Sold beyond capacity

Onboarding was probably the best experience you had had with a supplier in years. Fast replies, a proper kickoff call, a document that made it feel like someone had genuinely thought about your business. Then it slowed. A deliverable slipped by a week, then by a month. Replies took days rather than hours. And somewhere in there an email arrived introducing your new account manager, followed a few months later by another one just the same.

Worth naming plainly, because most owners take it personally: this is a capacity problem, not a character problem. The agency sold more work than its delivery side can carry, and you are standing in the queue that creates. It is number ten in my list of why SEO agencies fail, and it sits in the structural group because nobody has to behave badly for it to happen. It falls out of how the business is built.

Sales scales in a few weeks. Delivery does not.

A new salesperson can be productive quickly. There is a deck, a pricing sheet, an audit tool that produces something impressive in an afternoon. Someone capable is booking meetings inside a fortnight, because selling SEO is the same conversation over and over.

The person who runs your account is a different proposition. They need to understand your market, read a results page and know why it looks the way it does, and judge whether a page is worth rescuing or folding into another one. That takes months of supervision before they are useful and years before they are good. You cannot buy it in on Monday because someone signed a contract on Friday.

The money pushes the same way, so every business under pressure resolves the timing gap identically: sign first, staff later, fully intending to catch up. Some do. Often the sales side never eases off, because sales is the part that is working, and the part that is working is the last thing anyone slows down.

A signed client is revenue this month. A delivery hire is a cost this month and a useful person some time next year. That asymmetry is the whole story.

The last piece matters most to you. Sales strain is visible inside the business: there is a pipeline and a number being hit or missed. Delivery strain is invisible until it reaches a client. It shows up as your deliverable date moving, which makes you the measuring instrument.

Why the best onboarding you have ever had can be the warning sign

Good onboarding is genuinely good, and this is not a reason to be suspicious of competence. The point is narrower: onboarding is the most templated and best resourced part of an engagement, and its speed tells you about the sales side rather than the whole company. The kickoff deck is version four hundred of the same deck. The audit that landed in three days was largely produced by software. It is fast because it is a machine, not because the organisation is fast.

The work you actually bought is the opposite kind. Deciding what to do in month seven about a service page that climbed to the middle of page one and stopped needs judgement from someone experienced, and that queues behind every other client who needs the same scarce person. So watch the gradient rather than the level. Unremarkable in month one and still unremarkable in month twelve is not this failure. Outstanding in month one and hard to reach by month eight is.

The account manager carousel, and what it costs you

Turnover is not by itself an indictment. It becomes meaningful when it is the visible edge of overload: account managers carrying too many clients, all of whom are chasing, more time spent apologising than working, and the people with the most options leaving first. The cost to you is not learning a new name. It is context. Think about what the person who left knew about your business:

• That you do not want a page built for one particular suburb, because of a job that went badly there and a type of customer you would rather not attract again.
• That phone calls are the only enquiry that matters and form fills are mostly time wasters, so a report celebrating form submissions is measuring the wrong thing.
• That your busy season starts in September, so anything meant to help with it must be live by August.
• That there is one page you have asked twice not to touch, because a referral partner links to it and the URL cannot change.
• That your pricing sits at the higher end deliberately, so copy leaning on cheap and affordable actively hurts you.

None of that is in a CRM. It lives in the head of the person who was on the calls, and when they leave it goes with them. What you get instead is an invitation to explain your business again, which is unpaid work you are doing for your own supplier. By the third time most owners quietly stop bothering, because correcting small mistakes costs more than living with them. That is the real damage: you end up with generic work because you gave up on the process that made it specific.

There is a momentum cost too. Every handover contains a quiet month while the new person reads the file, and it is never the checklist work that gets dropped. Templated tasks survive; judgement calls evaporate. That is a large part of why a page can sit stuck at number three for a year: the next push was sensible, and the person who understood why it mattered had gone.

The ten-minute self-check

This needs nothing but your email archive and a piece of paper. Do not rely on your impression, because impressions get shaped by whoever was most recently apologetic on a call. Get the dates out.

1. Search your inbox for the provider's domain and scroll back twelve months. Count how many different people have been your main point of contact. One is normal, two happens, three or more is not about personalities.
2. Take your last three deliverables and write two dates next to each: the date promised, and the date it arrived. Find the promise in writing rather than trusting memory, and note whether the gap is growing.
3. Take the last five emails you sent that genuinely needed an answer and write down how long each took. Do the same for five from your first month.
4. Compare month one with last month on volume, not on how it felt. How many calls were held, and how much work actually landed on your website?
5. Ask one plain question by email: how many active clients does the person doing the work on my account handle? Note the answer, and note whether an answer arrives at all.

Now read back what you wrote. Turnover alone can be coincidence, one late deliverable is a bad month, slower replies might be one person having a rough patch. It is the combination that identifies the problem, and it will not be fixed by you being firmer.

Chasing harder makes you the squeaky wheel, not a priority

The natural response is to chase, and here is the frustrating part: it works, for a while. Your overdue thing moves up the queue and lands, which teaches you that chasing is the solution, so you chase more. But nothing was created. Capacity did not increase because you sent a follow-up, it was reallocated, away from a client who now waits until they chase too. Run that forward and the account is prioritised by who complains loudest, a competition you have to win again every month for something you already paid for.

Chasing does not create capacity. It only decides who gets the capacity that already exists.

It also changes what the relationship is about. Calls that used to cover strategy become status updates on overdue items, and the senior attention you do attract gets spent apologising rather than thinking about your website. You have converted a marketing budget into a project management job you do for free, and what all that effort buys is compliance, when judgement was the thing worth paying for.

The honest counter-case

Now the other side, because I would rather you diagnosed this correctly than went looking for it. Good providers have genuinely difficult months. Someone goes on parental leave, a large website migration eats February, a key person resigns with no notice. One slipped deliverable is not a pattern, and two staff changes across three years is ordinary turnover, not a carousel.

Two things separate a bad month from a capacity problem. First, the direction of travel over six to twelve months: a bad month is a dip in a line that returns to where it was, a capacity problem is a line that keeps sloping the same way. Second, whether you were told before the date passed. A provider under temporary strain contacts you in advance, gives a new date, and hits it. An overloaded one lets the date go by silently and responds once you notice.

Check your own side of the ledger too. Content sitting in your inbox for three weeks awaiting approval, photos requested in May and never sent: a fair share of what feels like a slow agency is a stalled approval loop with your name on it. Work out where the waiting actually happened on those three deliverables. If the delay is yours, the fix is easy.

What a capacity-honest provider does instead

The alternative is not a provider who works faster. It is one who decided in advance how much work they can carry, and then behaves in ways that look commercially irrational:

• They know how many active clients each person doing the work carries, and will tell you the number without hedging.
• They offer a start date rather than an immediate yes. A waitlist of a few weeks beats availability this afternoon.
• They say no. Not yet, not this quarter, or not at all because it is the wrong spend for your situation.
• They quote dates with slack built in, and tell you before a date moves rather than after.
• When something slips, a person tells you why, instead of it arriving as a silence you had to break.

I should declare an interest. I run a deliberately capped practice, so I turn work away and sometimes ask people to wait, because I would rather say I can start in six weeks than sign someone today and be the reason their October slipped. Treat that as disclosure rather than a pitch, because the useful part is not that I do it: ask any provider what their cap is and how they know when they have hit it. Someone who has thought about it answers in four seconds. Someone who never has will say they scale to demand, which is a description of the exact problem in this article. The question belongs alongside the rest of how to choose an SEO provider.

What your realistic options are

Assume the check came back clearly. You cannot fix an agency's staffing from the client seat, so the honest list is short. It is still a real list.

1. Name it once, in writing, to someone senior. Not another chase: a short summary of what you found. Three contacts in twelve months, deliverables late by two, four and six weeks, replies averaging four days. Then ask what changes, and by when.
2. Shrink the scope to what they can actually deliver. Two things a month that reliably happen beat an ambitious plan permanently three weeks late, and the fee should come down to match.
3. Take back whatever is stalling in a place you control. If the hold-up is approvals or publishing, moving that step to your side removes it from their queue.
4. Set a review date with conditions: sixty days, three named deliverables, dates agreed in writing. Judge it on those dates, not on how warm the calls felt, because warmth is the one thing that never runs short.
5. Leave, if the review date comes and goes. Before you give notice, make sure you hold owner access to your Analytics, Search Console and Google Business Profile, and that the domain is in your name.

If you do leave, go in with your eyes open about what you are inheriting. A year of slow delivery usually means less was done than you paid for rather than something actively harmful, which is a far better position to be in. The starting point is the same one I would use if you had paid for SEO and gone backwards: establish what was actually built, what was actually linked, and what state the site is in before anyone sells you the next plan.

Diagnosing this rather than stewing on it matters because it changes the question. While you believe it is about attitude you will keep applying pressure, and pressure has no effect on how many people work there. Once you can see it is structural, the question becomes whether the service you are getting, at the speed you are getting it, is worth what you are paying. That one you can settle this afternoon.