How to choose the best SEO agency (and the promises that should scare you)

Why SEO is the easiest service to fake

Because it is slow. A paid media agency is exposed in six weeks, the money goes out, the leads come in or they do not. An SEO agency can bill for a year before anyone can prove the work was worthless, and "SEO takes time" is a genuinely true sentence that also happens to be the perfect alibi.

So the best SEO agency is not the one with the most confident deck. It is the one that will show you, month by month, what it changed and what moved, and will tell you early when the thing holding you back is not SEO at all.

Promises that should worry you

• "We guarantee page one." For which keyword, and who chose it? Guaranteed rankings usually mean a keyword nobody searches for.
• "We have a special relationship with Google." Nobody does. Google's SEO documentation is public and there is no back door.
• "We will build 500 backlinks a month." Volume link building is a good way to buy a problem you will pay someone else to clean up later.
• "We will publish 20 blog posts a month." For whom? Content with no search demand and no opinion is just hosting costs.
• "Rankings are up 300%." From position 90 to position 30 is a 300% improvement that earns exactly zero clicks.

The common thread: every one of those is a metric an agency can produce without moving your revenue an inch. If a proposal is full of things that can be true and useless at the same time, treat that as the finding.

The questions that actually reveal something

1. 1"Which keywords will you NOT chase, and why?" A real strategy excludes things. A keyword list is not a strategy.
2. 2"What will you fix in month one?" Technical foundations and the pages that already rank on page two. If the answer is "start publishing", ask why.
3. 3"How do you connect rankings to revenue?" Rankings are a means. If nobody is looking at enquiries, nobody is looking at your business.
4. 4"What would make you tell us to stop?" Everyone should have an answer.
5. 5"Who writes the content, and do they talk to our customers?" Content written by someone who has never heard your customers speak reads exactly like it.

And here is the opinion we will defend: SEO and paid media should not fight for credit. They should share search intelligence. Your paid search account already knows which commercial angles convert and which objections come up, that data should be shaping the SEO plan before a single page gets written. Agencies that keep the two in separate silos are wasting the most useful dataset you own.

When not to hire an SEO agency

SEO compounds, which is the good news, and it is slow, which is the catch. There are situations where hiring one is simply the wrong call.

• You need revenue in 60 days. SEO will not save you. Paid search buys demand that already exists, today.
• Your website cannot convert the traffic it already has. Ranking better just sends more people to a page that does not work. Fix the site first.
• Nobody searches for what you sell. Rare, but real. Then you have a demand creation problem, not a ranking problem.
• You cannot commit past six months. Half an SEO program is not half the results. It is closer to none.

That second one is the most common by far. We have seen traffic and rankings look perfectly healthy while enquiries stayed flat, because the site buried the offer, loaded slowly, and made people hunt for the next step. Fixing the website did more for growth than any amount of additional ranking would have.

When the foundations are right, though, SEO is the cheapest long-term traffic you will ever own. That is what our SEO work is built around, and if local search is where your customers actually are, that is a different job with a different plan.

What the first three months should actually contain

Because SEO is slow, the first ninety days are the only real evidence you get before you are a year in. Here is what should be happening in them.

Month one is technical and structural. Can Google crawl and index everything that matters? Is the site fast enough to clear Core Web Vitals? Is the internal linking sending authority to the pages that earn money, or to the privacy policy? None of this is exciting and all of it is load-bearing.

Month two is the pages that are nearly there. Every site has content sitting at position 8 to 15, close enough that a genuine improvement moves it onto page one. This is the fastest return in SEO and it is routinely ignored in favor of publishing new articles, because new articles are easier to invoice for.

Month three is where new content starts, informed by what the first two months revealed and by what your paid search data already knows about which angles convert. Publishing before this point is guessing with a content calendar.

• Ask for the crawl report. If they have not run one, they have not started.
• Ask what they fixed, not what they published. Fixing beats publishing for the first quarter, nearly always.
• Ask which existing pages they improved. If the answer is none, they are ignoring the cheapest wins on the site.

An agency whose first ninety days is "we published twelve blog posts" has done the easy, visible thing and skipped the work. Google's own SEO documentation is unambiguous that the foundations come first, and there is no way around them.

Frequently asked questions