An Independent Look at Nir Levi's Search and AI Visibility Consulting
Nir Levi's consulting practice occupies an interesting position in search marketing. It is not a software subscription, and it does not operate on a fixed package menu. The Haifa-based consultant has worked in search since 2008, founded Godrank in 2015 and Madrank in 2018, and sells a scoped investigation that ends in tasks, briefs, and priorities a client team can execute. Across the collected pages, the emphasis is consistent: agree on a business question, inspect the available evidence, and make the findings usable.
What the practice actually covers
The consulting work spans four connected areas: SEO strategy, technical SEO, GEO and AEO for AI visibility, and content strategy. The distinction matters because each area starts from a different failure. SEO strategy work typically examines search demand, competitors, and missed opportunities, then builds a roadmap. Technical SEO reviews look at crawling, indexing, rendering, canonical URLs, internal links, and multilingual site structure, with recommendations aimed at developers. GEO and AEO work focuses on how a company appears in AI answers, what sources those answers cite, and whether the underlying website information is accessible. Content strategy covers which pages to create or improve and produces briefs explaining the intended audience and questions.
| Service area | Typical focus | Stated outputs |
|---|---|---|
| SEO consulting and strategy | Search demand, competitors, opportunities | Findings, roadmap, priorities |
| Technical SEO | Indexing, rendering, canonicals, site structure | Developer recommendations |
| GEO and AEO | AI answers, citations, factual accuracy, content access | Answer logs, content findings, task list |
| Content strategy | Customer questions, page selection | Page briefs or rewrites |
This is not a software dashboard. The deliverable is human judgment backed by evidence: a consultant investigates the situation, records what was found, and translates it into work that product, content, and engineering teams can act on.
How an engagement is structured
The process described on the site starts before any work is scoped. A company states its website, goals, and the question it wants answered. Levi and the client then agree on the scope, deliverables, languages, platforms, and price. After that, the work can include collecting evidence, reviewing pages and sources, and preparing the findings. The site also says clients work directly with Levi rather than through junior account managers.
That structure has a practical consequence. Because responsibilities are agreed in advance, the service can fit teams that already have an SEO function but need a second opinion or a focused review. It can also fit teams that are not sure whether to start with technical SEO, content planning, or AI visibility. The initial conversation is intended to determine that entry point rather than requiring the buyer to know the exact service name.
AEO is narrower than GEO in this practice
The distinction between AEO and GEO is easy to blur, but the site keeps a useful boundary. AEO work, as described on the AEO expert page, concentrates on questions and answers on the client's own pages: whether a product page explains who the product is for, whether pricing information stands on its own, and whether Schema markup matches the visible content. GEO work extends into external sources and broader visibility: it examines review portals, comparison content, publications, and the domains that AI systems actually cite.
The research published on the site reinforces that external focus. In a small 24-answer Perplexity panel collected on 11 October 2026, tracked brands appeared in all 24 category answers, but official-domain links appeared in 22. The average cross-language source-host overlap between Hebrew and English answers was 34.7%. Even same-language repeats averaged only 47.2% source overlap. Levi's interpretation is appropriately cautious: a mention is not the same as a citation, a citation is not a recommendation, and none of these observations proves that a website change caused an answer to change. That framing is more useful than the usual AI visibility pitch.
Free tools and an unusual research habit
For an individual consultancy, the public site has an unusual amount of original material. There is a free AI visibility checker for ChatGPT and Gemini, plus a no-signup CSV worksheet for manual checks. The worksheet separates mentions, citations, recommendations, and accuracy rather than collapsing them into a single visibility score. That is a small but meaningful design choice because it pushes teams to collect the underlying record instead of treating AI visibility as one number.
The site also publishes worked examples, including a B2B keyword research walkthrough using Ahrefs data and search-result observations, and a guide that checks eight claims from the Google leak against their original sources. One published article about a 431% reported traffic increase from digital PR is explicitly labeled as an analysis of a third-party case study, with the caveat that client analytics were not independently audited. These are not anonymous marketing posts. They show the method: source the evidence, note the limitations, and separate what is supported from what still needs checking.
Pricing is not published
No collected page lists a fixed fee, currency, subscription plan, or billing period for the consulting itself. The site states that cost depends on the review scope and the support required, and that a proposal covering work, deliverables, and schedule follows the initial discussion. The GEO and AEO service page repeats that languages, platforms, scope, and price are agreed before work begins.
For a buyer comparing options, this means there is no self-serve package to preview. It also means the quote is tied to a defined review rather than a generic monthly retainer. The free AI visibility checker and worksheet are exceptions: both are presented as free, with the worksheet requiring no signup.
Who the practice is aimed at
The published pages name SaaS companies, startups, and large organizations as the main audience. The selected client experience list includes monday.com and ClickCease in B2B SaaS, Casino.com, 10Bet, and MansionBet in regulated iGaming, Sensica in medical aesthetics, and T·TIME in digital health. The multilingual angle is a recurring theme, with work available in Hebrew and English and discussion of Hebrew-English source divergence.
That client mix suggests the service is not aimed primarily at small local sites looking for a cheap one-off fix. It fits organizations that already have a product or content team, a site large enough to create technical trade-offs, or a market where AI answers and multilingual demand create measurable but difficult visibility questions. The site is also clear that completing a review does not guarantee a search ranking or an AI citation. That limitation is built into the process: the aim is to produce inspectable findings and a plan the team can use, not to promise a specific search outcome.
What stands out after reviewing the practice
The strongest thread across the site is the separation of an observation from a conclusion. A brand mention is not a citation. A citation is not a recommendation. A search-volume estimate is not confirmed demand. A traffic increase reported by a third party is not an audited result. This is not dry caution for its own sake; it directly affects what a client is asked to do next. Instead of reacting to a single ChatGPT answer or a ranking movement, the engagement is meant to produce a recorded baseline, a defined set of checks, and a prioritized list of tasks with owners and verification steps.
That approach has limits. Teams wanting a fixed package price, standardized deliverables without scoping calls, or guaranteed implementation may find the model harder to evaluate from the public pages. But for organizations that already know they have a search or AI visibility problem and need a credible way to turn it into internal work, the practice is built around a coherent method rather than a flashy promise.