What your SEO retainer is actually paying for

Every month you pay your SEO agency a retainer, and every month some portion of it turns into work on your website. The gap between the two numbers is the part nobody puts on the invoice. It is not concealed out of malice. No agency itemises the office rent, the sales commission and the software licences against your account, so the only figure you see is the total.

This is the eighth failure in my series on why SEO agencies fail, and the first of the structural ones. The earlier failures were mistakes: things done badly that the same people could have done well on a better day. This one is not a mistake, it is arithmetic. Once you understand where a retainer dollar goes, a lot of otherwise baffling agency behaviour stops being baffling.

Where a retainer dollar actually goes

The fee passes through a series of layers before any of it reaches your website. Roughly in order:

• Winning you in the first place. The advertising that found you, the salesperson who called you, the proposal built before you had paid anything, and the commission paid when you signed. Agencies lose most of the proposals they write, so you are also paying a share of the sales effort spent on businesses that never became clients.
• Account management. The person who runs your monthly call, chases approvals and coordinates between you and whoever does the work. This is real labour, and it is the layer you have most contact with, which is part of why an agency can feel busier than the work log suggests.
• The software stack. Rank tracking, crawlers, backlink data, reporting dashboards, call tracking. Some of it is genuinely expensive at agency scale, and a slice of each licence is allocated to each client.
• General overheads. Office space, insurance, accounting, recruitment, training, annual leave, the time senior people spend managing others rather than doing client work, and the dead time between one client leaving and the next arriving.
• Profit margin. An agency is a business and is entitled to make money. A firm running at zero margin will not be there next year, which is not in your interest either.
• Delivery. The person who opens your site, changes something, writes something, or goes and earns a link.

Notice which one is last. Delivery is what remains after the other layers have taken what they need, and that ordering is not a moral failing. It is how a services business with a sales function and an office is built.

There is a version of this article that treats that as an expose. It is not one, and you accept the same structure everywhere else you spend money. When a plumber charges an hourly rate, you know it covers the van, the insurance, the person who answered the phone and the unpaid hour driving to you. SEO feels different only because the deliverable is invisible: there is no fixed pipe to calibrate against, just a fee at one end and a report at the other.

The problem is never that an agency has overheads. The problem is when so little of the fee survives them that a month of SEO is one blog post and a report.

Why the delivery share shrinks as the price drops

This is the part that matters most, because it explains why the cheapest retainers deliver the least value per dollar rather than simply less value in total. Most of the layers above delivery are close to fixed per client. Winning a client costs roughly the same whether the retainer is small or large, because the same salesperson runs the same process either way. Account management time and software allocated per client barely move with price either. Those costs do not scale down neatly just because the fee did.

So as the monthly fee falls, the fixed layers eat a larger proportion of it and delivery gets squeezed from both ends. That mechanism explains behaviours that otherwise look like laziness. If only a few hours a month can be afforded, they have to be cheap hours, which means junior staff following a template. The work has to be repeatable across many accounts, which means a generic blog post rather than a considered fix. Links have to come from whatever is fast and cheap, which is exactly the route by which cheap SEO can damage your domain rather than help it. Nobody in that chain decided to do a bad job. The price decided it for them, and it is why the honest answer to a very low quote is often that no version of the work exists at that price which will move anything.

The ten-minute self-check

You can test your own retainer without knowing anything about SEO. It is one question and one division.

1. Email your agency and ask, in plain words: how many hours of work went into my account last month, and who did them? Ask for roles as well as names, and do not apologise for asking.
2. When the answer arrives, divide your monthly fee by the number of hours. That gives you an effective hourly rate for the work you received.
3. Look at the roles attached to those hours and decide, as the person paying, whether that rate is reasonable for the seniority involved. You do not need a benchmark: you know what a competent professional costs and what somebody in their first year costs.
4. Separately, count how many hours changed your website, content or links, and how many went into reporting, meetings and admin. Both are legitimate, but only one can improve your position.
5. If nobody can tell you the hours at all, stop there. That is the finding.

The maths, done properly, with an illustrative example

The numbers below are invented to show the shape of the calculation, not survey data or a benchmark.

Take a hypothetical retainer of 1,500 dollars a month. You ask the question and the answer is six hours, mostly a junior executive, including an hour of account management. Six hours into 1,500 dollars is an effective rate of 250 dollars an hour for work largely done by somebody junior. Strip out the account management hour and the hour spent building the report and only four hours touched your site, which is 375 dollars per delivery hour. You can decide whether that is sensible for junior execution.

Now a second hypothetical: 3,000 dollars a month, fourteen hours, of which nine were a senior specialist on the site and the links, three were content and two were reporting and the call. That is roughly 214 dollars per hour overall and about 333 dollars per senior delivery hour. The fee is double, the work bought is more than double, and the seniority is higher. That is the pattern I see most often: the more expensive engagement is often better value per dollar, not worse.

One caution, because this number misleads if you lean on it too hard. Hours are an input, not an outcome. A senior person who identifies the one thing holding your site back and fixes it in two hours has done more for you than a junior filling twenty hours off a checklist. If a page has sat just outside the top spots for months, the constraint is usually authority rather than effort, and the reason a page gets stuck at number three is not solved by adding billable hours.

So use the rate as a sanity check, not a verdict. It tells you when a number is absurd. If your fee buys three hours a month of junior time in a competitive market, no amount of talk about strategy makes that add up.

What a reasonable answer sounds like

A good agency will not be rattled by the question, and the answer need not be precise to the minute. You are listening for whether an answer exists at all.

• A number, or a range, with roles attached: about twelve hours, of which eight were the specialist, two content, one account management and one reporting.
• A list of what those hours produced. Pages changed, content published, links earned, technical fixes made. The hours and the output should look like they belong to each other.
• An honest split between delivery and admin. If reporting and meetings are a third of the time, a confident agency says so rather than burying it.
• A willingness to talk about the trade: at your fee level you get roughly this much work a month, here is what that can realistically achieve, and here is what would change that.
• No defensiveness. The question is not an accusation, and treating it as one tells you more than the number would have.

This is the same instinct behind everything I have written about how to choose an SEO provider. You are not looking for perfect answers, but for somebody comfortable being specific about their own work.

Why not tracking hours is itself informative

Sometimes the reply is: "we don't track hours per client, we work to deliverables". That is not automatically a bad answer. Plenty of good firms price on scope rather than time. But it changes the question rather than ending it.

If nobody counts hours, ask for the deliverables instead, itemised for last month: every page touched, every piece of content published, every link acquired with the URL it sits on. A deliverables-based agency should produce that list quickly, because the list is the product it sold you.

You now have a two-part test. Either the agency can tell you the time, or it can tell you the output. If it can do neither, you are not buying a defined service. You are paying a subscription with no specified product, and the work reaching your site is whatever survived the other layers. That is what quietly produces the accounts I get called into once somebody has paid for SEO for months and gone backwards.

What the overheads genuinely buy, and what the alternative costs

Now the fair side of the ledger, because it would be dishonest to call those cost layers pure friction. Overheads buy real things, and for some businesses they are worth the money.

• Continuity. If one person is sick, on leave or resigns, an agency still turns up, and the account does not stop for three weeks.
• Breadth of skills. Technical SEO, content, digital PR, development and analytics are genuinely different disciplines, and a team can put the right specialist on each instead of stretching one generalist across all of them.
• Capacity. If a project needs two hundred pages migrated to a deadline, an agency can put several people on it at once. One person cannot.
• Process and review. Checklists and second opinions catch the kind of error a solo operator working alone at nine at night misses.
• Lower key-person risk. The firm will probably still exist in three years, and your data and history sit in systems rather than one person's head.

Those advantages are real, and a business at a certain scale often should hire an agency. The hourly-rate question is not there to prove agencies are bad. It is to find out whether you are paying for that structure and getting its benefits, or paying for it and getting a template.

Since I am a solo consultant, it would be convenient to stop there, so here is the other column. One person is one point of failure: if I am ill or away, nothing moves that week. There is a hard capacity ceiling, and a client needing a large project delivered fast may be better served elsewhere. My skill set is narrower than a full team's, with no in-house developer or designer on call. The advantage of the solo model is not that the work is inherently better. It is that the layers are shorter, so more of the fee reaches delivery, and the person you brief is the person doing the job. That is a trade, not a free win.

The question underneath all of this

The effective hourly rate is useful, but it is not the end point. The real question is a threshold one: in the market you compete in, is the work your fee buys each month enough to change your position?

Some jobs have a minimum viable dose. A quiet regional market with weak competitors might genuinely move on a few focused hours a month. A competitive metro market where every business above you is investing properly will not. Below that threshold the money is not stolen, it is spread too thin to reach the point where anything changes, which is the more common and more expensive failure.

So take the hours answer and ask the follow-up: given what my competitors are doing, is this enough? A good provider will engage honestly, and may give you the uncomfortable version, which is that your budget will not win the terms you want and you would do better targeting something narrower or waiting until you can fund it properly. That answer costs them revenue, and it is exactly what I would want to hear before signing anything.

Ask the question this month, and divide the number. Whatever comes back, you will know more about what you are buying, and you will be a considerably harder person to sell to.