What does fair, transparent pricing look like for AEO tools?

Fair pricing for answer engine optimization software has four parts.

1. Published per-plan limits. The price and the allowance for each tier sit on a public page, not in a proposal.
2. Clear units. You know what one unit is: a tracked prompt, a completed audit, an engine, a brand.
3. No demo gate. You can run a first check, or at least see a sample report, before a sales call.
4. Honesty about the ceiling. The vendor states what the subscription cannot buy.

The reasoning is simple. An AEO tool's costs scale with how many answers it collects. Each answer is a call to an AI platform, and some calls carry extra search fees. Anthropic, for example, charges $10 per 1,000 searches on its API web search tool, on top of tokens. A vendor that hides its units is hiding the one thing you need to compare plans. For a wider view of what these tools do, see the guide to AEO tools.

Most AEO and GEO tools price on some mix of the same four units. A transparent page names all of them for every tier.

The engine question matters more than it looks. As of October 2026, Google says AI Overviews and AI Mode may use different models and techniques, so they can show different links for the same query. A plan that lists "Google AI" as one engine may be measuring only one of them.

As a reference point, 99Visibility (the publisher of this site) lists its plans with these units on its pricing page: keywords, audits a month, platforms and brands per tier, starting with a free plan. For typical price bands across the market, see how much AI visibility tools cost per month.

Here is the reality. Engines decide what they retrieve and cite, and no vendor controls that. A fair vendor says so on the pricing page or the sales page.

Things outside any tool's control:

• Citations. A critical survey of 45 GEO studies (arXiv, July 2026) found no technique with a stable, cross-platform causal effect on organic discoverability.
• Rankings inside AI answers. Google's generative AI report in Search Console exposes impressions only. There is no position data to check an "AI rank" against.
• The actual queries. OpenAI says ChatGPT search typically rewrites your question into one or more targeted queries. A tracker sees the prompt it typed, not the search the engine ran.
• Whether search was on. Retrieval is discretionary per query, so any single visibility score averages answers given with and without live web access.

What a price can buy is measurement and diagnosis. That means seeing which engines name you, what facts they get wrong, and whether something on your side blocks them. The fixes with the strongest evidence are access fixes: server-rendered content, per-bot robots.txt permission, a firewall that lets AI crawlers through, and index and snippet eligibility.

Walk away, or ask hard questions, when a plan is priced on outcomes nobody can deliver.

• "Guaranteed citations" or pay-per-citation tiers. The vendor cannot control the engine, so the guarantee is either hollow or hedged in the contract.
• "AI ranking position" as a paid feature. No ground truth exists for it on Google.
• E-E-A-T or domain authority scores. Google says E-E-A-T isn't a specific ranking factor, and no engine computes Domain Authority.
• Schema or llms.txt as the headline deliverable. Ahrefs tracked 1,885 pages that added schema and found AI Overviews citations fell 4.6%. Google says Search ignores llms.txt.
• "Up to 40% more visibility." That figure comes from a lab engine with sources already loaded into context.
• Demo-only pricing for a self-serve product. If the tool runs prompts automatically, there is little reason to hide the price.

Agency retainers that bundle a tool with services deserve the same scrutiny. The guide to a fair scope for AI SEO packages covers what should be in writing.

Run this check on any shortlisted tool.

1. Find the public pricing page. If you can't, ask for limits in writing before any call.
2. Count engine-answers per month. Multiply prompts by engines by runs. Divide the price by that number to compare plans on equal terms.
3. Ask how engines are queried. Is search turned on, and are logged-out sessions used? This decides what the numbers mean.
4. Ask what a finding looks like. You want specific items: a wrong price, a blocked crawler, content that only appears after JavaScript runs.
5. Run a free first check. Several tools offer one without a sales call; this list of free audits with no sales demo is a starting point. The free checker on this site runs up to three buyer questions as a one-off snapshot.
6. Check the exit. If you already pay for a tool, see what to look for when switching, including history export and plan downgrades.

The next move is to pick two or three tools that pass step one, then compare them on step two.

Why do AEO tools price by keywords or prompts? Each prompt the tool tracks becomes one answer per engine per run, and every answer costs the vendor an API call. Some engines add search fees on top. Pricing by prompts ties your bill to that real cost. Ask whether a prompt counts once or once per engine, because that changes your effective allowance a lot.

Is a cheaper AEO tool always worse? No. Price mostly reflects how many answers the tool collects, not how accurate they are. A cheaper plan with fewer prompts can be enough for one brand in one market. What matters more is whether each engine is measured separately, whether search state is disclosed, and whether findings point to a specific fix.

Should an AEO tool require a sales demo? For self-serve tracking, it usually shouldn't. The product runs prompts automatically, so a public price and a free trial or free check are practical. A demo can make sense for enterprise contracts with custom terms. If a small plan is demo-only, ask for the price and limits in writing first.

Can any AEO vendor guarantee my brand appears in ChatGPT? No. OpenAI says ChatGPT ranks results using multiple factors and that placement is not guaranteed. What you control is access: letting OAI-SearchBot crawl your site, making sure your firewall or CDN doesn't block it, and putting key facts in the HTML. A vendor promising placement is promising something OpenAI itself does not.

What should a monthly AEO plan actually deliver? A per-engine view of whether AI answers name your brand for your buyer questions, a list of facts the engines state wrongly, and checks on crawler access and rendering. Each finding should name what is broken and what fixing it involves. A single blended score without that detail is hard to act on.