Every B2B company makes this call wrong at least once. The damage is not obvious for six months, then it is. A quarter of wasted spend, a brand voice that has drifted, a team that cannot explain what the company stands for because the agency owns the answer.
The build-vs-buy question is not about cost. It is about where judgment lives. Get the function map wrong and the company ends up renting its own strategy back from a vendor.
The function-by-function matrix
Marketing is not one function. It is eight to ten, and each one has a different answer to the in-house-vs-outsource question.
Content and narrative. In-house, or in-house plus a strategic partner who sits close. The person writing the company's story needs to be in customer calls and in the room when positioning shifts. A contractor who drops in once a month writes content that reads fine and lands flat.
Design production. Outsource for most companies under $20M ARR. A strong external design partner is cheaper, faster, and more versatile than a single in-house designer. Bring design in-house when brand becomes a competitive surface the company wins on, which is rare below Series C.
Paid media. Outsource execution, own strategy. A good agency or fractional paid operator runs the accounts better than a generalist in-house hire. What stays in-house is the decision about which channels to fund, what the offer is, and what "working" means.
Web development. Outsource almost always. Unless the website is the product, an agency or contractor build-and-maintain model is faster and cheaper than hiring. The in-house counterpart is a marketer who owns the site and directs the contractor.
Analytics and ops. In-house from the moment the company has more than two channels. Marketing ops is where judgment about attribution, lead flow, and reporting accumulates. An outsourced ops vendor turns into a black box nobody can debug before a board meeting.
PR and comms. Outsource to a specialist on retainer. The relationships, the media list, and the pitch craft are specialist skills that do not justify a full-time seat until the company is doing meaningful news. What stays in-house is the spokesperson and the POV.
Product marketing. In-house the moment the company has more than one product or more than two buyer personas. It cannot be done by someone without daily access to product, sales, and customer data.
Events. Outsource logistics, keep the strategy. Who the company sends, what they say, and which events matter are decisions that stay inside.
The principle underneath the matrix
The rule is simple. Outsource execution, never outsource judgment.
Judgment is the decision about what the company says, to whom, and why. Positioning, narrative, category, and the choice of what matters. These compound only through daily proximity to customers, product, and sales. An outside operator can shape them but cannot own them.
Execution is the craft of turning those decisions into work. Copy production, ad ops, video editing, site builds, design systems. These skills transfer cleanly because the deliverable is objective and the judgment is already made.
Every function on the matrix sorts into one of those two buckets. When a founder cannot tell which bucket a function sits in, that is the signal to slow down before making the hire.
Where the rule bends
Three situations bend the principle.
Tiny teams. A pre-Series A company with two people cannot house the full strategic function. The right move is a fractional operator who carries senior judgment while the founder stays close to the work. Fractional is the in-house-shaped solution for a company too small to house the seat.
Specialist work. Some skills are too narrow to justify a full seat, even at scale. Technical SEO audits, international expansion research, crisis comms. These are episodic deep-expertise engagements that belong with specialists permanently, not contingent on company size.
Launch sprints. A product launch, a rebrand, or a category move compresses three quarters of work into three months. Outside capacity is often the only way to meet the deadline without breaking the team. The rule here is that the sprint ends and the core function goes back in-house.
None of these contradict the judgment rule. They adjust where the judgment sits temporarily while the company figures out what the permanent shape looks like.
What breaks when you outsource the wrong thing
Three failure modes show up consistently.
Message drift. The company outsources content without keeping an internal editor. Three months in, every post reads fine in isolation and nothing sounds like the company. Sales starts hearing "your marketing feels generic now." This is what happens when the writer stops hearing customer calls. For more, see B2B marketing without a full team.
Slow feedback loops. Paid is outsourced and the agency reports monthly. A bad campaign runs for three weeks before anyone internal notices. The spend that would have been killed on day three in-house runs to day twenty-one outside.
Strategic drift. The company treats a fractional engagement as a handoff instead of a partnership. Six months later, the founder cannot explain the marketing plan in a board meeting because it lives inside the vendor's head. Lean B2B marketing assumes the founder stays in the plan.
The three signals to bring something in-house
Knowing when to bring a function back is the other half of the matrix.
The first signal is that the function has become a standing part of how the company competes. If paid media is 40 percent of pipeline, a full-time paid marketer inside beats an agency on any measure that matters. When something crosses from a channel the company runs to a channel the company wins on, it comes in-house.
The second signal is frequency. A function that needs daily attention outsources poorly. Content at a cadence of three posts a week is in-house work. Content as a quarterly campaign push is outsourceable. The breakpoint is usually whether the work needs to sit in Slack every morning.
The third signal is that the vendor is now explaining the company back to the company. When outside operators know the customers better than the internal team does, the judgment has migrated out of the building. Bringing it back in requires a senior hire who can absorb that knowledge and own it going forward.
The matrix is not static. Functions move in and out based on stage, strategy, and team shape. What stays constant is the principle: judgment in-house, execution wherever it gets done best. Companies that hold that line build marketing orgs that compound. The ones that blur it pay for advice they cannot remember receiving.
Common questions.
Which B2B marketing functions should stay in-house?
The functions that carry judgment should stay in-house: positioning, narrative, category strategy, customer research, and the relationship with product. These do not survive being outsourced because they require context that compounds only through daily proximity to the business. Everything downstream of those decisions is fair game for outside help.
What are the safest B2B marketing functions to outsource?
Execution-heavy specialist work outsources cleanly: paid media management, SEO technical audits, web development, video editing, design production, and PR outreach. These are skills with clear deliverables, measurable outputs, and a low penalty for the operator not living inside the company. Anything where the work is mostly craft and the judgment is small.
What actually breaks when you outsource the wrong function?
Three things break in sequence. Message drift, where every asset starts sounding slightly off-brand because the operator does not hear customer calls. Slow feedback loops, because the outside team cannot iterate at the speed an internal team can. Finally, strategic drift, where the company stops making choices about what matters and starts running whatever the vendor proposes next.
When does the rule of keeping judgment in-house bend?
It bends when the company is too small to house the function, when the work is short-horizon and specialist, and during launch sprints where outside capacity is faster than building internally. A pre-Series A company often cannot house a full content team, and a fractional operator with judgment is the right answer. The rule is about where judgment lives, not where hands live.
