DIY SEO vs Hiring an Agency: An Honest Comparison

Almost every version of this comparison opens with a price table, which is why almost every version is useless. Price is the easiest variable to research and the least predictive of the outcome. Owners have paid nothing and built a presence that still brings in work five years later; others have spent forty thousand dollars across two years and hold a folder of PDF reports.

Two things separate those outcomes, and you can assess both this afternoon without requesting a quote. First: do eight to ten focused hours a month genuinely exist in your calendar, as opposed to your intentions. Second: can you watch a channel produce almost nothing for two or three quarters without abandoning it.

We sell agency services, so read the verdicts knowing which invoice they favour. This page names the situations where hiring us would be the worse decision.

Why the price comparison settles nothing

The standard framing sets a monthly retainer beside a software subscription and invites you to notice the gap. The gap is real and it tells you little, because the two are not the same kind of thing. A retainer buys executed work. A subscription buys the ability to see problems you may not have time to fix.

The second problem runs in the direction that flatters agencies. Most comparisons inflate the DIY side by assigning your hours a high billing rate, then presenting the result as a hard cost. The page currently ranking first for this search does exactly that: it values the owner's time at a fixed rate, multiplies by ten hours a week, and arrives at a DIY figure several times larger than the retainer it goes on to recommend. The arithmetic is fine. The assumption underneath it is doing all the work, and it is never examined. We examine it below. For pricing on its own terms, see how much should SEO cost.

What Google actually tells small businesses

Google publishes guidance on this exact decision, and its position is friendlier to DIY than most agencies would like. Its documentation on whether you need an SEO opens by telling readers that if you run a small local business, you can probably do much of the work yourself, and sends them to the free SEO Starter Guide before it discusses hiring anyone.

Read that fairly in both directions. The Starter Guide is good, free, and covers the fundamentals that account for most of the upside on a simple site. It is also explicit about its limits, telling readers that if they are short on time or not feeling adventurous, they might consider hiring a professional. Google is describing a fork, not a recommendation.

Test one: do the hours exist?

This is where the DIY case usually dies, and rarely for lack of ability. For a single-location business past setup, a realistic rhythm is eight to ten focused hours a month: one substantial piece of content, a pass over the Google Business Profile, a look at Search Console for crawl errors and query drift, one small technical fix. The first month runs heavier, twenty to thirty hours.

Notice the distance between that and the fifteen to twenty-five hours per week quoted in agency-published comparisons. That is a full-time job description, and it appears almost exclusively in articles three paragraphs away from recommending a retainer. Inflating the time requirement is the same manoeuvre as inflating the hourly rate.

So the honest test is not whether nine hours can be found. It is whether you have ever protected recurring hours for something with no deadline, and whether that survived a busy month.

The true hourly cost of DIY, calculated honestly

Now the arithmetic, without a thumb on the scale. The cash is small: Search Console, Analytics, Business Profile and Google's documentation cost nothing, and a starter research tool runs roughly $100 to $150 CAD a month.

The hours are where the disagreement lives. The relevant rate is not your billing rate, it is the value of the next best use of those hours. For most businesses that swings hard with capacity, and the distinction moves the answer further than anything else here.

Two of the four rows point away from hiring anyone, and they are not edge cases. A landscaper in February, a tax preparer in July, an owner who finds the work interesting: ordinary situations where the honest recommendation is to keep the money.

One caveat cuts the other way. Hours are not the only input; accuracy is the other. Nine hours of confident but misdirected work can leave you worse off than nine hours of nothing, with pages competing against each other and a settled belief that the channel is broken.

Test two: can you sit inside the feedback loop?

The second test is temperamental rather than logistical, and it ends more DIY efforts than the first. Search work pays out late and reports back slowly, so you will spend months acting without confirmation. Many people abandon it exactly where it was starting to compound.

The data is more interesting than the folklore. Ahrefs, in a study of one million URLs updated May 2025, found that only 1.74% of newly published pages reach the top ten within a year. That gets deployed as discouragement, which misreads it: the sample is all pages, including vast volumes nobody optimised or intended to rank. The useful number sits beside it. Among pages that did reach the top ten inside a year, 40.82% arrived within the first month.

That turns a waiting game into a checkable cycle: publish, wait eight weeks, watch impressions rather than rank, and treat flatness as a verdict on the page rather than the channel. Ahrefs' companion finding is the counterweight: the average page ranking first is about five years old, and 72.9% of top-ten pages are more than three years old. Early signals read quickly, but durable positions are slow.

Where DIY wins outright

Not where DIY is a tolerable compromise. Where it is the better decision.

A fifth case is common and rarely admitted: you have already been burned. An owner two years into a retainer with nothing to show has usually lost faith in the category, not one vendor. Six months of doing it yourself teaches you what good looks like from the inside, and makes you a much harder buyer afterwards. Our guide on how to choose an SEO agency is written to be used against us as readily as against anyone else.

Where DIY fails, and it fails the same four ways

The failure modes are consistent enough to list, and none are about intelligence or effort.

Competitive verticals. If your market holds three or more well-funded competitors publishing consistently, the bar is not "do the fundamentals," it is "out-execute an organisation that employs people for this." Fundamentals that win a quiet local market are table stakes in a contested one.

Accumulated technical debt. A migration that went wrong, duplicate templates, an accidental noindex, a redirect chain nobody documented. Diagnosis is the hardest thing to self-teach, because you cannot see the shape of what you are missing. A pattern we see repeatedly in audits is a year of competent content work sitting on top of a technical fault.

Four surfaces instead of one. Search work now spans crawl and rendering, editorial production, local entity signals, and whether generative engines treat you as an extractable source. If your gap is one of those, one person can close it. If it spans three, you are not doing SEO in nine hours a month, you are project-managing four disciplines badly. We weigh that against staffing in SEO agency vs in-house, and against a single specialist in SEO agency vs freelancer.

The hours were never real. The most common failure by a wide margin. Setup month goes well, month two is decent, month three brings a staffing crisis, and the project quietly ends. Not the failure of a plan, the failure of a calendar.

Side by side, with verdicts

The summary sits here rather than at the top, because verdicts only mean something once the reasoning is on the record.

Four rows to DIY, three to the agency, one split, tallied by an agency. If your situation loads onto the top three rows, do it yourself. If it loads onto diagnosis, competition and speed, buy help.

The six-question decision framework

Answer these in writing, because the honest answers are less flattering on paper.

Five or six points, do it yourself and start with the free Starter Guide. Three or four, the hybrid below is your answer. Zero to two, hiring is the rational call and the question becomes who rather than whether, which is what our questions to ask an SEO agency are for.

The hybrid most owners actually land on

Treating this as a binary hides the option most businesses eventually choose. Buy the diagnosis, do the maintenance. A one-time audit establishes what is broken and what to publish in what order. You execute it yourself in your eight to ten hours, and buy interventions when you hit something outside your range.

It splits the decision where the two options genuinely differ. Diagnosis is hard to self-teach and cheap to buy once. Execution is expensive to buy repeatedly and entirely learnable. It also fails predictably: the plan runs for two months, then stops, and a year later you hold an expensive document. The hybrid still depends on test one.

Declaring our own interest

Formative Digital is an agency in Brantford, Ontario. We are one of the two options here and have an obvious preference about which you pick, which is why the concessions above are specific rather than decorative.

We tell prospects to do it themselves regularly, on the criteria already listed. In those cases a retainer with us is worse value than a free guide and a protected Sunday evening. Saying so on the call is cheaper than discovering it in month four, and a client who hires us for the wrong reason leaves anyway.

Where we do take the work, the terms are on the record. Results Guarantee: if your existing domain shows no measurable organic search results after 12 months of work with Formative Digital, we work for free until you see them. That is scoped to existing domains, because a new domain's timeline is inherently longer. It is a continuation-of-work commitment rather than a refund, and what counts as results is agreed in writing before we start. The methodology, including Vector 1 Diagnose and Vector 11 Measure, sits in our research library.