Sometime in our second year running PipelineRoad, Bruno and I made a decision that seemed small at the time and turned out to be one of the most consequential choices we’ve made. We decided to build software.
Not hire a dev shop to build something. Not white-label someone else’s product. Actually build it — write the specs, design the interfaces, commission the engineering, and ship a product under our own name.
We were a marketing agency. We had no engineering team. We had no product experience. What we had was an observation from dozens of client engagements that crystallized into a conviction: there was a problem nobody was solving well, and we understood it better than anyone because we lived inside it every day.
That conviction was the easy part. Everything that followed was hard.
The Observation
Running marketing for B2B SaaS companies means spending a lot of time inside other companies’ go-to-market machines. You see the tools they use, the processes they follow, the gaps they work around. And after enough engagements, the gaps start to form a pattern.
The specific gap we kept noticing was in how companies managed their pipeline development workflows. The CRMs were fine for tracking deals once they existed, but the upstream work — the research, the targeting, the sequencing, the multi-channel coordination that turns a cold account into a warm lead — was scattered across spreadsheets, disconnected tools, and tribal knowledge.
Every client had their own duct-taped version of this system. Every one of them wished it were better. None of them had found a tool that solved it elegantly.
We saw this pattern across eight clients, three verticals, and two years of engagement. At some point, the question shifted from “why doesn’t someone build this?” to “what if we built this?”
The Identity Crisis
The moment you decide to build a product, you enter an identity crisis that nobody warns you about. You’re not quite an agency anymore. You’re not a product company yet. You’re something in between, and that in-between state affects everything.
Your team doesn’t know which entity they work for. Your clients don’t know if you’re still focused on them. Your market doesn’t know how to categorize you. Your own internal conversations start fragmenting — half the meeting is about client deliverables, the other half is about product roadmap, and neither half gets the attention it deserves.
Bruno and I had different instincts about this. I wanted to go all-in on the product and let the agency become a supporting function. He thought we should keep the agency as the primary business and treat the product as a side project. We were both wrong — the answer turned out to be a more nuanced third path — but the disagreement itself was productive. It forced us to articulate what we were actually building and why.
What we eventually landed on was a model where the agency and the product are symbiotic but operationally separate. The agency work gives us insight into the problem space. The product is our bet on scaling that insight into something bigger. Neither one subsidizes the other. Each needs to stand on its own economics.
Getting to that clarity took months of conversation and a few genuine arguments. The identity crisis doesn’t resolve itself. You have to resolve it deliberately, and you have to keep resolving it as the balance shifts.
Things Nobody Tells You
Here’s what surprised me about building software as a non-software company.
First: the cost isn’t the engineering. The engineering is expensive, yes, but it’s a known expense. The real cost is attention. Every hour you spend thinking about product is an hour you’re not thinking about clients. Every mental cycle spent on feature prioritization is a cycle not spent on campaign strategy. Attention is zero-sum, and building a product while running an agency means you’re running both at a permanent cognitive deficit.
Second: the timeline is longer than you think, and then it’s longer than that. We planned for a six-month development cycle. It took over a year to get to something we were willing to show people. Not because the engineers were slow — they were excellent — but because we kept learning things from our agency work that changed what the product needed to be. The feedback loop between client insight and product design is valuable, but it also means the spec is never truly frozen.
Third: the market doesn’t care about your backstory. We assumed that our agency experience — the fact that we’d seen this problem up close at dozens of companies — would be a compelling differentiator. It is, to a point. But when a prospect is evaluating your product, they care about whether it works, not about how you came to understand the problem. Your origin story is good for a blog post. It’s not a substitute for a product that delivers results.
Fourth: fundraising as a hybrid is awkward. Investors like clean stories. “We’re an agency” is clean. “We’re a SaaS company” is clean. “We’re an agency that’s also building a SaaS product and neither one is big enough to stand alone yet” is a messy story that makes investors nervous. We learned to lead with the product vision and mention the agency as a strategic asset rather than the other way around.
What the Agency Gives the Product
Despite the challenges, there’s a genuine structural advantage to building software from inside an agency, and it’s one that pure-play SaaS companies don’t have.
We have paying customers before we have a product. Not for the software — for the agency services that the software will eventually automate or enhance. That means we understand our users at a depth that no amount of customer discovery interviews can replicate. We’ve done their job. We know what’s hard, what’s tedious, what’s broken in the existing workflow, because we’ve experienced it firsthand, repeatedly, across multiple companies.
We also have a built-in distribution channel. Every client engagement is an opportunity to introduce the product. Every prospect call is a chance to demonstrate the thinking behind it. The agency is the product’s first growth engine, and it costs us nothing in customer acquisition.
And we have revenue while we build. The agency pays the bills. The product doesn’t need to generate revenue immediately, which means we can take the time to get it right rather than shipping something half-finished because we’re running out of runway.
The Long Game
Building software when you’re not a software company is an act of faith in your own understanding of a problem. You’re betting that what you’ve observed across dozens of engagements is a real pattern, not a coincidence. You’re betting that the solution you’ve imagined is better than what exists. You’re betting that you can learn the entirely new discipline of product development while maintaining the existing discipline of client services.
Some days the bet feels genius. Some days it feels insane. Most days it feels like a lot of work.
But I keep coming back to the original observation — that pattern we saw across every engagement, the problem nobody was solving well. That observation hasn’t changed. If anything, it’s gotten stronger as we’ve worked with more clients and seen the same gap in more contexts.
The identity crisis is real and ongoing. We’re not fully an agency and not fully a product company. But maybe that’s not a crisis at all. Maybe it’s just what it looks like when a company evolves into something that didn’t have a category when it started. We’ll find out.