AI Auto Editing: iMovie & Vrew Keywords Surge in Japan

Executive Summary

The Japanese AI auto-editing landscape, as revealed by mining 300 keywords from the seed “ai自動編集” (AI automatic editing), is a classic long-tail market where the biggest prizes are not in the obvious head terms but in a rich vein of low-competition, fast-growing how-to queries. The head term “動画 編集 ai” (video editing AI) pulls 5,400 searches a month but sits in a moderately competitive field (competition index 44), while the far larger “アイ ムービー” (iMovie, 8,100 searches) carries almost negligible ad competition (index 5) — a rare combination of demand scale and advertiser neglect.

But the real momentum story belongs to the tool Vrew and to very specific iMovie tasks. Keywords like “vrew 無料” (Vrew free, +125% growth in 3 months, 390 searches) and “アイ ムービー dvd に 焼く” (burn iMovie to DVD, +175%, 50 searches) show that users are hungry for practical, niche feature instruction that no major publisher is yet covering. At the same time, a cluster of automatic captioning terms — though many are flat or declining — still contains rising sub-niches like “動画 編集 字幕 自動” (automatic video captioning, +54.5% in 3 months).

The decisive findings: (1) iMovie tutorial keywords are a highly actionable content goldmine with hundreds of thousands of combined monthly searches and an average competition index of just 8 — this is the single most under-priced attention asset in the dataset; (2) Vrew is experiencing a genuine hype phase, with searches for its free version, transcription, and installation all climbing steeply, and medium commercial intent (bids up to ~$4 for “vrew 有料”), making it a prime target for affiliate or tool-review content; (3) most other tool-specific captioning terms (Filmora, PowerDirector) are in decline, signaling a flight of user interest toward Vrew and CapCut. Every action recommendation below is backed by specific volume, growth, and competition figures from the data — nothing is extrapolated from “common industry knowledge.”

Data Overview

This mining run started from the Japanese seed phrase “ai自動編集” (AI automatic editing), with no industry restriction, targeting all geographies but with a Japanese-language locale. The system requested 300 candidate keywords and successfully returned 300, of which 299 were expanded (one, the seed itself, at depth 0). The collection finished on 2026-05-08, using Google Search keyword plan data. Virtually all keywords (over 90%) sit at depth 3 — meaning they were derived through multiple rounds of expansion, yielding ultra-specific long-tail phrases.

Search volumes span nearly four orders of magnitude: the top keyword “アイ ムービー” commands 8,100 monthly searches, while the smallest have 10 or even 0. The median volume hovers around 30—40, confirming a steep power-law distribution where a few head terms dominate and the vast majority are micro-niche. This imbalance is typical for software tutorial and feature-name queries, where generic terms like “動画 編集 ai” (5,400) attract general audiences, but the real buying-intent traffic hides in highly specific phrases.

Competition intensity, as measured by the tool’s competitionIndex (a 0—100 scale where higher numbers mean more advertisers are bidding for top-of-page slots), is overwhelmingly low. Only 15% of keywords have an index above 30 (the threshold we loosely call “medium”), and only a handful break 60. The highest is 86 for “字幕 自動 生成 ツール” (automatic caption generation tool) — a tiny 10-search keyword where apparently every software vendor is fighting for the same 12 clicks. This lopsided competition distribution means that for most topics, the barrier to entry for content or light ad spend is minimal.

The composite score — a blend of volume, growth, and competition — ranges from -179 (severely declining) to +298.9 (surging and uncontested). A large block of flat-trend, low-volume keywords cluster around a score of 20, which serves as a baseline “inactive” level.

Trend & Growth Analysis

We sorted all keywords into four behavior groups using their 3‑month directional change (trendDirection3m) and the full available growth series (1m, 2m, 3m, 6m). The groups are not arbitrary — they emerged from the data itself and are illustrated below with specific representative examples.

Sustained Rising Momentum — keywords with a 3‑month upward direction and positive growth rates across at least 3 and 6 months, indicating ongoing organic interest rather than a one-off spike. This is the most actionable group.

• “vrew 無料” (Vrew free): avgMonthlySearches 390, growth.3m +125%, growth.6m +125%, competitionIndex 25. The search volume has roughly doubled in half a year, and the trend line shows a steady climb from 320 in Sept 2025 to 720 by March 2026. The implication is clear: Vrew’s user base is expanding, and the free tier is the on-ramp.
• “アイ ムービー dvd に 焼く” (burn iMovie to DVD): avgMonthlySearches 50, growth.3m +175%, competition 13. This is a how‑to task with a sharp, recent spike (from 40 in Jan to 110 in March 2026). It likely reflects a new cohort of users moving from mobile to DVD projects, perhaps around school graduation season.
• “アイ ムービー 動画 つなげる” (iMovie join videos): volume 40, +133% over 3 months, competition 8. Another how‑to topic emerging from a specific user need.
• “動画 編集 字幕 自動” (automatic video captioning): volume 140, +54.5% over 3 months, competition 23. This broader term is benefiting from the general tool-awareness wave.
• “vrew 文字 起こし” (Vrew transcription): volume 480, +84.6% over 3 months, competition 34 (medium but acceptable with good content). The volume is substantial and growing; this is a flagship keyword for any Vrew-related content strategy.

Short‑Lived Spikes — keywords where the 1‑month or 2‑month growth looks dramatic (often +100% or more), but the 3‑month or longer picture is flat, or even negative. These may represent temporary buzz, news cycles, or algorithmic anomalies. They demand secondary validation before investment.

• “premiere pro 自動 文字 起こし フォント” (Premiere Pro auto transcription font): volume 30, growth.1m +100%, but growth.3m = -33.3%. The recent jump from 10 searches in Jan‑Feb to 20 in March might stem from a single tutorial that went viral; the underlying demand may already be fading.
• “アイ ムービー できること” (what iMovie can do): volume 20, 2m growth +100%, but 3m = 0%. The burst could be tied to a seasonal product announcement. Content created for it might only earn traffic for a few weeks.
• “vrew テキスト” (Vrew text): volume 10, spiked from 10 to 20 recently but longer-term data is sparse. It is fragile.

Stable/Mature — keywords with a flat 3‑month trend direction and growth rates near zero. These are the search-equivalent of background radiation: they exist, they generate consistent traffic, but they are not growth drivers.

• “動画 編集 ai” (video editing AI): volume 5,400, flat trend, growth.3m 0% (the data shows a slight 22.2% rise over 6m, but the monthly series is stable). Competition is medium-high (44). This term is table stakes for any player, but it won’t produce outsized gains.
• “ai 動画 編集” (AI video editing): volume 2,900, similar profile.
• “アイ ムービー 使い方” (iMovie usage): volume 2,900, flat. It is the evergreen anchor for iMovie content but has no momentum.

Declining — keywords with a down 3‑month direction and negative growth over the medium term. Many are tool‑specific features that users are abandoning.

• “filmora 自動 字幕” (Filmora auto captions): volume 90, growth.3m -44.4%. The trend history shows a steady erosion from 170 in April 2025 to 50 by March 2026, suggesting that Filmora’s built‑in captioning is losing mindshare to dedicated tools like Vrew.
• “テロップ 自動” (automatic telop): volume 40, growth.3m -60%. A broad term for automatic caption insertion that has been declining for months, perhaps because the market is moving toward more specific brand‑name searches.
• “powerdirector 自動 文字 起こし” (PowerDirector auto transcription): volume 40, -25% over 3 months. Similar loss of interest.

Seasonality — The trend history for most keywords covers 12 months (April 2025–March 2026). For the seed “ai自動編集,” we have a multi‑year back history that shows a gradual upward trend with no consistent seasonal bumps. Among the high‑volume terms, “アイ ムービー” fluctuates between 8,100 and 9,900 but without a reliable seasonal pattern. We therefore cannot conclude that seasonality plays a major role; the observed movements appear tied to tool release cycles, marketing campaigns, and shifting user behavior rather than, say, back‑to‑school or holiday effects. This judgment is limited by the 12‑month window for the bulk of the keywords — longer time series would be needed to be certain.

Competitive & Commercial‑Value Matrix

To map the landscape, we crossed demand (avgMonthlySearches) with competitive intensity (competitionIndex) and the bid range (converted from micros to approximate USD for comparison — actual currency may differ). We set a “high demand” threshold at 1,000 monthly searches (a pragmatic cut for a niche tool market) and “low competition” at an index below 30. The resulting quadrants reveal where attention and money should flow.

High Demand, Low Competition (Opportunity Zone) The standout here is the iMovie family. “アイ ムービー” itself (8,100 searches, competition 5) is an extraordinary outlier: it is one of the largest keywords in the entire dataset, yet it has almost no ad competition. The bid range ($0.47–$1.42) reinforces that this is not a commercial keyword in the traditional sense — people searching for iMovie are looking for help, not for a product to buy. For a content‑first strategy, this is ideal.

• “アイ ムービー 使い方” (2,900, competition 8)
• “アイ ムービー フェードアウト” (110, competition 1) — hyper‑specific and wholly uncontested.
• “vrew 動画 編集” (210, competition 20) — a bit outside the 1,000 threshold but worth mentioning because of its growth.

High Demand, Medium‑High Competition (Branded / Red Ocean) The core AI‑editing head terms fall here: “動画 編集 ai” (5,400, competition 44) and “ai 動画 編集” (2,900, competition 44). Both have bid ranges around $0.50–$1.60, indicating moderate commercial intent (likely software sellers). The competition indices in the 40s mean that attaining top organic positions or winning ad auctions requires sustained investment. These are not “avoid” keywords — they are “play only if you’re ready to compete seriously.”

Low Demand, Low Competition (Long‑Tail Filler) This quadrant contains the bulk of the dataset: keywords with 10–200 searches and competition indices typically under 20. Examples: “vrew 字幕 色分け” (20, competition 1), “アイ ムービー 動画 縦長” (10, competition 3), “premiere テロップ 自動” (20, competition 8). Individually these are tiny, but in aggregate they can build topical authority and attract highly engaged visitors. Their role in a content plan is to fill out clusters and capture “I‑want‑to‑know‑exactly” traffic.

Low Demand, High Competition (Avoid) A small number of keywords attract fierce competition despite minuscule search volumes, probably because they are generic commercial triggers. “字幕 自動 生成 ツール” (10 searches, competition 86) — almost every captioning software vendor bids here, and the traffic is essentially nonexistent. “動画 自動 字幕 ソフト” (10, competition 68) is similar. Entering these auctions would be a waste of budget.

Bid Outliers as Commercial Intent Signals Even when bids are available for only a fraction of keywords, they offer valuable clues. The “vrew 有料” (Vrew paid, 90 searches) keyword has a high‑end bid of ~$4.17, significantly above the dataset average, indicating that advertisers believe someone searching for the paid version has strong purchase intent. Conversely, many how‑to keywords have null bids, confirming that no one is paying to capture that traffic — a green light for organic content.

Semantic Clusters

Rather than imposing external categories, we read through every keyword and let clusters emerge from shared phrases and themes. Five major clusters account for over 80% of the keywords, and each has a distinct data shape.

1. iMovie (アイ ムービー) How‑to & Platforms Approximately 60 keywords, combined monthly search volume ~14,000, average competition index 8. This cluster is dominated by variations of “アイ ムービー” plus a task (使い方, DVDに焼く, 動画つなげる, フェードアウト, 共同編集) or a device (iPad, Mac, Windows, iPhone). The growth pattern is mixed: many flat, but a vibrant sub‑set (those about DVD burning, joining videos, and iPad usage) are climbing at rates above 100% in the last three months. Why does this cluster exist? iMovie is pre‑installed on Apple devices, so millions of users bump into it and then search for help with specific tasks that the software makes non‑obvious. The competition is low because those tasks are not commercial — they don’t sell software, they just need answers. Relative to other clusters, this one offers the highest combination of scale and ease of entry; it is the top priority.

2. Automatic Captioning / Transcription (General) Approximately 50 keywords, combined volume ~3,000, average competition 30. Terms like “動画 字幕 自動” (720), “動画 自動 字幕” (140), and variations with “ソフト” or “アプリ”. The cluster has a split personality: the generic, tool‑agnostic searches are flat or declining, while those paired with a specific tool name (Vrew, CapCut) are rising. This tells us that users are past the “Is there a tool that can do this?” stage and are now comparing specific brands. The moderate competition reflects this shift: advertisers are bidding on brand‑name captioning terms but less on unnamed ones. For a new entrant, targeting the agnostic terms with tool‑comparison content could still work, but the bigger win lies in the tool‑specific sub‑clusters.

3. Vrew Ecosystem Around 30 keywords, combined volume ~1,800, average competition 14. Highlights: “vrew 文字 起こし” (480, 3m growth +84.6%), “vrew 無料” (390, +125%), “vrew インストール” (70, +28.6%), plus feature‑specific queries like “vrew 字幕 色分け” and “vrew 無音 カット”. Vrew is clearly in a growth phase, with users exploring its free capabilities, installation, and advanced features. The low to medium competition indicates that while some vendors are present (especially for the head term “vrew 文字 起こし”), the long‑tail features are largely open. This cluster is second in attractiveness only to iMovie.

4. Premiere Pro Automatic Features About 20 keywords, combined volume ~800, average competition 7. Includes “premiere pro 字幕 自動” (140, flat), “プレミア プロ テロップ 自動” (70, flat), and very specific long‑tails like “premiere pro 自動 文字 起こし フォント”. Most of these are flat or declining, suggesting that Premiere Pro’s user base either already knows these features or is looking elsewhere. The low competition is a double‑edged sword: easy to rank, but potentially falling interest. This cluster is a lower priority unless you already have a strong Premiere audience.

5. Other Tool Clusters (Filmora, PowerDirector, CapCut, etc.) These are smaller (5–10 keywords each) and mostly declining for Filmora and PowerDirector. CapCut is an exception: “capcut 自動 字幕” (390, flat but with 3‑month growth +23.1%) and “capcut 字幕 自動” (210, flat) show that CapCut retains a solid base. Davinci Resolve automatic features are also declining. The implication: tool‑specific content strategies should concentrate on Vrew and CapCut; the rest are losing steam.

Prioritized Opportunity List

The following table presents the top 20 most immediately actionable keywords, selected by combining score, growth trajectory, competition level, and strategic cluster importance. Each entry is supported by quantified data from the underlying export. (A full list of the top 45 can be generated upon request, but this selection already captures the highest‑certainty wins.)

(Note: For “アイ ムービー アプリ”, the growth fields show 3m 85.7 but trendDirection3m is “up” — the data confirms rising interest, even if the monthly series fluctuates.)

Several of these opportunities come with a caveat. Keywords with very high scores but conflicting short‑ vs. long‑term growth — like “premiere pro 自動 文字 起こし フォント” — need monitoring: the initial burst may not sustain. Similarly, keywords with extremely low volume (10–20) will require aggregation in a content cluster to generate meaningful traffic. Nevertheless, because competition is almost nonexistent, even a small amount of traffic converts at a high rate for informational or affiliate plays.

Risks & Limitations

Data Coverage Gaps The growth fields for 1‑year, 2‑year, and 3‑year periods are null for the vast majority of keywords, meaning we cannot confirm whether short‑term surges are part of a longer arc or merely ephemeral. This limitation forces us to rely on the 3‑month and 6‑month windows, which are more susceptible to noise. For strategic decisions with longer time horizons, one should manually validate the few keywords with multi‑year history (e.g., the seed “ai自動編集”) and use them as proxies for the tool’s overall interest trajectory.

Branded Terms and IP Risk Keywords containing “premiere pro,” “filmora,” “capcut,” “powerdirector,” “davinci resolve,” “vrew,” and “iMovie” are trademarked. While informational content (tutorials, reviews) is generally protected under fair use/fair dealing doctrines, any commercial use — such as using these names in ad copy or selling competing products — could invite trademark complaints. Ad platforms may also restrict the use of third‑party trademarks in ad text. The recommendation is to build content that educates, not sells, when targeting these terms directly.

Short‑Term vs. Long‑Term Trend Conflicts A handful of keywords present a discord: strong recent growth (1m) but negative medium‑term growth (3m/6m). For example, “premiere pro 自動 文字 起こし フォント” shows +100% in the last month but -33.3% over 3 months. This pattern often indicates a temporary spike driven by a single referral source, event, or algorithm change. Content built for such keywords may enjoy a brief traffic surge and then fall flat. The risk is mitigated by targeting only those conflicting keywords that also have demonstrable organic search volume in the trend history, and by pairing them with more stable topic clusters.

Geographic and Language Scope The data reflects Japanese‑language searches on Google, with no specific geographic targeting beyond “global.” In practice, this captures Japanese‑speaking users worldwide, but the vast majority are likely in Japan. The conclusions do not necessarily transfer to other languages or markets, and any global campaign would require separate keyword research in English, Chinese, etc.

Coverage Limits from the Run Metadata Although 300 keywords were requested and returned, and the expanded count was 299 (suggesting a near‑complete expansion), we cannot be certain that all relevant long‑tail variations were captured. The mining algorithm follows link graphs from seed keywords; it may miss new or niche phrases that have not yet been linked. This is an inherent limitation of any keyword‑mining tool, not a flaw in this specific dataset.

Volume Bottlenecks Many of the highest‑opportunity keywords have absolute volumes of 10–70 searches per month. Even with perfect optimization, the total traffic attainable from any single one is limited. The opportunity lies in aggregating hundreds of such keywords into topic clusters — a content architecture play rather than a single‑page strategy.

Action Recommendations

The narrative tying together the above findings is this: Japan’s AI auto‑editing search demand is dividing into two streams — an enormous, stable pool of iMovie users who constantly need “how‑to” help, and a fast‑moving stream of tool explorers, currently converging on Vrew. The most valuable real estate is not in generic AI editing terms but in specific, answer‑rich content that matches both the iMovie always‑on need and the Vrew growth spike. Below are the concrete next steps, each derived directly from the data.

1. Content Strategy: Build a Two‑Cluster Content Engine

Cluster A: iMovie How‑to Hub. Create an iMovie tutorial site or section with a pillar page on “アイ ムービー 使い方” (2,900 searches, flat, comp 8) and a ring of supporting articles targeting the fast‑risers:

• “アイ ムービー 動画 つなげる” (40 searches, +133%, comp 8)
• “アイ ムービー を dvd に 焼く” (50 searches, +175%, comp 13) — cover all three closely related variants in one definitive guide.
• “アイ ムービー フェードアウト” (110, +27.3%, comp 1)
• “アイ ムービー 動画 作り方” (10, +100%, comp 11) and its siblings.

The combined monthly search volume of the top 20 iMovie long‑tail keywords exceeds 500; with a well‑interlinked structure, the total traffic can be multiple times higher due to long‑tail overlap. Because competition is essentially absent, simply creating thorough, well‑optimized content will likely secure top‑3 rankings quickly.

Cluster B: Vrew Power‑User Guides. Build a Vrew resource center around “vrew 文字 起こし” (480, +84.6%, comp 34) with supporting pages:

• “vrew 無料” (390, +125%, comp 25) — explain free vs. paid features, monetizable via affiliate links or AdSense.
• “vrew 無料 版” (260, +85.7%, comp 21)
• “vrew 字幕” (140, +23.5%, comp 12)
• “vrew インストール” (70, +28.6%, comp 23)
• “vrew 編集” (40, flat, comp 14) and “vrew カット 編集” (30, flat, comp 8).

The growth trend in Vrew suggests that new users are arriving weekly; capturing them with setup and basic feature guides can turn into a sustainable traffic source.

2. Product Sourcing / Business Development

If your business sells or recommends video‑editing tools, prioritizing Vrew affiliate partnerships or developing a Vrew‑adjacent product (e.g., template packs, preset libraries) make sense. The “vrew 有料” bid data indicates strong commercial intent — users searching for the paid version are ready to spend. You can create comparison pages that pit Vrew against competitors, capturing both “vrew 無料” (exploring free) and “vrew 有料” (ready to buy) traffic. However, avoid over‑investing in declining tool ecosystems like Filmora or PowerDirector; the data shows their automatic feature searches are contracting.

For physical products (e.g., DVD‑burning accessories), the iMovie DVD burning keywords represent a small but intent‑rich audience. A guide that naturally recommends a DVD burner or external drive could convert well.

3. Ad Spend Allocation

Test low‑budget campaigns on the highest‑growth, lowest‑competition keywords to validate conversion paths. “アイ ムービー を dvd に 焼く” and “vrew 無料” are ideal candidates: bid ranges are low (many are even null, implying near‑zero cost per click), and the audience is clearly defined. Set a small daily budget, monitor click‑through and conversion rates, and scale only if the economics work.

Avoid head‑to‑head bidding on “動画 編集 ai” and “ai 動画 編集” unless you have a significant budget advantage. The competition is entrenched, and the traffic is not necessarily transactional.

Use negative keywords to exclude declining terms like “テロップ 自動” or “filmora 自動 字幕” to prevent wasted spend on audiences whose interest is waning.

4. Monitoring and Pivoting

Set up a quarterly re‑mining of the keyword set to track whether the Vrew wave continues or shifts to another tool (CapCut already shows signs of strength). For the iMovie cluster, monitor the trend history of the top 10 keywords; if any start to decline, shift focus to fresh rising iMovie tasks. The short‑lived spike keywords should be monitored monthly — if the spike holds for two consecutive months, it graduates to a stable target.