SEO Agency Evaluation

The hardest part of hiring an SEO company is that you are being asked to evaluate technical claims you have no way to check. That asymmetry is uncomfortable, and some companies rely on it.

The way out is to stop trying to judge the technical claims and judge the behaviour instead. You already know how to evaluate behaviour.

And if you have three quotes in front of you, you have probably noticed the other problem: they are not comparable. One lists deliverables. One promises outcomes. One quotes hours. The prices differ by a factor of three and you cannot tell whether that reflects scope, quality, or confidence.

This is not accidental. A quote that cannot be compared cannot be undercut.

First: normalise the quotes

Before price enters the conversation, build a simple grid. Six rows, one column per agency.

1. Google Business Profile work. Is it included, and what specifically? Category audit, service area, hours, services, photos, review process. For most local Henderson businesses this is the highest-value work in the engagement, and it is the thing most often assumed rather than stated.

2. Content. How much, on what, and who writes it. "Four blog posts" is a volume, not a plan. Ask what the first three would be about and why.

3. Technical work. What is included, and is there a cap? Some quotes include an initial fix and nothing after.

4. Reporting. What is in it, how often, and — critically — is map position measured from one location or a grid of points across the city?

5. Lead tracking. Is a tracking number included? Is form attribution set up? Who owns the number? Without this, nobody can tell later whether the engagement worked.

6. What is explicitly excluded. Ask directly. The exclusions are usually where the price gap actually lives.

Once that grid is filled in, the quotes are comparable, and quite often the cheapest one turns out to be cheapest because it does not include the part that produces revenue.

The five questions, and their right answers

Send these in one email to all three. The content of the answers matters. The manner of the answers matters more.

1. Do you measure map position from one location or a grid of points?

Right answer: a grid.

Distance from the searcher is one of the three things Google weighs for local results, and the map pack is recalculated for every location a search happens from. Henderson is large enough that Green Valley, Anthem and the Water Street District return different results for the same query at the same moment.

So a single ranking screenshot is a result from one point — very often your own address, which is the easiest place on earth for you to rank. An agency that tracks from a grid can tell you you are winning Green Valley and losing Anthem. An agency that tracks from one point can only tell you a number that may not be true anywhere your customers stand.

This one question separates the field faster than any other.

2. Which accounts will be registered in my name?

Right answer: all of them.

Domain, hosting, Google Business Profile, Analytics, Search Console, and the tracking number. You should have owner-level access to every one, from day one, in your own account.

Any hesitation here is the single biggest structural risk in the relationship. Accounts held by an agency turn leaving into a negotiation, and a tracking number owned by the agency takes its entire call history with it when it goes.

3. Will you show me a dated work log, at task level?

Right answer: yes, monthly, without being asked.

"Content optimization" is a category. "Rewrote the water heater page title and meta description on 14 March" is a task. The difference is whether you can verify anything.

An agency doing real work already produces this. One that hesitates, or offers a summary instead, is telling you what the reporting will look like in month six. If producing a work log takes a week, it is being created for the first time.

4. How are leads tracked and attributed?

Right answer: a tracking number and source-recording forms, set up before work starts.

Without attribution, you will be comparing gross revenue against an invoice and guessing at causation. That is how good engagements get cancelled and bad ones get renewed.

Note the timing in the answer. Tracking set up at the start gives you a baseline. Tracking added in month four gives you an argument.

5. What is the notice period?

Right answer: thirty days or less, clearly stated.

SEO genuinely takes six to twelve months to mature, and that is an argument for patience — not for surrendering your ability to leave. An agency that must re-earn the relationship every month has a permanent, structural reason to keep the reporting honest. That is the entire argument for month to month.

If a long contract is genuinely required, ask what specifically happens in month eleven that could not happen if you were free to leave in month four. The answers are usually about cash flow.

How to read the replies

The five questions above are not hard for an agency doing real work. None of them require preparation.

A good reply answers all five, in one email, quickly, and volunteers a limitation somewhere. Real work has constraints, and an agency comfortable with its own work will name them without being pushed.

A bad reply substitutes. It offers a meeting instead of an answer, a case study instead of a process, or a rankings chart instead of a lead count. Watch particularly for a reply that answers three of the five and quietly ignores the two about ownership and notice period.

A disqualifying reply guarantees rankings. More on that below.

Judge the behaviour, not the claims

Three situations separate good from bad more reliably than anything technical, and none of them require you to know what a canonical tag is.

When they explain what they will do. A good agency names specific things: "Your primary category is set to Contractor rather than Roofing Contractor, which is costing you relevance on the searches you want. That is the first fix." A bad one deals in categories: "We'll optimise your online presence and build authority." The difference is whether the sentence contains something you could go and verify.

When results are bad. A good agency tells you early, with a diagnosis and a changed plan, because real work has bad months. A bad one never reports a setback. Twelve months of uniformly positive reporting is not a sign of excellence; it is a sign that the reporting is a marketing document.

This is the single most reliable indicator there is, because it is the one thing a bad agency structurally cannot do.

When your business is a bad fit. A good agency says so. There are real cases where local SEO is the wrong instrument — you need revenue in sixty days, your job value is too low to support the fee, nobody searches for what you sell, or your actual problem is that calls go unanswered. An agency that has never turned work down is optimising for its pipeline rather than your outcome.

Red flags, sorted by severity

If every warning sign is treated as disqualifying you will rule out the entire market, including the people who would have done good work. So these are sorted.

Disqualifying — walk away

Guaranteed rankings. Nobody controls Google's results. A guarantee is therefore either meaningless, or it will be satisfied in a way that costs nothing to achieve and earns nothing either — by ranking you for a long, precise phrase nobody types. "Emergency plumber Henderson NV 24 hour same day service" is easy to rank for and has approximately no search volume. The guarantee is met, the report looks excellent, and the phone does not ring. Ask which specific terms are guaranteed and what their monthly search volume is. The question usually ends the conversation for you.

Secrecy about method. "Proprietary techniques we can't disclose" is not a thing in local SEO. The mechanics are public and have been stable for years — profile accuracy, proximity, prominence, relevant content, reviews. Secrecy is either a shield around techniques that will get you penalised, or a shield around the absence of any technique at all.

Refusing to put accounts in your name, and refusing a dated work log — questions 2 and 3 above. A no to either is disqualifying rather than negotiable.

Serious — ask directly, and get it in writing

A twelve-month contract with automatic renewal. Long contracts are common enough that walking away from all of them will narrow your options considerably. But the automatic renewal is the part that costs people money — a further twelve months unless you cancel inside a narrow window you have forgotten about. If you sign a long term, strike the auto-renewal clause. If they will not, that tells you which part of the arrangement mattered to them.

Notice periods over thirty days, or notice by post. A ninety-day notice period is three months of paying an agency you have already decided to leave. Notice-by-registered-post clauses exist to add friction to exactly that moment.

A non-disparagement clause. A clause preventing you from discussing the engagement publicly exists for one reason, and it is not your protection.

Yellow — worth a question, not a verdict

The price is far below everyone else. Below roughly $1,000 a month there is not enough time available in the month to do local SEO properly, which usually means it is not being done properly. But price alone is a weak signal. A cheaper agency that sends a dated work log every month may be doing more real work than an expensive one that sends a chart. Ask about the work log before you draw conclusions from the number.

They cannot name any client. Some clients genuinely require confidentiality, particularly in legal and medical categories. That is legitimate. What is harder to explain is being unable to produce a single reference in any category.

They lead by attacking your last agency. This one turns entirely on specificity. "They bought links — here is the Search Console report showing them" is a diagnosis: checkable, and useful. "They clearly had no idea what they were doing" is a sales technique aimed straight at your frustration. An agency willing to use your last bad experience as leverage is demonstrating how it operates.

The contract clauses above are the ones that cost real money, and they are easiest to catch with the agreement actually in front of you:

What the quotes will not tell you: who actually does the work

This is worth asking outright, and almost nobody does.

Who writes the content, and where are they? There is nothing wrong with subcontracting or with writers overseas — plenty of excellent work is produced that way. What matters is whether anyone who understands your business reads it before it goes live. A page about desert roofing written by someone who has only ever seen hail markets is wrong in ways a Henderson homeowner notices immediately.

Who will you actually speak to? The person selling is frequently not the person doing. Ask who handles your account day to day and whether you can talk to them before signing.

How many clients does that person carry? There is no correct number, but there is a revealing one. Someone carrying forty accounts is doing a checklist. Someone carrying eight might be thinking.

Do they work with your competitors? Not disqualifying — a specialist in your category often knows more than a generalist. But you should know, and you should ask how they handle two clients competing for the same searches, because one of you will lose that trade.

Reading a proposal's silences

The gaps in a proposal are usually more informative than the contents, because the contents were written to impress and the gaps were written by omission.

No baseline. If the proposal does not describe measuring where you currently stand, then in six months there will be nothing to compare against, and any claim about improvement is unfalsifiable.

No mention of your Google Business Profile. For most local service businesses this is where the calls come from. A proposal that jumps straight to content and links is describing a national SEO engagement sold to a local business.

No mention of what happens if it does not work. Every engagement has months where nothing moves. A proposal that contemplates only success is a sales document.

No named first three months. "Ongoing optimization" is not a plan. Ask what specifically happens in month one, and whether they can tell you before they have looked at your account. If they can, it is a template. If they say "I would need to look first," that is the better answer.

Comparing the final two

At some point you will have narrowed it to two that both answered well. A few tiebreakers that actually predict how the engagement goes:

Which one told you something you did not want to hear? An agency that pointed out your job value is too low for the fee, or that your real problem is unanswered calls rather than visibility, has demonstrated it will tell you things later that are inconvenient to tell you. That is worth more than enthusiasm.

Which one asked more questions than it answered? A first conversation dominated by their process is a pitch. One dominated by questions about your close rate, your capacity, and where your customers actually come from is a diagnosis.

Which one was easier to get a straight answer from? How they communicate while trying to win you is the best version you will ever see. It does not improve after signing.

Which one would you find it easier to leave? Deliberately unromantic, and the most predictive question on this page. If the answer is "both, easily," you have chosen well regardless of which you pick.

Also worth doing before you sign: call a reference in a comparable category — not a logo on a website, an actual client, on the phone — and ask whether the agency has ever lost a client, and why. An honest answer describes a specific mismatch. An evasive one describes a difficult client.

What actually protects you

Most of the list above collapses into two structural things. If you get these right, being wrong about the rest is survivable.

You own every account. Then leaving costs you nothing but the notice period.

The notice period is short. Then the agency has to keep re-earning the relationship, which is the only reliable pressure toward honest reporting that exists in this industry.

Everything else — the reporting quality, the work log, the strategy — is easier to fix when leaving is cheap. That is the whole argument for a month-to-month arrangement, and it is why we run one.

The uncomfortable part

We are an SEO agency in Henderson, so treat this page as what it is — advice from an interested party. The defence is that every test above is one we would rather you applied to us than skipped, and that a few of them we would fail if we were not set up for them.

If the local-versus-national question is the one actually holding you up, that comparison includes the three cases where we are the wrong choice. If you have already signed and want to know whether it is working, SEO reporting covers what the monthly document should contain.