Run Your Marketing Website Like a Product Team

A recent client came to me with two lines going the wrong way. Pipeline was down. Organic traffic was down. Everyone could see the charts.

What nobody could agree on was where to start.

That is a hard place to be. When the numbers drop, every cause looks plausible: search changes, the messaging, the navigation, the campaigns that ended, the tracking itself. The list of possible fixes keeps growing. Nothing moves.

My first step was diagnosis. I went deep into their GA4, Search Console, and Ahrefs data to find the patterns underneath the decline. That work is still in progress, and I will not pretend it has a tidy ending yet. But it has made one thing clear to me. The team is not short on data or ideas. It is short on a repeatable way to get from a problem to a decision, and then to the next one.

Product teams have that. Most marketing websites do not.

The big launch is an expensive way to learn

The usual answer to a struggling website is a big launch. A redesign, a relaunch, a new site.

I understand the appeal. It feels decisive. But big launches are expensive in money, time, and people, and they are hard to learn from. When the new site goes live and the numbers move, what caused it? The messaging? The navigation? The new forms? The launch changed everything at once, so the answer is usually a guess.

Then attention moves on. The site starts aging the day it launches, while buyers, search demand, the product, and the positioning keep changing around it.

A redesign is sometimes the right call. Even then, it works better as one release inside a longer way of operating than as the way of operating.

Product teams already solved this, in more than one way

Software teams live with the same problem: a system that is never finished, users who keep changing, and a limited number of people to work on it. They have built many ways to handle it. Agile. Lean. The minimum viable product. The vocabularies differ, but they share one idea: don’t bet everything on one big release. Learn in small, regular steps.

The framework that shapes how I apply that idea to marketing websites is GPS, from Matty Wishnow’s Listening for Growth. It draws on lean and agile product thinking and lays out the full loop in six steps, which he calls poses:

1. Desires: what the founder and the company fundamentally want
2. Goals: those desires turned into measurable targets
3. Problems: the root obstacles most responsible for blocking the goals
4. Hypotheses: specific solution claims tied to those problems
5. Experiments: a planned way to test each hypothesis
6. Outcomes: what the team learned, who needs to know, and what comes next

What I value most is that it doesn’t start with the website. It starts with what the business actually wants. And it doesn’t stop at the fix. It keeps going into what the team learned and how that feeds the next cycle.

At its core, GPS is about listening: to the business, the founder, the CMO, and the customers. Then you turn what you hear into goals, problems, hypotheses, and experiments that move you toward those goals.

What the loop looks like on a marketing website

The translation is fairly direct:

Product team	Marketing website team
Product outcome	Pipeline outcome
User research	Buyer and customer research
Product analytics	Website and funnel analytics
Product backlog	Website improvement backlog
Hypothesis	Messaging, UX, content, or technical change
Sprint or release	Experiment or site improvement
Product manager	Accountable website owner

When I ran a website this way, no single practice made the difference. They worked together. We kept a backlog of changes and ideas. Each one had a clear goal, a testable hypothesis, and a documented outcome. And we shipped on a regular schedule.

Two-week sprints worked well. Each release had acceptance criteria and a short story about what it should accomplish.

This was looser than standard product agile, and it had to be. Marketing moves fast. A campaign can land in the middle of a sprint and change what matters. That is fine. The backlog absorbs the change, and the loop keeps running.

Brand judgment and the MVP

Not every decision on a marketing website waits for data. Sometimes the story, the brand, and the design execution matter a little more than the numbers. It has to be a balance.

At a SaaS company where I ran the website, we sometimes had ideas for larger projects, like a big year-in-review data piece. The instinct with a project like that is to wait until every part is right. We tried to treat it the way a product team treats a new feature: get a minimum viable version live in time for the campaign, then expand and improve it over the next several sprints.

I have seen the opposite happen enough times. “Perfect is the enemy of good” is an old line, and websites keep proving it. Meaningful improvements get lost to time because the effort, the people, and the months all went into launching something exact.

Borrow the logic, not the theater

None of this requires full Scrum, more meetings, or a new vocabulary for the marketing team.

Wishnow describes GPS as tunable. A team decides how much proof a decision needs, which he calls fidelity, and how deeply the practice is built into daily work, which he calls rigor. A lean team can start simple:

• A clear goal
• A regular look at the evidence
• A prioritized backlog
• Explicit hypotheses
• A steady release rhythm
• A look back after each release

Someone has to own the loop

This is the part I think gets missed most.

Ideally, the whole team is aligned across the whole loop. The gaps appear when it isn’t.

Many marketing leaders work at a higher strategic level. That is a real part of the job. They set the goals. The ICs execute. Between the two, the loop can go unowned. The leader is not always managing down into the problems and hypotheses. The IC is not always managing up, or is so focused on shipping the next thing that nobody steps back to ask what the last thing taught the team.

Nobody is doing anything wrong here. Everyone is doing their own job. The loop just falls between the jobs.

I think that is part of what is happening at my recent client. There is plenty of data and no shared way to move from the problem to the next decision. My next step with them will likely be the full GPS method as a roadmap: from what they want, to measurable goals, to the few problems most worth solving, and on from there.

On a lean team, that ownership has to live somewhere. Sometimes it is the marketing leader. Sometimes it is a senior IC with the time and permission to own it. Sometimes it is a fractional website owner, which is a lot of the work I do now through Plerom.

Start with one turn of the loop

You do not need a new process to begin. Pick the one action on your website that matters most. Write down one goal, one problem you have evidence for, and one hypothesis about what might fix it. Ship something in the next two weeks. Write down what happened.

Then do it again.

It is less dramatic than a launch. It is also how a website keeps getting better after everyone has stopped looking at it.