The Missing Piece Inside HubSpot's Loop Marketing Model

Episode Transcript

Imagine you're the company that basically invented content marketing. And then, in about 10 months, you lose 80% of your organic search traffic. Sadly, this is not a hypothetical situation. It's what actually happened to HubSpot from 2024 into 2025. Their CEO called it an apocalypse. And here's the part that should make every marketing leader sit up and take notice: while this was happening, their revenue went up nearly 20%.

So what did they do? They did not try to figure out how to write a better blog post. They totally rebuilt how their marketing works — and they gave it a new name: Loop Marketing. They say it has four stages: Express, Tailor, Amplify, Evolve. It's a system that is supposed to get smarter every time it runs. And I love it.

I genuinely do. But I want to talk about the thing that decides whether your loop actually compounds or just spins. Because ultimately, a loop is only as fast as its slowest cycle. And on most marketing teams, the slowest cycle is no longer how long it takes to make stuff. It's the plan, the pile of work in progress, and how often the team pauses to fix how the whole thing works. That is what we're talking about today: Loop, the vehicle — and Agile, what makes it move or not.

---

Welcome to The Agile Marketing Edge, the first podcast dedicated to turning Agile theory into real-world marketing breakthroughs. I'm Andrea Fryrear, CEO of AgileSherpas and your guide on this climb to smarter, faster, outcome-driven marketing. Every week we dig into the what, who, and how behind Agile marketing — from building high-velocity workflows and cutting waste, to measuring what actually matters and scaling success across teams.

Today we're putting two frameworks side by side: HubSpot's brand-new Loop and the Agile Marketing Operating System that a lot of you are probably already using. They were built by different people for very different reasons, but they land in a surprisingly similar place.

---

We need to take a second to get the framework straight first, because it's newer than most people realize. The book just came out in September 2026, so it's very fresh. It has four stages.

**Express** is where you define who you are and who you serve. HubSpot calls this a taste profile — it's your brand's point of view plus a rich picture of how your customer thinks. **Tailor** is where you turn that into personalized content at scale, made possible by AI doing a lot of the production. **Amplify** is the distribution stage: answer engine optimization, generative engine optimization, ads, creator collaborations — wherever your buyers are, getting your personalized content into their sphere. And **Evolve** is where you measure what worked and feed it back into Express, so the next turn of the loop starts from a smarter place than the last one.

A quick note, because I've already seen people getting this mixed up: if you're thinking "didn't HubSpot already kill the funnel with a flywheel — attract, engage, delight?" — short answer is yes. That was the 2018 model, but Loop is different. The flywheel describes momentum, whereas Loop is about learning. Their own shorthand is basically: the funnel still processes, but the loop learns.

You're probably hearing a lot of parallels between this and what I talk about on this podcast. So it's probably obvious why I'm interested in it. But the part that's really getting my attention is when you get to the section of the book about how to actually run this — the operating model. HubSpot starts describing small cross-functional pods, lifted pretty much straight from software teams. They describe reviews every two weeks, monthly KPI reviews, quarterly retrospectives, and a weekly decision meeting that includes a pre-read. A lot of Agile echoes throughout.

But in the whole book, as far as I've been able to find, the word "agile" does not appear once. And I'm not saying that to score a point. I'm saying it because it is one of the most important pieces of evidence in this whole episode. One of the biggest marketing companies on the planet went looking for the best way to operate in the AI era. They started from a traffic collapse, not from an Agile training, and they arrived at what are essentially sprints, retros, and cross-functional teams. This is not a coincidence. This is a convergence. When two groups start from totally different places and land on the same inherent approach, it's usually because the answer is solving a real constraint that a lot of people are grappling with.

So here's how I want to overlay the two. Loop's four stages live mostly in the What circle of the Agile Marketing Operating System — strategy, content production, distribution, metrics and analysis, what work gets chosen and how it's measured. Super real, super valuable. Loop is doing a great job there. But a loop doesn't just spin because you drew it like a circle — or in this case, like an infinity symbol. It turns and things move through it because of how the work moves and who owns it.

Loop can be like a cool new app on your phone, but you can't install it on top of your old operating system and just expect it to start humming along perfectly. Maybe it'll let you install it and maybe it kind of works at first, but pretty soon you're going to see performance degrade. So this is the "yes, but" part of this episode.

When you rediscover a system on your own — like HubSpot did — you tend to rediscover the pieces you needed most urgently, not always the whole thing. And the pieces being skipped are probably the ones that will come back and hurt you later. For the rest of this episode, I want to look at three places where the loop can stall out — not because it's wrong or bad or broken in any way, but because these are the exact spots where Agile teams already have the operating system designed to keep things spinning, and where everybody else who tries Loop without Agile behind it is about to discover they've got some painful gaps.

---

**Important pause before we move on.** If your team wants to adopt Loop — or anything in that same family — ask one question at your next leadership meeting: which of our current meetings and habits would actually make this cycle turn faster? If you can't think of one, you are about to install a new app on a really old operating system. It's not going to work nearly as well as you think.

---

So let's talk about learning, because it's a big part of both Loop and Agile. HubSpot makes a claim in the book that I completely agree with: a team running 26 learning cycles a year is going to beat a team learning four times a year, even if the slower team has better people and a bigger budget. That's compound interest applied to learning, and it's a big part of the argument for using sprints in marketing.

But we need to pause and ask: what actually counts as a completed cycle? Because it's not "we looked at the dashboard." A cycle closes when what you learned changes what you do next. That's why in the book, Evolve has to go back into Express and change a priority. Something has to be different on the next run of the loop. Otherwise, you didn't run a loop — you just wrote a report about one loop that you ran.

This is where your plan comes in, because your plan needs to be the gate that every insight has to pass through. In this year's State of Agile Marketing Report, we asked how often teams update or adjust their plans. 89% of Agile marketers do it at least monthly — probably more often if they're on two-week sprints. Non-Agile marketers: 66%. And 66% might sound fine — so let's look at it from the other direction. A full third of non-Agile marketers, 34%, adjust their plans quarterly or less. That's three times the rate of Agile teams, where it's 11%.

Think about what that means for a loop. If your plan only changes once a quarter and you can learn something every week, nothing is going to land until the next planning cycle. No matter how fast the loop itself is going, the effective speed is still four times a year. It doesn't matter how fast AI is churning out content or how fast you're collecting data. If you're still only planning quarterly, you built a race car and then parked it behind a toll booth.

The second piece of this is fuel, because the Express stage runs on customer understanding. HubSpot's whole taste profile idea depends on knowing your customers' language, triggers, and tolerances. And in our State of Agile Marketing Report, 87% of Agile marketers say they often use customer insights or data to guide what they work on, versus 72% of non-Agile marketers. And 6% of non-Agile teams say they rarely or never use customer insights, compared to just 1% of Agile teams. That's the difference between a taste profile that's alive and resonant and one that's just a Word document nobody is ever going to open again.

HubSpot actually warns about this exact failure, which I appreciate. As Teresa Torres would put it, discovery is not a phase — it is a habit with a cadence. If customer insight only shows up at your annual persona refresh, then your loop has nothing new to learn from.

And here's the perennial objection to Agile in marketing: "If we change our plan all the time, won't it just be chaos?" No. And there's data to back that up. We asked how confident teams are that they'll stay organized when priorities change — and they will, you can't avoid it. 82% of Agile marketers said they were extremely or very confident, versus only 58% of non-Agile marketers. And importantly, not one single Agile marketer — 0% — said they were hardly or not at all confident.

Why? Because frequent replanning is not the same as constant pivoting. Agile teams change the plan at a known rhythm — monthly or every sprint. In between, the plan holds, and that's what makes change feel survivable. The chaos comes from teams that never formally replan, so every new insight or idea becomes a catastrophic interruption.

**What to do about this:** measure your insight-to-plan latency. Next time someone on your team learns something real — a new test result, a new customer quote, data about shifting channel behavior — write down the date. Then write down the date when your priorities actually changed because of it. That gap, not your content production velocity, is your real loop speed.

---

Now let's talk about making stuff — and how that is actually no longer really a competitive advantage.

AI made it really cheap to start things. It did not make it cheaper to finish them. One of the most exciting ideas in Loop is the concept of a full-loop marketer: one person owns an idea from first thought to last insight. On Monday they have an idea, they test it against an AI simulation of the buyer. By Wednesday it's live. Friday, performance data. Next Monday, building version two. It's a great representation of what it means to be full-stack in 2026.

And on a team doing one or two of those things at a time, it can work. But here's what I've seen happen on real teams. AI drops the cost of starting something to basically zero — so people start more things. A lot more things. But the cost of finishing — incorporating the review, the decision that requires five people to weigh in, the integration into tools and workflows and data collection — that does not get cheaper nearly as fast. So what happens? The work in progress piles up. Hello, bottleneck.

In our State of Agile Marketing Report, we asked current Agile users what's getting in the way of working in a more Agile way. The number one barrier, named by 47% of teams, was too much work in progress. But when we segment this and ask that question of teams with an extremely successful Agile implementation, the number who name work in progress as their primary barrier drops to just 33%. That 14-point gap is one of the biggest in the report.

And there is a mechanism behind why this happens. It has a name: Little's Law. Daniel Vacanti has written about this beautifully for knowledge work. It says your average cycle time equals your work in progress divided by your throughput. Let me repeat that for those of you listening and not quite able to visualize equations: your average cycle time equals your work in progress divided by your throughput.

Here's why that equation matters. If your throughput stays the same — which it probably will if you don't make a fundamental adjustment to your process — and you double what's in flight because AI makes it easy to start things, everything will take twice as long. Apply that to a loop. If your full-loop marketer is running six experiments at once instead of two because they can now, each loop takes roughly three times as long to close. Their 26 cycles a year quietly becomes just eight or nine. And nobody will notice because everyone looks incredibly busy.

Here's the irony: HubSpot opens that chapter of the book with a story about Airbnb's design team being like a deli with a line around the block — a three-month wait for a button. That, my friends, is a work-in-progress problem. Removing the handoffs by turning people into full-loop marketers fixes part of it. But if you give every marketer their own deli counter with no limit on the orders they can take, you're just going to get very long lines and still very unhappy people who aren't getting their buttons — or their sandwiches.

This is the Who part of the Agile Marketing Operating System. When one person owns the full loop, somebody still has to decide what that person does not start. In Agile, that's the job of a WIP limit and a prioritized backlog. It's not bureaucracy. It's the thing protecting your cycle time.

**What to do about this:** before you start handing people the whole loop to own, set a cap. One or two active experiments. A new one can start only when an old one completely closes — meaning the learning has been written down and fed back into the plan.

---

New question: this loop is supposed to learn. Does your team learn?

This might be the one I care about most. Chris Argyris — an organizational learning researcher — drew a distinction between single-loop and double-loop learning. Single-loop is like a thermostat: it's cold, so you turn on the heat. You adjust the action. Double-loop learning pauses to ask whether 68 degrees was the right setting in the first place, or whether a thermostat is even the right tool. This is where we adjust the system that selects the action — not just take the action.

The Evolve stage of Loop is excellent single-loop learning. We find out that short-form video beats long-form for a particular segment — important learning. We update the taste profile, we make more short-form content. The content and amplification get smarter. But who is asking whether the team itself is working as well as it could while producing that video? Who is asking whether the handoffs, the meetings, the approval steps are still working the way they should? That is double-loop learning. And in Agile, that's the retrospective.

The Loop model treats retros a little differently — they happen quarterly, which is just four times a year. By HubSpot's own math, that means the team's process improvement is running at four cycles a year while the content is running at 26. To be fair: at the scale of a very large marketing organization, a quarterly reset is sometimes all you can realistically get. And the four retro questions they suggest using to run this meeting are good ones. But I'm going to push here and say your team can go faster than that — and the data says it should.

In our State of Agile Marketing Report, retrospectives showed the biggest gap of any Agile practice between the most successful Agile teams and everybody else. More than half — 52% — of teams with an extremely successful implementation run retros. For everybody else, it's only 35%. And here's the bit that I find surprising: those same top teams were actually less likely to work in two-to-four-week sprints (22% versus 35%) and less likely to hold formal sprint reviews (15% versus 25%). So the takeaway is that it is not the ceremonies automatically that separate the best teams from everybody else. It's the learning habit — the practice of looking back and discussing what works and what doesn't. One practice that improves how the team works means all the work that flows through it going forward gets better.

This is also a complicating point for the Agile community: if you're copying a calendar — Loop or Scrum or whatever — you're copying the visible part. But the compound improvement comes from the part that's easiest to skip.

---

HubSpot has given us a great map of what the cycle should be. Our job is making sure the cycle can actually turn the way it's meant to. Here's the order I'd do it in.

**One:** Find out where insight dies in your current system. Take your last month of work and trace one real learning from Evolve — the end of the loop — back to Express — the beginning. Where did it stall? Was it waiting for a planning meeting? Is it buried in a dashboard somewhere? That stall point is the slowest part of your cycle, and it's where you need to start. Think of it as your operational bottleneck.

**Two:** Put your plan on a rhythm. If you're replanning quarterly, move to monthly. Not constant change — but a known, regular moment where the plan is allowed to move. That rhythm is why 82% of Agile marketers feel confident, not chaotic, when their priorities inevitably change.

**Three:** Cap your work in progress before you scale anything. One or two active experiments per person. AI is going to tempt you to start 12. Do not do it. Finished loops get the opportunity to compound. Open ones just pile up.

**Four:** Retro at the speed you work. Pick a cadence that matches your replanning and protect it. Block those calendars. This is the double-loop learning moment where your team gets smarter about how it works, not just what it does.

And here's the question I'll leave you with: if your marketing learns 26 times a year, but your team only improves how it works four times a year — which one is actually governing your speed? Go ahead and hit back ten seconds and re-listen to that if you need to. I'll be here.

---

Until next time, I'm Andrea Fryrear. And remember: the struggle is real, but so is Agile marketing.

If you want to be able to do everything I just talked about in this episode, we have free micro-learning courses on our online learning platform, The Ropes, at theropes.agilesherpas.com. You can learn retros, you can learn prioritization — all of our courses there are seven to ten minutes long. And if you bring your whole team, we'll give you a free hour of coaching every single month so you can make sure you're doing it right. It's very silly not to take advantage of that. Head to theropes.agilesherpas.com and go learn how to do this right now.