Almost every software company we talk to has already decided it needs help. What it has not decided is the shape of that help. That question, build or buy, gets settled by whichever proposal or job description landed first, which is a bad way to commit six figures a year.
This guide walks the decision the way we would run it if we were on your side of the table, including the cases where hiring internally beats hiring us. We sell retainers, so treat the framing as informed and biased, and run the arithmetic with your own numbers. The engagement models themselves are compared on our B2B marketing firm page.
Start with the shape of the work, not the budget
Budget tells you what you can afford. The shape of the work tells you what to buy. There are two shapes, and they point in opposite directions.
Front-loaded work concentrates effort at the start and then drops to maintenance. A website rebuild, a technical SEO and information-architecture overhaul, an entity and schema layer for AI search, a content system with the first thirty pages, a paid account restructure: all of that is heavy for a quarter and light afterward. Hiring permanent staff for a front-loaded shape means you hire for the peak and then pay for the trough.
Steady work is the opposite. Lifecycle email, community, sales enablement, event programs, and product marketing run at a constant level and depend on internal context that an outside team acquires slowly. Buying steady work from a firm means paying agency margin for something a good hire does better.
Most software companies below roughly $10M ARR have a front-loaded problem and treat it like a steady one. That mismatch is why so many first marketing hires spend their first two quarters fighting a website they did not choose.
The total cost comparison people skip
The usual comparison, one salary against one retainer, is not a comparison of the same thing. Do it properly with four lines.
Line one, loaded people cost. Take the base salary of each role you would hire and add employer taxes, benefits, equipment, and software seats. Whatever multiplier your finance team uses on base salary, use that, not the base.
Line two, the specialisms your hires cannot cover. Write down every capability the program needs: technical SEO, AI search visibility, content strategy, writing and editing, design, front-end, paid search, paid social, creative, analytics. Mark which your planned hires genuinely do well. Everything unmarked is contractor spend, and contractor spend without internal management is the line that overruns.
Line three, tooling. Analytics and warehousing, SEO and AI search visibility platforms, content and design tooling, CRM add-ons, testing. A firm usually absorbs some of these into its retainer, which is worth checking rather than assuming.
Line four, time to first output. Multiply the months of hiring and ramp by whatever a month of a stalled funnel is worth. If you cannot value that, use your monthly new-pipeline target and take a fraction of it. The number will be uncomfortable, which is the point.
Add the four lines, then divide by the number of channels you need running. Compare that per-channel number against a retainer. Our breakdown of what agencies actually charge gives the market ranges for the second half of that comparison, and our own tiers are published on the pricing page.
A scoring model you can run in an hour
Score each statement from one to five, where five means strongly true of your company today. Sum the first block and the second block separately.
Block A, favors in-house. Our channel needs are narrow and stable. We have a specific senior person we can hire and someone who can manage them. Our product requires deep context to market credibly. We expect the same work to run at the same level for at least two years. Our sales cycle depends on tight daily collaboration between marketing and sales.
Block B, favors a firm. We need four or more specialisms running at once. Our website or technical foundation blocks progress regardless of who owns marketing. We cannot absorb three to six months of hiring time with the funnel unattended. Our biggest near-term work is a build rather than an ongoing program. We would rather buy an owned system now and staff around it later.
A gap of five or more points between blocks is a clear answer. Within five points, you are in hybrid territory, and the structure matters more than the vendor.
The hybrids that work, and the one that does not
Three hybrids hold up over time. The first is an internal owner plus an execution firm: one person internally holds strategy, budget, and the customer relationship, and the firm runs the channels on a shared plan. The second is a build-then-staff sequence: a firm ships the foundation in a bounded engagement, then an internal hire arrives to run and extend it. The third is split scope by decay rate, keeping product marketing, sales enablement, and community internal while search, content production, and paid sit outside.
The hybrid that reliably fails is a firm hired to be the strategy with nobody internal accountable. Every plan involves tradeoffs against product timelines, sales priorities, and budget, and an outside team cannot arbitrate those. It shows up as slow approvals, drifting scope, and a quarterly review where both sides are politely disappointed.
Practically, the internal sponsor does not need to be a full-time marketer. It needs to be someone with authority who will spend 30 to 60 minutes a week on decisions. We say no to engagements where that person does not exist, which is documented in the self-qualification section of our B2B marketing agency page.
How to avoid paying twice
Paying twice happens two ways. Either you hire internally and keep the full outside retainer for a quarter of unclear overlap, or the firm's work is not transferable and your new hire rebuilds it. Both are avoidable in the contract.
Require that you own everything from day one: the site and its repository, the content, the ad and analytics accounts, the search and AI visibility data. Require documentation as an ongoing deliverable rather than an exit favor, meaning playbooks, naming conventions, dashboards, and recorded walkthroughs kept current. Then define the taper: when you hire internally, reduce firm scope one channel at a time with one overlapping cycle each, rather than a single cliff.
A firm that resists any of this is telling you its retention depends on lock-in. The way we handle it month by month is on our methodology page.
What in-house does better, honestly
An internal team wins on context, availability, and institutional memory. A good internal marketer sits in customer calls, hears the objection language before it reaches a transcript, and can reroute a week's priorities in a hallway conversation. They also accumulate knowledge that stays when a vendor relationship ends.
No firm replicates that fully. What a firm can do is reduce how much of it a program needs, by owning the parts that are craft rather than context: technical foundations, search and AI search visibility, production throughput, paid structure. The best programs we run have a sharp internal owner supplying context and a team outside supplying craft and volume.
Where we fit
We are an integrated firm for B2B and B2C software companies: website, SEO, AEO, GEO, content, and paid under one strategist with one measurement layer, published pricing, and a six-month minimum. That is the right purchase when your constraint is fragmentation or a foundation that blocks content and conversion work, and when you are not ready to hire a senior marketing leader.
It is the wrong purchase when Block A of the scoring model wins clearly, when you need positioning or category work as the main deliverable, or when outbound conversations are the actual gap. We turn those down and say why.
If you want a second opinion on the build-versus-buy call, bring your draft job description or your vendor shortlist to a 30-minute growth call. If you would rather keep reading first, the vendor evaluation scorecard covers the diligence step once you have chosen to buy.