DemandLoops 2025 Annual Letter
A letter written from the trenches of solo-to-intentionally-small-shop
This week I was attempting to hit inbox zero and came across a letter from Immad Akhund, CEO at Mercury. He later ended up sharing to his LinkedIn community, too.
But I read the whole thing. Twice. The first time through I was reading as an avid member of their community. I’m very bought into the product they’re building and the problems they’re solving for founders. But the second time through I was reading with my own “founder” hat on. I love the vulnerability, the transparency, the rhythm of doing something like this annually. So…I thought welp, why the heck not. And now I’ve spent half my day writing my own version of an annual letter for DemandLoops.
I’m hoping this serves a few purposes:
To anyone wanting to leave W-2 life: I’m not a “founder” or CEO or inventor. I’m just an operator who got good at one thing (demand gen) and built a business around it. If I can do this, you can too.
To anyone trying to up-level in their current role: The gaps we’re solving across clients aren’t unique. Pattern matching, org-wide visibility, managing up - these skills matter everywhere, not just in consulting.
To anyone we might work with: This is how we think about problems and why we operate the way we do.
So, here it goes.
Dear DemandLoops community, readers, friends,
I started DemandLoops mid 2023 with a belief that felt obvious but kept getting ignored.
I believe growth is not linear.
Careers aren’t. Companies aren’t. Demand gen definitely isn’t.
But most of the industry treats all three like straight lines:
More tools → better outcomes
More leads → more revenue
More effort → more progress
But I believed back then, and still today, that progress actually comes from compounding feedback loops:
What you measure changes behavior
Behavior shapes systems
Systems reinforce outcomes
Outcomes then justify the original measurement
Companies then, and now, keep hiring new talent, or new agencies, to “fix” demand gen. They get new campaigns. Fresh creative. Maybe better automation. Six months later, nothing fundamental has changed.
Sales still complains about lead quality. Attribution still makes no sense. Pipeline forecasts are still vibes and spreadsheets.
What I kept seeing was this: AI was scaling bad data. Sophisticated tools were amplifying broken processes. Short‑term pipeline pressure made it impossible to build anything sustainable.
So I built DemandLoops to be the opposite. Originally DemandLoops was purely a solo venture. A chance for me to break the cultural norms of having to climb the corporate ladder in order to be successful, time to focus on client projects that kept my interest, and a position where clients paid for an honest perspective on their growth opportunities.
What We Keep Finding In Client Systems
Account selection is still the Wild West
Most companies confuse their TAM with their ICP. And worse, they confuse their ICP with whoever came inbound last quarter.
I wrote about this in October because I kept seeing the same pattern across engagements. Marketing would proudly show me their “ICP definition” - usually a neat little chart with company size, industry, tech stack. Then I’d ask sales what their best deals looked like. Completely different answer.
One client came to us saying they couldn’t get inbound to work. When we dug in, they were targeting their entire TAM (tens of thousands of companies) in their paid programs. No prioritization. No tiering. Just “spray and hope someone converts.”
The fix isn’t complicated, but it requires admitting something uncomfortable: you can’t go after everyone, even if they technically could buy from you.
Here’s what we’ve seen work:
Start with your best customers (highest LTV, shortest sales cycle, actual success with your product)
Document the firmographics, technographics, demographics, AND behavioral signals they share
Build account tiers based on fit, not just size
Accept that some accounts in your TAM shouldn’t get the same attention as others
The hardest part isn’t the analysis, but getting leadership to agree that tier 3 accounts don’t deserve the same resource investment as tier 1. That conversation is where most companies get stuck.
The measurement problem is always an alignment problem
This one showed up in literally every engagement in 2025.
Sales and marketing weren’t fighting about lead quality, but instead are really fighting because they were measuring success using different definitions.
Marketing measured MQLs, form fills, campaign influence. Sales measured closed-won deals and pipeline created. Neither team’s metrics connect to the other’s.
One CMO told me: “We hit 120% of our MQL goal last quarter and sales still said we weren’t performing.” When we mapped it out, turns out those MQLs had a 3% conversion rate to closed-won. Marketing was generating volume. Sales needed velocity.
The pattern I keep seeing: companies layer on more sophisticated attribution tools thinking that will solve alignment. It doesn’t. Because both teams are still optimizing for different outcomes.
What changed things for clients:
Getting sales and marketing leadership in the same room to define entrance/exit criteria
Building metrics both teams believed in (usually around qualified pipeline/revenue, not MQLs generated)
Creating visibility into what happens after the handoff, not just before
Admitting that some of their “best performing” campaigns were only best at generating activity, not revenue
The breakthrough moment usually happens when we show them their top 10 campaigns by MQL volume next to their top 10 campaigns by closed-won revenue. Almost never the same list.
AI is exposing every process gap you’ve been ignoring
This is the one that’s evolving fastest and causing the most chaos.
Every client conversation in Q4 included some version of: “We’re implementing AI tools but not seeing the efficiency gains we expected.”
Well, AI is really good at doing what you tell it to do. If your process is broken, AI just executes that broken process faster.
So things like:
Clean CRM data (accounts properly structured, duplicates merged, fields actually used)
Clear signal definitions (what counts as intent, what’s just noise)
Documented workflows (so AI can follow a process that works)
Measurement systems that connect activity to outcomes
…are becoming more critical than ever.
The AI stuff everyone’s excited about - signal orchestration, predictive scoring, automated personalization - only works if the foundation is solid.
The Shape of What We’re Building
The clients we said no to (and why that matters)
27 times in 2025, we turned down potential clients. Clients that were ICP fit, but either they weren’t ready for the work, or we didn’t have capacity.
On the work front, the pattern was always the same. Leadership wanted better pipeline numbers. They thought the solution was new campaigns, better tools, maybe some AI magic. They wanted tactical fixes to organizational problems.
One VP of Marketing called in March. Their CEO wanted to “fix demand gen in 90 days.” Had this hard out he was adamant about including in the contract. When I asked about their current attribution setup, said “we track everything in Salesforce.” When I asked how sales and marketing defined a qualified lead, said “that’s actually been a point of tension.”
That’s code for: nobody agrees on what success looks like, but leadership wants it fixed by Q2.
I could have taken that engagement. Delivered some new campaigns. Generated some activity. Collected the check. But six months later, they’d still have the same tension, just with new dashboards.
The clients we said yes to had something different. Not bigger budgets. Not simpler problems. They had leadership who was willing to hear that their measurement was wrong, their ICP needed work, or their tech stack was amplifying their dysfunction.
It’s some uncomfortable stuff to admit. But you can’t fix what you won’t acknowledge is broken.
The red flags we learned to spot:
Leadership using phrases like “quick win” or “low hanging fruit” in the discovery call
Sales and marketing leaders not both involved from day one
Timeline pressure that doesn’t allow for learning (everything needs to be “done by end of quarter”)
Unwillingness to change measurement, only tactics
Every no felt so expensive in the moment. But every yes to the wrong client would have meant less capacity for the companies ready to do this work differently.
An attempt at productization
Q1 2025, I decided to productize some of my processes.
The thinking made sense: clients kept asking for the same initial audit. I’d built a mostly-repeatable framework. Why not turn it into a self-serve offering? Scale the business without scaling my hours.
I spent a few several nights and weekends thinking through and documenting how I wanted to build it out. What the experience might look like. How I’d market it. It literally never left my drafts. That was a year ago now.
In that moment I was still working solo, trying to keep too many irons in the fire. The irony was that I needed to take some of my own advice and focus. Do less, better.
The hire that changed everything
I’d become this accidental cheerleader for going solo. Built a big part of my online presence around it. So the idea of hiring felt like... I don’t know, admitting I couldn’t hack it? Or something equally stupid.
I’ll never forget it. I was on the front porch, multi-tasking while the kids played, and within 30 minutes, I said “I think this might be too good to be true.”
And just like that, took the leap from solo to company of two.
I’d resisted hiring for almost two years. Solo was simple. No management overhead. No risk of a bad culture fit.
But solo also meant turning down good clients because I was maxed out. It meant every vacation required either overworking before I left or disappointing clients while I was gone. It meant my business could only grow as much as I personally could scale. And it meant making every decision alone.
The hiring process was simpler than expected. I found the perfect hire in my network. Someone who could think strategically and also execute. Who understood the technical side of MarTech and could talk to executives about business outcomes. Who was comfortable with ambiguity because I was still figuring things out.
The part I didn’t anticipate: the founder-sales problem. People got used to seeing me in their feeds. Reading this newsletter. Had heard me on podcasts, etc. When they realized I wouldn’t be on every client account, they got nervous. Early contracts had addendums guaranteeing my involvement in weekly calls, strategy sessions, anything to give them that safety net.
That lasted about three months.
By December, something shifted. Clients realized she wasn’t just qualified - she was bringing perspective and skills I couldn’t. The quality of our work improved because I wasn’t context switching between four client calls a day. And I could actually think about where this business should go instead of just surviving the current quarter.
Looking back, I should have hired sooner. But I also don’t think I could have hired well until I felt that constraint. Sometimes you need to hit the ceiling before you’re ready to raise it.
The Weird Choices We Made About Growth
Here’s what we decided about how DemandLoops would grow:
Bootstrapped, profitable
Work‑life integration, not balance
Learning in public, including the failures
Intentionally small team, high leverage
Those choices absolutely capped certain growth paths.
They also attracted clients who were ready to fix root problems instead of leverage RAM (random acts of marketing).
What We’re Building Toward Next
In 2026, outside of maintaining our standard for client delivery, we’re focusing on three things.
Making the positioning switch official
We’re fully committed to building a small, intentional team. Not scaling to 20 people. Not becoming a traditional agency. Just a tight group of operators who can solve hard problems for clients ready to do the work.
That means updating the website. Rewriting messaging. Finally announcing who joined the team and what she brings to clients. All the positioning work I’ve been quietly putting off because I wasn’t sure how to talk about it yet.
Now I am.
Hiring again (and we already found them)
This year we’re making our second hire so we can say yes to more clients. Still selective. Still high-touch. Just not so limited by capacity anymore.
The best news? We’ve already signed the offer for the perfect candidate. 🤗
Building more candidly in public
With all the AI slop crowding my feed, I’ve been thinking about what got me here in the first place.
I got my start sharing pretty candidly on LinkedIn. The good, the bad, the ugly. From there came the Demand Gen Chat podcast at Chili Piper. That format of connecting with the community, learning from others, sharing wins, elevating voices so people could be heard above all the noise - that’s how I learned and grew in this craft.
I really haven’t been on camera much since leaving Chili Piper. Definitely outside my natural comfort zone. But I built my platform through writing and podcasting because it helped me stay connected to this work and this community.
Getting back into that rhythm is my goal for 2026.
So expect some video content this year. Probably a small series on LinkedIn. Maybe some uncomfortable first attempts. But that’s kind of the point.
This probably doesn’t read like a traditional annual letter. But traditional isn’t really working in demand gen right now.
If you’re dealing with any of the patterns I mentioned - misaligned ICPs, broken measurement, AI amplifying dysfunction - I’d genuinely love to hear about it. Hit reply. I read everything.
Here’s to growth,
Kaylee ✌


Love this!
You read mercury's letter twice, and I read yours twice over. Look at that. Super excited to see more content from you this year! I've been in the demand gen seat for a little over a year, and it feels like a never-ending fire hose of things to learn. Your newsletter has been so helpful to me. Thank you for being so generous and transparent with your learnings!