Agency, Freelancer or In-House for SaaS Marketing?

Agency, freelancer or in-house: which should a B2B SaaS company choose?

All three, in a sequence, and the sequence matters more than the choice. Freelancers for defined pieces of work, an agency for capability you do not have and do not want to build, and in-house for anything that has to accumulate knowledge about your product and your buyers.

We should declare the obvious: we are a studio, so we are one of the options being compared. We will be as straight as we can about where we are the wrong answer, because a client who hires us for the wrong thing is a client we lose in six months.

Here is the honest comparison, starting with the number everyone guesses at.

What does an in-house hire actually cost?

More than the salary, and the salary is higher than most founders assume. The US Bureau of Labor Statistics reports that "the median annual wage for marketing managers was $166,790 in May 2025", with the lowest 10 percent earning less than $90,260 and the highest 10 percent more than $293,610.

The industry spread matters too. BLS puts marketing managers' median wages from $165,630 in wholesale trade up to $192,600 in information services, which is the band software companies compete in. So a mid-level marketing manager in tech is a six figure commitment before benefits, tooling, recruiting cost and the time your team spends hiring.

For build work the numbers are lower and still substantial. BLS gives a May 2025 median annual wage of $92,650 for web developers and $104,000 for web and digital interface designers, with the combined category at $99,520. Two in-house builders is most of a quarter of a million dollars a year in salary alone.

How do the three compare on paper?

The trade is consistent across every dimension, which makes it easier to reason about than it first appears.

Dimension	In-house hire	Agency or studio	Freelancer
Time to start	Weeks to months of hiring	Days to weeks	Days
Cost shape	Fixed and ongoing	Project or retainer	Variable, per piece
Range of skills	One person's range	A team's range	One specialism, usually deep
Product knowledge over time	Compounds	Partial, project shaped	Rarely accumulates
Main risk	Wrong hire, slow to correct	Priorities split across clients	Availability and continuity

Read down the product knowledge row and you have the real decision rule. Work where accumulated understanding of your buyers is the value should be in-house. Work where current craft is the value can be bought.

That is why positioning belongs in-house and a website rebuild usually does not. One is a body of knowledge. The other is a capability you need at full strength for eight weeks and at zero afterwards.

Where does in-house genuinely win?

Anything that requires being in the room. Product marketing depends on hearing sales calls, sitting in roadmap discussions, and noticing that three customers said the same odd thing this month. None of that is transferable in a brief.

The Product Marketing Alliance's State of Product Marketing Report 2025 puts a useful number on how tight that access is: "88.8% work closely with product, and 81% with marketing". That level of day to day proximity is difficult to buy externally, and the roles that need it are the roles to hire.

In-house also wins on response time for anything customer facing. A competitor changes their pricing page on a Tuesday and someone internal can react that afternoon. An external partner reacts next sprint, and usually should.

Where does an agency or studio genuinely win?

When you need several specialisms at once for a bounded period. A website rebuild needs design, build, performance work, content structure and search understanding simultaneously. Hiring five people for eight weeks is not an option, and one generalist doing all five is how you get a site that is adequate at everything.

The second case is a capability you never want to own. Nobody's B2B SaaS strategy depends on having an in-house Webflow specialist forever. It depends on having a fast, well-built site, which is a different requirement.

The third is pattern exposure. A team that has built thirty B2B sites has seen which pricing page structures cause support tickets and which navigation patterns confuse technical buyers. That is real and it is genuinely hard to replicate internally, because your team will see your site and a handful of competitors.

Where does a freelancer win?

On defined, deep, repeatable pieces. A technical SEO audit. A set of illustrations. Twelve articles against an approved brief. A one-off migration. Anything where the scope can be written down precisely and the output judged on its own.

They also win on cost for exactly that kind of work, because you are buying the hours and none of the coordination. An agency running a twelve article content programme carries account management the freelancer does not.

Where freelancers struggle is continuity and ownership of an ongoing system. If the work needs someone to notice that a process has quietly broken, that is not a scoped deliverable, and it will not get noticed. The realistic version of that constraint is in running a content engine as a team of one.

What is the honest disadvantage of each?

For in-house, it is correction speed. A wrong hire takes three months to identify and another two to resolve, during which the function does not happen. That risk is why the first hire should be scoped narrowly enough to be assessed.

For agencies, it is divided attention, and we will not pretend otherwise. We have other clients. A good agency manages that with clear scope and honest capacity limits, and a bad one manages it by being slow and vague. Ask any agency how many active projects a lead is carrying, and treat a reluctance to answer as the answer.

For freelancers, it is single point risk. One person gets ill, takes a bigger contract, or stops replying, and the work stops. That is survivable on a discrete project and serious on anything your pipeline depends on weekly.

What does the team size benchmark tell you?

That almost nobody is fully in-house. The Product Marketing Alliance report states that "44.3% of PMM teams are still just 1-2 people". If nearly half of product marketing teams are one or two people, the rest of the work is being done by somebody else: agencies, freelancers, or nobody.

The same report notes that "30.7% reported increased investment in PMM this year", so the direction is toward more internal capacity rather than less. But starting from one or two people, more internal capacity still leaves most execution outside.

The useful read is that hybrid is the normal state, not a compromise. Designing for it deliberately beats arriving at it by accident, and it is why we push clients to decide what is theirs before deciding what to outsource.

What sequence do we actually recommend?

Three steps. First, hire one in-house person for the thing only an insider can do, which at early stage is almost always positioning and messaging. Second, buy the bounded capability projects: the site, the design system, the technical foundations. Third, use freelancers for repeatable production against briefs that your in-house person owns.

The ordering matters because the in-house hire is the one who writes the briefs. Buying execution before you have someone who can judge it is how companies end up with a lot of output and no consistency.

What we would avoid is the common inverse: a large agency retainer with nobody internal accountable for the outcome. That arrangement reliably produces activity reports. How the spend should be split sits in how to allocate a B2B SaaS marketing budget.

How should this change as you scale?

The line moves inward. Work that was bought becomes work that is owned, in roughly the order that it becomes continuous. Content production comes in-house when it is weekly. Design comes in-house when there is a product interface as well as a website. Build usually comes last, because it is the most bursty.

What should not change is the test. Does this work compound knowledge about our buyers, or does it need current craft? The first is a hire. The second is a purchase. Companies get into trouble when they hire for craft and buy the compounding work, and BLS salary levels mean that mistake is expensive in both directions.

If you want an honest answer about whether your next piece of work should be a hire or a project, we will tell you when it is a hire. That conversation starts at phoenix.studio, and the build-specific version of this question is in Webflow agency against in-house.