Most companies decide this on feel. In-house sounds cheaper and more committed. An agency sounds faster and safer. Both stories are wrong about half the time, and the deciding factors are boring: how many distinct skills the work needs, how much of the job is execution versus strategy, and who is going to manage the person doing it.
Here is the framework we use.
What each option actually costs
| | In-house hire | Agency | Hybrid | | :-- | :-- | :-- | :-- | | Cash cost | $85,000-$140,000/yr fully loaded | $3,000-$15,000/mo | $70,000-$110,000/yr + $2,000-$6,000/mo | | Skills you get | One person, one skill set | Strategy, content, technical, links, reporting | Internal owner plus specialist coverage | | Ramp time | 30-90 days to hire, 30-60 to context | 2-4 weeks | 30-90 days | | Context on your business | Highest | Lowest at first, grows | High | | Continuity risk | One resignation ends the program | Account manager churn, not program churn | Lowest | | Fires you can be | Slow to fix a bad fit | 30-60 day exit | Split |
Fully loaded means salary plus payroll taxes, benefits, tools, and training. A $95,000 SEO salary is closer to $125,000 in real cost, and the tool stack alone runs $300-$1,500 a month before anyone does any work.
The three questions that actually decide it
1. How many distinct skills does the work need?
SEO is not one job. It is technical auditing, content strategy, writing, digital PR and link acquisition, local search, analytics, and increasingly AI search visibility. One hire covers one or two of those well. If your plan needs four, an agency or a hybrid is cheaper than three hires, every time.
2. Is the bottleneck strategy or volume?
If you know what to do and cannot get it published, you need production capacity, and that is what agencies and contractors are good at. If nobody in the building knows what to do, you need senior judgment before you need volume, and hiring a mid-level in-house SEO to figure it out is the most common expensive mistake in this decision.
3. Who manages the work?
An in-house SEO with no manager who understands SEO is an unmanaged hire. An agency with no internal owner is an unmanaged vendor. Either way the same failure shows up in month five: activity reports, no accountability, nobody able to say whether it worked. If you cannot name the person who will own the relationship internally, fix that before you choose.
When in-house is the right answer
- Your site is the product. Ecommerce, SaaS, marketplaces, publishers. Search work touches the roadmap daily and an outside team cannot move at that speed.
- You publish constantly. High content volume with tight brand or subject-matter constraints is easier with someone inside the building.
- You already have senior direction. A CMO or head of growth who can set strategy and judge output turns a mid-level hire into a good buy.
- Compliance or access is heavy. Regulated industries where an outside vendor cannot touch systems without a month of paperwork.
When an agency is the right answer
- You are a local or multi-location service business. Local search, reviews, paid search, and site work are several skills and none of them is a full-time job on its own.
- You need results inside two quarters. Hiring takes 60-120 days before anything ships.
- You have never done this before. Buying a pattern that has worked in your category beats paying someone to learn on your budget.
- The work is seasonal or campaign-shaped. Paying a salary through a slow quarter is worse than pausing a retainer.
Compare shortlists by category on the directory, and see how we score them in the methodology.
The hybrid most growing companies land on
One internal owner, typically a marketing manager or a mid-level SEO, plus an agency for the specialist layers: technical audits, link acquisition, and content production at volume. The internal person owns priorities, the calendar, and access to the rest of the business. The agency owns execution depth.
This costs more than a single hire and less than a full internal team, and it fails in exactly one way: when the internal owner becomes a message-forwarder instead of a decision-maker. Give them the authority to approve, reject, and reprioritize, or the arrangement is just an expensive layer.
The numbers to judge either one on
- Cost per qualified lead or booked job, not rankings or traffic.
- Non-brand organic revenue or pipeline, tracked monthly. Brand search flatters everyone.
- Publishing throughput: pages shipped per month against the plan.
- Time to first meaningful change, which should be weeks, not a quarter of onboarding.
- Concentration risk: what breaks if this person or this account manager leaves next month.
Set the same scorecard for both options before you choose. Most in-house versus agency arguments are really arguments about two different scorecards.
When not to do either yet
- You cannot track a lead to a source. Fix measurement first or you will be unable to judge the result. See leads vs qualified leads vs booked jobs.
- Nobody answers the phone or the form fast. Demand is not your constraint.
- You have no budget past six months. SEO under-delivers when it is cancelled at month five, which is roughly when it starts working.
- You want one person to do SEO, paid, social, and email. That job posting describes four people, and you will get the weakest version of all four.
The short version:
- In-house costs $85K-$140K fully loaded and buys one skill set with deep context.
- An agency costs $3K-$15K a month and buys several skills with less context and faster ramp.
- Decide on skill count, whether the bottleneck is strategy or volume, and who manages the work.
- Most growing companies end up hybrid: one internal owner with real authority plus specialist execution.
- Judge both on cost per qualified lead and non-brand pipeline, using the same scorecard.
- Compare agencies by category on the directory, read how much a marketing agency costs and agency vs freelancer, and see the methodology.