How to Measure AI Search Demand With the AI Optimization API

Your audience is no longer starting every search in Google. More and more of them are relying on an AI assistant like ChatGPT, Perplexity, and Google’s AI Overviews feature to get answers to the questions they used to type into a search box. That search demand is real, but most keyword tools don’t show it. Those tools measure searches in traditional engines, not what people ask AI. DataForSEO’s AI Optimization API gives you the missing view through a metric called AI Search Volume. This post shows how to measure AI search demand with it, and how to use that data in your SEO, GEO and content strategy.

AI search is creating demand your keyword tools can’t see

People tend to ask AI assistants longer, more detailed questions than they type into a search engine. Compare “best tie brands” with “what are the best tie brands for men.” Both describe the same intent and topic, but the second query is longer, more specific, and more conversational. That’s how people naturally phrase questions in AI tools.

This matters because traditional search volume and AI Search Volume measure different behaviors and different search environments. A keyword can show modest volume in traditional search and strong demand in AI tools, or the other way around. The two surfaces don’t move together. To understand that distinction and make sure you don’t miss real opportunities in AI search, you need a metric that measures AI demand directly. That is what AI Search Volume gives you.

What “AI search demand” actually measures

AI Search Volume is the estimated frequency with which a keyword is used in AI search environments. The value is calculated based on our proprietary algorithm that considers multiple signals, including data on the “People Also Ask” section of Google search results from DataForSEO’s extensive SERP index.

For each keyword, you get last month’s AI Search Volume and a 12-month trend. For example, “best tie brands” returns an ai_search_volume of 412 for the latest month. The strongest demand occurred in August–September 2025, peaking at 641 in September, and declining from October to January 2026, with the weakest point at 307. From January to May 2026, the keyword shows renewed interest after the winter dip. You can see the full response array example below.

For the full definition and the calculation nuances, see our related article about the AI Search Volume metric. If you want to get these numbers for your own keywords, that takes just one request.

How to pull AI Search Volume in one request

The AI Keyword Search Volume endpoint reads your list of keywords and returns AI Search Volume for each one. It runs in Live mode, so one request returns results with no separate POST and GET steps, and you can send up to 1,000 keywords at a time. The request body and an example response look like this:

[
  {
    "language_code": "en",
    "location_code": 2840,
    "keywords": ["best tie brands"]
  }
]

The response returns ai_search_volume for last month plus an ai_monthly_searches array holding the 12-month trend:

{
    "version": "0.1.20260610",
    "status_code": 20000,
    "status_message": "Ok.",
    "time": "0.1793 sec.",
    "cost": 0.0101,
    "tasks_count": 1,
    "tasks_error": 0,
    "tasks": [
        {
            "id": "06261220-8284-0619-0000-bcca7c0a6b25",
            "status_code": 20000,
            "status_message": "Ok.",
            "time": "0.1171 sec.",
            "cost": 0.0101,
            "result_count": 1,
            "path": [...],
            "data": {...},
            "result": [
                {
                    "location_code": 2840,
                    "language_code": "en",
                    "items_count": 1,
                    "items": [
                        {
                            "keyword": "best tie brands",
                            "ai_search_volume": 412,
                            "ai_monthly_searches": [
                                { "year": 2026, "month": 5, "ai_search_volume": 412},
                                { "year": 2026, "month": 4, "ai_search_volume": 444},
                                { "year": 2026, "month": 3, "ai_search_volume": 442},
                                { "year": 2026, "month": 2, "ai_search_volume": 311},
                                { "year": 2026, "month": 1, "ai_search_volume": 307},
                                { "year": 2025, "month": 12, "ai_search_volume": 349},
                                { "year": 2025, "month": 11, "ai_search_volume": 384},
                                { "year": 2025, "month": 10, "ai_search_volume": 511},
                                { "year": 2025, "month": 9, "ai_search_volume": 641},
                                { "year": 2025, "month": 8, "ai_search_volume": 629},
                                { "year": 2025, "month": 7, "ai_search_volume": 516},
                                { "year": 2025, "month": 6, "ai_search_volume": 404}
                            ]
                        }
                    ]
                }
            ]
        }
    ]
}

For the full parameter list, see the AI Keyword Search Volume endpoint docs. You can test it in our free Sandbox before spending a credit.

The three sections below show you how to put AI Search Volume data to work.

Use case 1 — Prioritize keywords for AI search optimization (GEO)

The most direct use is to rank your keyword list by ai_search_volume and spend your GEO effort where AI demand is highest. Take a list you already have (a content plan, a Search Console export, last quarter’s keyword research), send it to the endpoint, and sort by ai_search_volume, descending. The keywords at the top are the questions people ask AI tools most often, so those are the ones to optimize for first across ChatGPT, Perplexity, and Google AI Overviews.

Volume alone can mislead, though, so read it next to the trend. When a keyword’s ai_monthly_searches climb month over month, that signals growing AI interest, often a better bet than a higher-volume term that has gone flat. Rising terms let you publish into demand that’s still building instead of catching a topic on its way down. The trend data deserves a closer look on its own.

Use case 2 — Track AI search demand trends & seasonality

Knowing a keyword’s demand today doesn’t tell you where it’s headed. Publish against a topic that has already peaked and you show up late. Miss a seasonal swing and you ship your best content in the wrong month. The 12-month ai_monthly_searches array shows you the shape of demand, so you can avoid both issues.

Take a look at the “best tie brands” again. Its AI Search Volume sits near 400 in late spring, but the trend tells the fuller story. The demand pattern is not flat: it builds through summer, peaks in September, then has a winter drop, and an early spring recovery. The trend suggests this keyword likely performs best around late summer / early fall, possibly tied to wedding season, back-to-office shopping, fall fashion, or formal event planning. Read that data shape and the plan writes itself: get your GEO content indexed and refined before the late-summer / early-autumn rise, not after demand has already peaked. So, level tells you how big the keyword is in AI search right now, trend tells you when to act. The next question is how this compares with traditional search volume, where the same keyword may follow a different curve.

Use case 3 — Compare AI Search Volume vs. traditional Search Volume

On its own, AI Search Volume can’t tell you whether a topic is an AI-specific opportunity or simply popular everywhere. Without a traditional search baseline, you can’t see where AI demand and Google demand split apart, and that split is where the opportunities may sit. Pair the AI Keyword Search Volume endpoint with our Keywords Data API (Google Ads search volume) on the same keyword set, and two patterns show up. Keywords strong in AI but weak in traditional search point to topics for which users want explanation, comparison, or recommendation, so content targeting these opportunities needs to give clear definitions, ranked options, pros and cons, use cases, and trustworthy supporting details.

Keywords strong in traditional search but weak in AI suggest users may be looking for a specific page, brand, product, local result, or quick transactional action rather than a conversational answer, so you should not force an AI-first strategy. Start with strong SEO performance, then add AI-friendly elements like concise summaries, entity-rich explanations, comparison sections, or FAQs, where they make sense.

The example below illustrates how a combination of data can reveal whether a topic is AI-leaning, traditional-search leaning, or strong in both.

The comparison becomes useful when you look at the relationship between the two numbers, not either metric in isolation. A high AI-to-Google ratio marks a topic users are more likely to explore through AI answers, which makes it worth GEO investment. A low AI-to-Google ratio marks a topic that still lives mostly in traditional search. Before you trust any single number, though, it helps to know how the metric behaves.

How to read AI Search Volume accurately (and how to use it well)

In practice, three habits keep your reading of the data sound:

• It’s an index of conversational demand, not a raw count of AI queries. ai_search_volume is derived from our proprietary algorithm, not actual logs from ChatGPT, Perplexity, or Google AI Mode. Treat it as a relative popularity indicator across keywords and time, not a precise query count from any single AI tool.
• Grammatical variants collapse into one keyword. Our algorithm treats different grammatical forms of the same word as one, so ai_search_volume for “tie” and “ties” will be identical. Don’t expect to differentiate singular vs. plural or tense variants with this metric — send root or concept terms instead of over-splitting variants.
• Multi-word phrases require all words present. For a phrase like “running shoes sale,” data is only counted from AI questions that contain all specified words. Add non-essential modifiers and you risk shrinking the matching set (and the volume) toward zero — so keep phrases to their core intent terms.

The takeaway: treat AI Search Volume as a directional prioritization signal, and always consider the context: seasonality, intent, topic type. Used that way, this metric will point your GEO efforts at demand you can verify rather than topics you assume are popular.

Measure the demand, then act on it

AI search environments carry measurable demand that traditional keyword tools don’t show, and AI Search Volume gives that demand a number. One endpoint lets you rank keywords for GEO, read the 12-month trend, and compare AI demand with Google search volume, all from data instead of guesswork.

Measuring demand is the first step. Once you know which topics matter, the next move is to track brand mentions across LLMs and see whether you’re showing up in the answers people get.

Try AI Search Volume for free