A strong proposal should let you picture the work happening inside your business. Which service comes first? What changes on the site? Who writes and publishes it? What will you review before committing to more?
If those answers are missing, you are comparing promises instead of plans. For an established consultancy, advisory firm or specialist service business, the useful question is whether the provider can turn your expertise into pages and evidence that help the right buyers choose you.
Use this guide to compare a shortlist, including us. If the scope is already clear, you can request a free Clarity Call directly. You do not need an audit packet or account access to start that conversation.
Still building a shortlist? Compare AI SEO companies by their published work, delivery responsibilities and evidence.
First, decide which responsibility you need to buy
An AI SEO, GEO or AEO label does not tell you who will do the work. Before comparing suppliers, identify the gap in your current team.
| Your situation | What to establish before choosing |
|---|---|
| Your team can make the changes but needs a direction | Whether a defined review and prioritized recommendations are sufficient |
| Your team has a plan but lacks delivery capacity | Who will research, write, implement and check the agreed changes |
| Your existing supplier already covers much of the work | Which responsibility is missing, and who will own it without duplicating effort |
| You mainly need to monitor selected questions | Whether software meets the need and who will act on what it shows |
Our AI SEO plan includes research, writing and implementation within an agreed scope. That is useful when you need the work carried through. If your existing team can resolve the problem, that may be the better first move. For the pricing differences between software, advice and delivery, read the cost guide.
Check whether the evidence matches your decision
A relevant case is more useful than a dramatic percentage without context. Ask to see the starting problem, the changes made, the dates, the measurement source and the result that matters to your business. If several channels or suppliers were involved, ask how that affects the interpretation.
The two readings below use our own work for Xporate, a named client in our existing public case study. This is website-delivery and Google-search evidence. It is not a study of AI citations or a test of another provider. Apply the same scrutiny to us as to your other candidates.
Inspect a delivered change
Our published Xporate case describes an earlier site led by workspace search. The retained before screenshot lets you inspect that presentation. The later navigation groups routes around starting a business, staying compliant and expanding.
On 7 October 2026, we checked the public Xporate homepage, Start a Business page and Compliance Management page. The pages returned successfully; the homepage contained the business-stage routes and the destination pages described the services. This was a page-content and response check, not a test of every interaction or search appearance.
| Inspect | What it establishes | What still needs an answer |
|---|---|---|
| The retained before and after screenshots | The visible presentation differs | Which business inputs justified the chosen routes? |
| The current homepage and destination pages | The documented routes and service explanations were published at our check | Which approval and acceptance checks accompanied publication? |
| Our published project account | Our account of the June 2026 launch and questions with the client | Where are the approved scope and dated change record for the work you are buying? |
The screenshot file path is not a capture date. Our account of the launch is not an independently verified deployment log. These records let you inspect a published change; they do not show how visitors used it or whether enquiries improved.
Use this when comparing a proposal: ask a provider to identify the actual output, its publication state and the checks they performed. A recommendation, an approved draft and a live change are different deliverables. If your team must publish, that responsibility belongs in the scope.
Inspect a reported result
The same case uses Xporate’s retained Google Search Console daily export, supplied on 29 September 2026. The comparison below was recalculated from its daily rows. It describes Google Web Search activity, not assistant recommendations, qualified enquiries or revenue.
| Measure | 1 March to 31 May 2026 | 1 July to 26 September 2026 |
|---|---|---|
| Days with retained daily rows | 92 | 88 |
| Impressions | 42,581 | 100,621 |
| Clicks | 651 | 1,250 |
| Impressions per day, rounded to two decimals | 462.84 | 1,143.42 |
| Clicks per day, rounded to two decimals | 7.08 | 14.20 |
Check the denominator. The periods differ in length. Dividing each total by its own day count makes the daily comparison readable. For example, 42,581 ÷ 92 = 462.84 impressions per day, rounded. Daily averages do not remove seasonality or other influences.
Check what happened before. Impressions were already rising before the June launch. The later window records more search activity, but this comparison does not isolate how much the rebuild caused. June is excluded because our project account identifies it as the launch month; it is not a controlled comparison.
Check what is missing. This retained record does not establish buyer fit, enquiries, sales or AI citations. The complete original extraction request is not retained, so we cannot independently verify every country or device filter. The public case displays the rounded readings; the underlying export is held in our evidence record.
Use this when comparing a proposal: ask which changes, demand shifts and other activity occurred between the periods, and what evidence connects search activity to qualified enquiries. A provider can have delivered useful work while the commercial effect remains unproven.
These examples help you assess the quality and limits of the evidence. They do not predict your result. For collection and reporting definitions, use the visibility and enquiry measurement guide. For delivery and review dates, see the AI visibility timeline.
Google’s hiring guidance recommends checking previous work, relevant experience, expected results, measurement and explanations of changes. Put those answers into the worksheet below, including unanswered questions.
Your proposal comparison worksheet
Put two proposals beside each other. Ask each provider to answer the same questions in writing, then record the answer and its source. A blank or vague answer is a follow-up question, not an automatic rejection. Look for a scope your business can use and sustain.
The fields below are for your own notes. Type directly into them, then use your browser’s Print command to print or save the worksheet. Notes are not sent to us or saved by this page; reloading or closing it can lose them. No form or email address is required.
Your working copy
Compare AI visibility proposals
Use the same questions for every provider, including SEO Informatica. Record the answer and a source reference. No automatic score.
What still needs an answer?
Share the right information at the right stage
You should not need to assemble a data room to make an enquiry. Start with your website, the service or market that matters most, and the decision you are trying to make.
| Stage | Useful inputs |
|---|---|
| Public first look or initial conversation | Website address, business context, priority and any known constraints. Our Outside View uses public information. |
| Agreed investigation | Relevant reporting access, the current scope, known site issues and evidence needed to answer the agreed questions |
| Implementation | Named people who can edit, check and publish; approved business facts, proof permissions and the access needed for each change |
For an ordinary reporting review, Search Console Restricted supports Performance data and limited URL inspection. GA4 Viewer supports data review. Analyst adds shared explorations; configuration work may require another role, depending on the task. Full Search Console access is not read-only.
In Tag Manager, Read, Edit, Approve and Publish are separate permissions. CMS and hosting access should follow the same principle: agree the change and give the access required to carry it out. Keep account ownership with your business. If your developer will publish the work, write that responsibility into the scope.
Choose the first conversation from what you need
If you want a written first look at how your business presents itself, request your free Outside View. We send how you look from the outside, who really competes with you and the ten fixes that matter, within two business days.
If you already know the problem or want to discuss a proposal, request your free Clarity Call. Bring your priority and the unanswered questions from the worksheet. We can establish the goal, work, responsibilities and price before you decide. An Outside View is optional.