SEO red flags
Test the SEO pitch before you buy.
Ask what you will be able to inspect.
A proposal offers page-one rankings, hundreds of links and a dashboard full of green arrows. Before comparing its price, find out what those promises mean for the customers you need.
Useful SEO work has an audience, a reason and something you can examine: a corrected technical fault, a stronger service page or an explanation of which searches bring suitable enquiries.
The first useful question
Which problem on our site would you address first, what would you change, and what evidence would show that the work helped?
The provider may need a paid audit before answering in detail. It should still explain the audit’s scope, output and access requirements. A diagnosis takes work; an unexplained promise does not replace it.
Three claims that need a direct answer.
“We guarantee first place.”
Ask which result, query, country and device the promise covers. Is it an organic result or a paid advertisement?
Google warns against ranking guarantees and claims of special access. Advertising does not purchase an organic ranking.
Look for: an achievable work commitment and a clearly described target, with uncertainty acknowledged.
“Our method is secret.”
A provider can protect its internal tools and still explain which pages, settings or external activities it intends to change.
Ask for one example of a proposed action, its purpose and how you approve it. Request an explanation of any material downside.
Look for: enough detail to understand what the business is authorizing and what the finished work contains.
“Hundreds of links, fast.”
Ask where the links would appear, why those sites would cite you and whether placement is paid.
Google’s spam policies address links created to manipulate rankings. A large count does not establish relevance or quality.
Look for: a legitimate reason for each type of placement, with the method and payment arrangements disclosed.
A concrete plan is easier to compare than a promise.
Here is an invented example for a local repair company. It illustrates a proposal you can inspect; it is not a client result or a universal SEO package.
Too vague to evaluate
“We will optimize your website, build authority and improve visibility every month.”
The wording leaves the audience, pages, work and evidence open. Two providers could deliver very different services under that description.
Specific enough to question
“First, review the repair-service pages and their current search data. Identify missing service-area and booking information. Propose revisions to the affected pages, check that the request form works on mobile and report the changes alongside relevant search visits and suitable enquiries.”
You can now ask which pages, who supplies accurate service details, who implements the changes and what is excluded.
Request a delivery schedule for work the provider controls. Ask what could delay it, including your own reviews or development work. Keep that schedule separate from any estimate of when search performance might change.
Make the report explain the work and its effect.
Ask to see an anonymized sample report before hiring. A useful report should let you follow one important page from the change made to the evidence collected and the next decision.
Activity without enough context
“Twenty keywords improved. Forty links acquired. Organic traffic increased.”
That does not tell you whether the searches matter, whether the links are appropriate or whether visitors could become customers. A graph can be accurate and still leave the business question unanswered.
Evidence you can examine
Look for the changed URLs and dates, the reason for each change, comparable search data and the intended customer action. Ask how suitable enquiries or orders are distinguished from general traffic.
The report should explain missing data and competing explanations. It should also identify useful work that failed to produce the expected result.
Keep access to the underlying accounts so you can inspect the same period and filters. Agree on a review date and the decisions it should support: continue, change the priorities or stop work that has not justified itself.
Keep access that survives the contract.
For an initial audit, Google recommends read access to Search Console, not write access. Match later permissions to the agreed work and use individual accounts.
Search Console distinguishes owners, full users and restricted users. A full user can take some actions, so it is not simply another name for read-only access. Keep a verified owner under the business’s control.
Define the handover before starting.
Record who controls the domain, hosting, website administration, reports and created content. Ask what you can export and what needs a continuing licence or subscription.
At the end of the work, review and remove access that is no longer needed. For a former verified Search Console owner, also review verification tokens using Google’s removal instructions; simply removing the user may allow them to verify again.
Distinguish material you own from third-party tools or external placements. You cannot assume ownership of another publisher’s link.
Get these answers in writing before deciding.
- Which audience and pages would you prioritize, and what evidence supports that choice?
- What will you deliver first, who implements it, and what is outside the quoted scope?
- How will you explain proposed links, content and technical changes before they are made?
- Which reports and customer outcomes will we review, and what uncertainty will remain?
- What access do you need, and what content, data and account control will we retain when the work ends?
Compare the answers with the proposal and sample report. If the provider clarifies gaps, you have a more useful basis for a decision. If it continues to substitute certainty or mystery for inspectable work, choose another offer.