's where the most interesting companies are emerging.
Closing
The reason for being public about a non-generic thesis is that it saves everyone time. Founders who match the thesis get a faster first call. Founders who don't, can find a better-matched investor without burning two weeks pinging me.
If you're building in the space — particularly if you've operated on the customer side of marketing tech before founding the company — I'd like to see the deck. Email is at the top of the site. The same applies in reverse: if you're outside the thesis, no reflection on the company, just not the right angel for it.
The more specific the thesis, the more useful the cheque. That's the thinking behind this one.