E34: Software Enabled Agencies with Taylor Holiday
In this episode of the SaaS Operators Podcast, we talk to Taylor Holiday, the Founder & CEO of Common Thread Collective, about the line between ecom SaaS and marketing agency services. Taylor explains how CTC evolved from an agency into a software enabled service business, why he doesn't call it a software company, and how a single growth map spreadsheet became their operating system. We talk about the differences in gross margin, NRR, and risk of ruin. Why agencies can actually have better unit economics than many SaaS companies, and why PE quietly loves marketing services. The episode closes on packaging and positioning, when to sell what you do as software versus service, and the very real arbitrage for operators building “Seans”, Software Enabled Agencies.
We opened this episode of the SaaS Operators Podcast with Taylor Holiday, CEO of Common Thread Collective, talking about the question everyone in our corner of the world circles around.
Is it better to build software or services?
The more we pulled on it, the more that question started to fall apart. What most people call “SaaS versus agency” looked a lot more like storytelling, accounting choices, and risk preferences than two cleanly different business models.
This conversation started from a story about how CTC turned a giant spreadsheet into their operating system. It ended with a pretty strong case that agencies might actually be the best business model in ecommerce.
Here is what I took away.
CTC Is Not A Software Company
At CTC, the whole “software enabled” thing started from a single internal artifact they call the growth map.
They began as a pure service business. Twelve years in, for the first seven they were not “tech enabled” at all. Just humans doing work.
Then they ran into the obvious scaling problem. At the edges of delivery, everything broke. Each account manager had their own way of doing things. Different sheets. Different logic. Different views of the truth.
To fix that, they did not go “we need to be a SaaS company.” They did something more boring and more powerful.
- Aggregate all the data into a single database.
- Build workflow on top of it.
- Use sheets as the interface because nobody wants to pay for more software and everyone is comfortable in spreadsheets.
They literally built a retrieval pipe into Google Sheets to avoid paying for Supermetrics. They took an old Lightspeed LTV template, adapted it, and wired their own data into it.
Over time, that sheet became software. It got formalized. But the important part is the frame:
CTC is a service company that happens to use software. Not a software company.
That distinction matters for everything that follows.
Gross Margin And NRR Are Mostly Storytelling
We asked Taylor about the “software enabled agency” idea and whether, in the extreme, a software enabled agency and a SaaS company converge on the same financial profile.
His answer was basically: they already have.
Most of the distinction is an accounting game. In SaaS, you call people “customer success” and stick them in sales and marketing instead of cost of goods. In an agency, those same people are your delivery team and sit in gross margin.
The work is the same. The label is different.
Taylor thinks:
- NRR and LTV to CAC can look just as good, or better, in a great service business as in a software business.
- The real constraint is not margin, it is velocity. How fast can you onboard new customers.
That is where the models diverge.
A lot of software companies literally cannot serve small accounts. Taylor is working on a deal with a big player in his space that has “hundreds of closed lost deals” tagged as “too small to support.” They want to send those customers to CTC as a preferred partner.
Which is wild.
You have a high margin software product, but you cannot profitably onboard small customers because the human support cost is too high. That is not the Slack model where one person or ten million can sign up and use it with almost no marginal cost.
So when you look at the financials, the difference between “software” and “software enabled services” starts to blur.
There Are Only Two Ways To Grow Fast Right Now
Rishabh made a simple observation that I think is worth writing in big letters.
There are basically two types of companies that grow insanely fast today:
- Product led growth, self serve, almost no humans.
- Forward deployed armies of people paid for by lots of capital.
Think Gamma, Lovable, that style of PLG. Pure self serve. No CSM org propping it up.
Then think Sierra, Decagon and similar “AI plus services” shops. They raise a ton of money and throw smart people at enterprise accounts.
The second bucket is way more common than the first. Because it is genuinely hard to build a product that is good enough to scale itself.
Most “SaaS” companies are much closer to “software wrapped around a small consulting firm” than people admit publicly.
Agencies Are Quietly Incredible Businesses
There is a running joke that we are all in ecom because we were not smart enough to be in finance.
Taylor leans into that. He also pushes back on the idea that agencies are this low tier business model with no upside.
He walked through his peer set. Names everyone in our world knows: Power Digital, Wpromote, Tenuity, MuteSix, Metric Digital. The shops CTC competed against for years.
Every single one of them has sold for hundreds of millions. Many of them have gone through multiple PE flips, moving up the chain from one fund to the next.
Why?
Because marketing services have:
- Tight cash flow to EBITDA.
- Contracts that banks like to lend against.
- A very clear acquirer universe, from mid sized PE all the way to the big holding companies.
Compare that with DTC brands. For a random niche brand, it is not obvious who the natural buyer is. For agencies, it is extremely obvious.
The problem is not that the outcomes do not exist. It is that they are under reported. Culturally we fetishize the SaaS exit and ignore the agency rollup.
Talent follows the story. Big upside narratives pulled the sharpest people into software. Service businesses got the “not cool” label.
Taylor’s view is almost the opposite: if you are smart and you want edge, go where the smart people are not.
Risk Of Ruin: Service Wins
One frame Taylor uses that I like a lot is risk of ruin.
Instead of starting with “what has the highest theoretical upside,” start with “what is least likely to kill me.”
- Ecom has brutal risk. You commit capital into inventory based on a forecast. If you are wrong, you are stuck with product nobody wants, no cash, and payables you still owe.
- Venture backed SaaS has a different risk profile. Your cap table often forces you into a high risk, high growth strategy whether you like it or not. Your investors’ business model gets baked into your own.
- A service business has almost no fixed commitments. If everything fell apart tomorrow, Taylor could fire everyone, keep one client, and still make money servicing them himself.
His committed costs are tiny. Cash conversion is fast. He can dial capacity up and down much more safely than someone sitting on a warehouse of dead stock.
If you rank models by “likelihood of dying,” his argument is that services are clearly the best business.
The market does not talk about that, because it is less sexy than “100x upside,” but it matters in the real world.
Packaging Is Just Marketing
We also got into the question of “what do you call yourself.”
If you want to raise money from Silicon Valley, you do not walk in and say “we are a services company.”
You talk about software, platforms, data, all of it.
If you want to sell to an ecommerce brand, the opposite can be true. Calling it “software” can actually make it harder to charge what you need to charge. It is easier to package a high ticket engagement as services.
So the label is really a marketing decision based on who you are selling to and what endgame you are optimising for.
Taylor also called out the “month to month nobility” a lot of operators, including Jeremiah, fall into. This idea that being cancellable at any time shows how confident you are.
Retention might be strong. Customers might stay for years. But from a bank or PE perspective, “they can leave whenever they want, but trust me, they will not” is not as valuable as a contract that says they will pay for twelve months.
If you care about valuation, leverage and acquisitions, contracts matter more than your personal sense of honor.
People As The Real Product
One thing I have seen in agency land is how much the experience hinges on the person on the call.
I mentioned a big agency where the team that showed up to a prospect call seemed scared to be there. Low energy, low conviction. You can burn millions of dollars in pipeline with a few of those calls.
Taylor is painfully aware of that dynamic and tries to design against it.
At CTC, they push hard on a “brick in the wall” culture. You are not there to build secret spreadsheets that only you understand. You are there to contribute permanent pieces to the system.
A feature. A process. A framework in their internal tool. Something you can point to when you leave and say “I built that.”
That does a few things:
- It reduces the dependency on any one person.
- It turns random Google Drive chaos into shared assets.
- It makes the service more like a product over time.
They also lean heavily on spreadsheets as an MVP sandbox. They have “data chefs” who can make Sheets feel like software. Every product starts there before it gets promoted into the actual app. Customers get value right away while the engineering team hardens it.
Interestingly, it worked almost too well. Their spreadsheet layer became so powerful that it delayed the need for formal software. When your sheets are that good, the urgency to “build a real product” goes down.
But it is still the same idea. Capture human problem solving in a reusable system instead of letting it evaporate in private docs.
The Labor Arbitrage Window
Toward the end, Taylor told a story that sums up where we are right now.
A friend of his went from agency world into VC land and then messaged him later saying:
“Labor is mispriced. I can charge hundreds of thousands of dollars for something I could build a tool for, and the market will pay it. If I offered the same thing as software, the price would be anchored way lower.”
That is the arbitrage.
Buyers still think of humans as “expensive but worth it” and software as “cheap and commoditized.” You can deliver the same underlying outcome with a blended model and make a lot more money today if you position it as service.
Taylor does not think that gap stays forever. Over time, software pricing compresses, labor gets repriced, younger operators care less about talking to humans, and it all converges.
But right now, there is a lag. Agencies and software enabled services sit in a mispriced corner of the world.
So What Do You Do With This?
If you are early in your career, or you are deciding what to build next, a few things feel clear to me after talking with Taylor:
- “Service vs SaaS” as a label is mostly branding and accounting. The real questions are: what outcome do you deliver, what does it cost to deliver it, and how fast can you scale that delivery.
- Services have lower risk of ruin and more flexible economics than people think.
- There is real, proven acquisition upside in agencies, even if it is not talked about as loudly as software exits.
- Product led growth at true scale is rare. Most fast growing “SaaS” companies are human heavy behind the scenes.
- There is a near term arbitrage in packaging high value work as service instead of software.
If you are building in this space, you do not need to cosplay as a pure SaaS company to build a meaningful asset. You might be better off embracing the hybrid reality, pricing for the actual outcome, and obsessing over risk of ruin instead of Twitter optics.
That is what I am taking from this one.
