Why SEO Agencies Fail

Write down the five phrases you would type if you were the customer, in the words a customer would use, with your city or suburb on the end. Do not think like a marketer. Think like the person with the blocked drain.

Now open the keyword list on your last report and count how many of those five appear. Anything not on that list is not being worked on and is not being measured. If three of your five are missing, the strategy is aimed somewhere other than your revenue.

Open your website in a browser tab and read the tab itself. That text is your title tag. Does it name the service you sell and the place you sell it in, or does it say "Home | Company Name"?

Do the same for your two most valuable service pages. If none of them contain the service plus the location, then months of fees have been spent without the cheapest, most basic piece of the job being done.

Ask for every link acquired in the last six months as a plain list: the URL of the page the link sits on, and the page of yours it points to. This is a request any provider doing the work can answer the same day, because they had to build that list to do the work.

Then open three of them and look for the link. If the list takes two weeks to arrive, comes back as a count with no URLs, or the links are not actually on the pages, you have your answer without needing to understand a single thing about SEO.

Take the same list from the check above and open five of the sites your links are sitting on. Read them as a human being for one minute each.

Does it look like something a person would actually read - a local paper, a trade association, a real business with staff and a contact page? Or is it a site with unrelated articles about loans, casinos and supplements, no named author, and a "write for us" page? You do not need a tool to recognise a website that no human reads. If most of them fail that test, the money spent has been buying risk, not authority.

Read your last three published posts and highlight every sentence containing a fact only your business could have supplied - a real price, a specific job, a step in how you actually work, an objection a customer raised last week.

If you cannot highlight anything across three posts, nobody with knowledge of your business was involved in writing them. That content is unlikely to earn a ranking and it will not persuade the people who read it.

Put last month's report next to this month's and check three things: is it the same list of keywords, in the same order, over the same length of comparison period?

If the keyword set changes between reports, or the date range shifts around, the two documents cannot be compared - and a report you cannot compare with last month's is not a measure of progress. Then count the lines that are about enquiries, calls or sales rather than rankings and impressions. On a lot of reports that count is zero.

On your next call, ask one specific question about a decision: "why did we target that keyword with that page, and what happens to the other page if it works?"

You are not testing whether they know the answer. You are testing whether an answer exists in the room. If every question of that shape has to leave the call and come back translated, then nobody you can talk to is close to the strategy.

Ask a plain question: how many hours of work went into my account last month, and who did them? Then divide your monthly fee by that number of hours.

You now have an effective hourly rate, and you can decide for yourself whether it is reasonable for the seniority of the person doing the work. If nobody can tell you the hours at all, that is the finding.

Ask three questions: who makes the strategic decisions on my account, how long have they been doing this, and how many accounts do they carry?

Then ask for one example, in writing, of a recommendation from the last six months that stopped or reversed something rather than adding to it. Experienced people say no to things - they kill a page that is cannibalising another, or tell you a piece of work is not worth the money. A file of pure additions is what a checklist produces.

Scroll back through twelve months of email and count how many different people have been your main point of contact. Then take the last three deliverables and compare the date they were promised with the date they arrived.

Turnover plus slipping dates is a capacity problem, not a personality problem, and it will not be fixed by you chasing harder. It gets fixed when their delivery side catches up with their sales side, which is not something you can influence from where you sit.

Ask what specifically changes at month four, month eight and month twelve. A plan built to keep you is asking a different question at each stage - build the foundations, then earn the authority, then hold the position and maintain it.

If the answer is the same monthly package repeating indefinitely with no stated end state, ask two more: what is your average client length, and is the person who sold this to me still involved in the work? You will learn a lot from how comfortable those questions are.

Get the scope in writing and go through it with a highlighter. Mark every line that is a deliverable - a page, a link, a press release, a fixed piece of work - and leave the process words alone: monitoring, management, optimisation, reporting.

Then add up six months of invoices and work out what share of that spend put something new on your website or a new link pointing at it. If the deliverable column is thin and the process column is thick, you have been buying supervision rather than progress.

Ask two questions before you sign anything. What happens if I give notice at month four? And would you do the same work month to month, at the same price, without the term?

The answers tell you what the clause is for. Plenty of good providers ask for a minimum term and are worth it - the point is that you should know which thing you are agreeing to, and the discomfort in the room when you ask is itself information.

Right now, try to log in to Google Analytics and Google Search Console with your own email address and see your website. Then check whether you are listed as an owner - not a manager, not a viewer - of your Google Business Profile.

If you cannot get in, ask for owner or admin access to all three, plus your domain registration and hosting. Time how long it takes to arrive. Whatever else is true about your provider, you should own your own data and your own domain, and the response to that request is one of the most revealing things you will ever ask for.