An audit that ends in a decision, not a PDF
An SEO audit is a diagnosis of why a site is not performing in search. A useful one is prioritised by impact and ends with a plan that names who does what. A hundred-page report listing every warning a crawler produced is a deliverable rather than a diagnosis.
Audits have become a product sold on length. Two hundred pages, every issue a tool could detect, colour-coded by a severity rating the tool assigned rather than by whether it affects anything. The result is a document that is genuinely thorough and almost never implemented, because nobody reading it can tell which six things actually matter.
The reason audits go unimplemented is that they do not distinguish between findings. A missing meta description on a noindexed page and a site-wide rendering failure appear in the same list, sometimes with the same severity marker. The reader has no basis for choosing, so the document is filed. Six months later the same issues are rediscovered by a different agency selling a similar audit.
When this work makes sense
Questions worth asking any supplier
Is the output prioritised by impact, or by tool severity? Those rank issues very differently.
Does it say who implements each item? A finding with no owner does not get fixed.
Does it cover content and competition, or only technical issues? Most stalled sites are competitively behind, not broken.
Will they walk you through it? An audit handed over without a conversation is usually the last anyone reads of it.
Is there an offer to implement? Not essential, but a provider who only audits has no stake in whether the findings work.
Deliverables
Full technical crawl with indexation and rendering analysis
Content coverage assessment against the searches your market actually performs
Competitive gap analysis — what the ranking sites have that you do not
Internal linking and site architecture review
Backlink profile overview, flagging anything that looks like a liability
Findings ranked by expected impact against effort, not by tool severity
A plan naming each item's owner and sequence
A walkthrough session, because the document alone rarely survives contact with a busy team
Built with
Process
Every audit ends with a walkthrough rather than a delivery. We go through the findings with whoever will implement them, in order of impact, and agree what is being done, by whom, and in what sequence. The document exists to support that conversation rather than to replace it. An audit whose only interaction is an email attachment has a very low implementation rate, and we would rather not produce one.
The standard nine stages — discovery, strategy, UX/UI, development, content and SEO, testing, launch, measurement, continuous improvement — apply to every project. The paragraph above is what differs for this one.
What moves the number
Pricing depends on scope, functionality, integrations and content requirements. These are the factors that change it most:
Automated report, technical audit, or full diagnostic
Automated tool report
Fits: A quick sense of technical health, usually free
Limits: No prioritisation, no competitive context, and a high proportion of noise
Technical audit
Fits: Sites with a suspected technical problem blocking everything else
Limits: Will not explain why you are behind competitively, which is the more common cause
Full diagnostic
Fits: Businesses deciding where to invest and needing the whole picture
Limits: Costs more and takes longer; unnecessary if the problem is already identified
Case studies
No published audit case study. Audit findings are client-confidential by nature and we will not publish sanitised versions presented as evidence.
MBGDesk operates from 4 offices — India, United States, United Kingdom, United Arab Emirates. We hold no industry certifications and do not imply otherwise.
Frequently asked questions
How long does an audit take?
Typically one to three weeks depending on site size and whether competitive analysis is included. Crawling is quick; the analysis and the prioritisation are the work, and rushing those produces the unprioritised list that nobody implements.
Can we just use a free tool?
For a first look, yes, and it will tell you whether something obvious is wrong. What it will not do is tell you which of its two hundred findings matter, or why a competitor outranks you when your technical scores are similar. That judgement is the point of paying for one.
Will the audit tell us what to write?
Content coverage analysis identifies the searches your market performs that you do not currently answer. Whether that becomes briefs or finished content is a separate piece of work, and we say which it is rather than blurring the two.
What if we cannot implement everything?
That is the expected case and the reason for prioritisation. The plan is ordered so that stopping partway still leaves you with the items that mattered most, rather than the ones that happened to come first alphabetically.
Do you audit sites you did not build?
Almost always — that is the usual case. We have no interest in concluding that a rebuild is necessary, and we will say when a site is fundamentally sound and the problem is competitive rather than technical.
What if the audit says our site is fine?
Then it says that, and the conversation moves to content and competition instead. An audit that finds serious problems on every site it examines is selling a conclusion rather than performing a diagnosis.
Services that connect to this
Talk to us about this
Tell us what you are trying to build and we will come back with scope, approach and an estimate.