---
title: "OpenAI Frontier Firms Data: The 6x Skills Gap and What 17.1x Means"
date: "2026-11-05"
video_id: "7704735864672341281"
tiktok_url: "https://www.tiktok.com/@therobertta_/video/7704735864672341281"
creator: "@therobertta_"
category: "AI"
tags: ["openai", "frontier firms", "ai skills", "enterprise ai", "token usage"]
excerpt: "OpenAI data shows frontier firms use skills at 19% vs 3% for typical firms. How to read the 17.1x token chart correctly, and why the 8x spend stat is a16z's."
---
OpenAI's August 12, 2026 enterprise data shows that 19% of weekly active users at "frontier firms" use skills, against 3% at typical firms, a gap of about 6x. A TikTok from Robert Ta (@therobertta_) reports those figures accurately. It misreads the chart behind the "17.1x" token growth figure, and it credits OpenAI for an 8x spending stat that comes from an a16z chart.

## What OpenAI's frontier firms data says about skills

The slide Ta shows is titled "Plugins and Skills adoption: top decile v. middle decile v. OAI" and credits "OAI Signals, 'What Frontier Firms are doing differently' Aug. 12, 2026." Its bars are:

- Skills: 19% for frontier firms (top 10% of enterprises), 3% for typical firms (middle 10%)
- Plugins: 21% for frontier firms, 9% for typical firms
- OpenAI itself: 95% and 93%

The video says "OpenAI's own data says frontier firms use skills at 19% and typical firms use it at 3%. That's about 6x gap on that one capability." 19 divided by 3 is 6.3, so "about 6x" is fair. A RuntimeWire report of the same release says OpenAI published its enterprise studies on 12 August 2026 and that the top 10% of enterprise users adopt skills about six times as often as typical firms.

The y-axis on the slide reads "Share of weekly active enterprise users." The percentages describe how many active users at each group of firms use skills. They do not describe how many firms use skills.

## Do skills make a firm a heavy AI user?

The data does not show that. The chart compares the top decile and middle decile of OpenAI's own enterprise customers, ranked by token use. Frontier firms are picked because they use the most tokens, so the chart shows that heavy users also use skills more. It does not show that adopting skills moves a firm into the top decile. The RuntimeWire report makes the same caution: the data comes from OpenAI's customers and does not show that these tools improve results.

## What "17.1x since April 2025" really means

The video says "frontier firms grew output tokens 17.1x since April 2025 and typical firms 2.1x." The chart's y-axis reads "Output tokens (x Apr 2025 typical)," and its subtitle reads "indexed April 2025."

The base for both lines is the typical firm's April 2025 level. So 17.1x means frontier firms ended June 2026 at 17.1 times what a typical firm produced in April 2025. It does not mean frontier firms grew their own output 17.1 times. The frontier line starts above the typical line at April 2025, so its own growth is smaller than 17.1x. The typical firm's 2.1x is a true growth figure because its base is its own April level.

The chart labels the gap between them "8.3x." That is not exactly 17.1 divided by 2.1, which gives 8.14, and the frame does not show whether the label uses unrounded values or per-user averages.

## How fast is the gap growing?

BankInfoSecurity's report of the OpenAI release defines frontier firms as the top 10% of enterprise customers ranked by output tokens per active user, with typical firms between the 45th and 55th percentiles. It says that in June, frontier firms produced 8.3 times as many output tokens per active user as typical firms, up from 2.6 times in January. That means the gap more than tripled in about five months, a fact the video does not use.

The same report notes that OpenAI treats token volume as an imperfect proxy for business value.

## Where the 8x top 1% spend stat comes from

The video says "the top 1% of AI spenders outspend the top 10% by about 8x. And that's just totally off a bell curve." Its opening line credits everything to "OpenAI's own data."

The slide behind that figure is a different chart. It is the left panel of an a16z Growth deck page titled "Power Users Are Pulling Away," with a footer reading "© 2026 Andreessen Horowitz." The chart is titled "AI Power Usage: Top 1% Outspends The Top 10% By ~8X," and its y-axis reads "Median Panelist AI Vendor Spend." That describes panel data on spending across AI vendors, not OpenAI usage logs. The footnote naming the dataset is too small to read in the video, so the 8x figure could not be traced past the slide.

## Is this "crossing the chasm"?

Ta says the gap is between "the innovators versus the everyone else right now" on Moore's technology adoption curve, and that "I don't even think we've reached the laggards of the market yet." Nothing on the slides places these groups on an adoption curve. Both groups are paying OpenAI enterprise customers, so even the middle-decile firm has already adopted the product. The innovator label is the speaker's framing.

He closes by recommending that teams treat skills as "first class artifacts in your code base" and says this will "optimize your token spend for value." No slide measures value per token or links skills to lower or better-spent tokens, and the clip shows no skill file or repository. The description links a longer YouTube video, which was not reviewed here.

## Bottom line

The 19% versus 3% skills figures and the 8.3x token gap match OpenAI's August 2026 release as reported by RuntimeWire and BankInfoSecurity. The 17.1x figure is measured against the typical firm's April 2025 level, not frontier firms' own growth. The 8x top 1% spend chart is an a16z panel, not OpenAI data, and none of the sources show that skills cause better results.

## Sources

- [OpenAI data finds top enterprise users adopt reusable AI skills six times as often, RuntimeWire, 13 Aug 2026](https://runtimewire.com/article/openai-enterprise-ai-skills-plugins-frontier-firms): the six-fold skills gap, the publication date and the customer-data caveat.
- [Enterprise AI Token Spend Shifts From Chat to Agents, BankInfoSecurity, 12 Aug 2026](https://www.bankinfosecurity.com/enterprise-ai-token-spend-shifts-from-chat-to-agents-a-32539): frontier and typical firm definitions, 8.3x in June against 2.6x in January, and the token-volume caveat.
- [How enterprises put AI to work, OpenAI](https://openai.com/index/how-enterprises-put-ai-to-work): the primary page named by the reports above, which returned an access error when fetched.


## Video description and verification notes

**The file is a 57.557 s portrait clip, 1080x1920, MP4 (isom) with HEVC Main video at 24 fps (303,128 bps) and HE-AACv2 stereo audio at 44.1 kHz (64,186 bps). Overall bitrate is 372,329 bps and the file is 2,678,768 bytes. metadata.json lists format bytevc1_1080p_372322-1 and a null fps field, so I took fps from ffprobe. The post timestamp is 1793898609, which is 2026-11-05 17:10:09 UTC. The uploader is @therobertta_, nickname Robert Ta, and the audio credit is "original sound" by Robert Ta. The description is about 480 characters and links to a longer YouTube video (https://youtu.be/MtrUaca1zfA), which I did not fetch. Engagement at archive capture (2026-11-07 13:01 UTC) was 1,612 views, 38 likes, 3 comments, 6 reposts and 28 saves. I read six frames: 1, 4, 8, 15, 22 and 29, which sit at roughly 0, 6, 14, 28, 42 and 56 s on the 2 s grid. I did not read the other frames, so I make no claim about them. Every frame I read uses the same split layout: a screen-share slide in the upper half and Robert Ta, a dark-haired man with a mustache, in a close-up webcam shot below it. Frame 1 is a title-style layout with the burned-in caption "Frontier firms use skills 6x more than you" and "For AI builders", plus two webcam tiles, Ta on the left and a second man in earbuds on the right. In frames 1, 4, 8 and 15 the slide is a green and cream deck page numbered 33, titled "Power Users Are Pulling Away", with the text "A tiny fraction are using orders of magnitude more AI than the field, and the gap has been widening since early this year." The footer reads "© 2026 Andreessen Horowitz", and the pages carry an "a16z Growth" mark. The left chart is titled "AI Power Usage: Top 1% Outspends The Top 10% By ~8X". The right chart is "AI Power Users Pulling Away From The Field", with the subtitle "Output tokens, top decile v. middle decile, indexed April 2025". Frame 22 zooms into the "All Industries" panel and shows end labels "17.1x" and "2.1x" and the label "8.3x gap". The neighboring "Information" panel shows 32.5x, 2.8x and an "11.7x gap". Frame 29 shows a bar chart of "Plugins and Skills adoption: top decile v. middle decile v. OAI". It has bars of 95% and 93% for OpenAI, 21% and 19% for frontier firms, and 9% and 3% for typical firms. The footnote reads "Source: OAI Signals, 'What Frontier Firms are doing differently' Aug. 12, 2026". Burned-in captions follow the speech word by word ("of AI spenders", "2.1x.", "don't even think", "spend for value."). No terminal, code editor, skill file or repository appears in any frame I read. The speaker argues that OpenAI data shows frontier firms using skills about 6x more than typical firms, that spend and token growth are far from a bell curve, that the market is still at the early-adopter stage, and that teams should write their skills down as first-class code artifacts to get value from token spend. The transcript has 10 segments and about 135 words, and the last segment ("for value.") runs 56.86 to 57.44 s, inside the 57.557 s audio, so the text is genuine and no hallucinated tail exists.**

## The 19% vs 3% skills figures match the slide, and "6x" is rounded from 6.3

At 0.00 s the video says "OpenAI's own data says frontier firms use skills at 19% and typical firms use it at 3%. That's about 6x gap on that one capability." Frame 29 shows exactly those bars: 19% for "Frontier firms (top 10% of enterprises)" and 3% for "Typical firms (middle 10% of enterprises)". The y-axis is "Share of weekly active enterprise users", so the percentages describe the share of weekly active users at each group of firms who use skills. They are not the share of firms that use skills. The ratio 19 divided by 3 is 6.3, so "about 6x" is a fair rounding. The slide credits "OAI Signals" with the title "What Frontier Firms are doing differently" and the date Aug. 12, 2026. A RuntimeWire report of the same release says OpenAI published its enterprise studies on 12 August 2026 and that the top 10% of enterprise users adopt skills about six times as often as typical firms. That report does not itself carry the 19% and 3% figures. A web search summary attributed the 19% and 3% numbers, plus Plugins at 21% and 9%, to OpenAI's Enterprise Signals material. That matches frame 29, but it is still a secondary summary. I could not open OpenAI's own page (https://openai.com/index/how-enterprises-put-ai-to-work returned HTTP 403), so I cannot quote the primary text. The numbers are therefore verified against the slide and against a second report of the release, and unverified against OpenAI's page itself.

The video's framing of this as a gap in "skills" also needs a limit. The chart compares the top decile and the middle decile of OpenAI's own enterprise customers, ranked by token use. The frontier group is selected by heavy usage, so the chart shows that heavy users also use skills more. It does not show that adopting skills moves a firm into the top decile. The RuntimeWire report makes the same caution: the data comes from OpenAI's customers and does not show that these tools improve results.

## The "8x" top-1% spend stat comes from a different chart with an unreadable source, not from OpenAI

At 13.96 s the video says "the top 1% of AI spenders outspend the top 10% by about 8x. And that's just totally off a bell curve." The opening line of the clip attributes everything to "OpenAI's own data." The slide in frames 4 and 8 shows that the "Top 1% Outspends The Top 10% By ~8X" chart is the left panel of an a16z Growth deck page, and its y-axis reads "Median Panelist AI Vendor Spend". That label describes panel data on spending across AI vendors, not OpenAI usage logs. The footnote under the chart is too small to read in the frames I have, so I could not identify the dataset or its vendor, and I will not guess one. What the slide supports is this: the 8x figure is an a16z-presented spend chart, separate from the OpenAI Signals token and skills charts on the right side of the same page. The OpenAI attribution in the video's first sentence fits the skills figure (frame 29's footnote) and the token chart (frame 22's subtitle), and does not fit the spend chart. I could not trace the spend figure to a primary source, so it stays unverified beyond "the slide says ~8X".

## "Frontier firms grew output tokens 17.1x since April 2025" misreads the chart's index

At 20.68 s the video says "frontier firms grew output tokens 17.1x since April 2025 and typical firms 2.1x." Frame 22 shows the y-axis label "Output tokens (x Apr 2025 typical)" and the subtitle "indexed April 2025". The index base is the April 2025 level of the typical firm, so 17.1x means the frontier line ends June 2026 at 17.1 times what a typical firm produced in April 2025. It does not mean the frontier group's own output grew 17.1 times from its own April 2025 starting point. The green line also starts above the red line at Apr-25, so the frontier group's own starting point was higher than 1x, and its own growth is smaller than 17.1x. The typical line ending at 2.1x is therefore a typical-firm growth figure, because its base is its own April level. The two numbers are on one scale, but they do not describe the same thing as "grew 17.1x" and "grew 2.1x".

The 8.3x gap is also the chart's own label, and the ratio is not simply 17.1 divided by 2.1 (which is 8.14). The label probably uses unrounded values or per-user averages; I cannot tell which from the frame. BankInfoSecurity's report of the OpenAI release says frontier firms are the top 10% of enterprise customers ranked by output tokens per active user, typical firms sit between the 45th and 55th percentiles, and in June frontier firms produced 8.3 times as many output tokens per active user as typical firms, up from 2.6 times in January. That agrees with the 8.3x label and with the June 2026 end point on the chart axis. It also shows the gap was 2.6x in January, which is a stronger fact than the video uses: the gap more than tripled in about five months. The same report notes that OpenAI treats token volume as an imperfect proxy for business value. The video's closing line, that skills "optimize your token spend for value", goes further than the data, because no slide measures value per token or links skills use to lower or better-spent tokens.

## The adoption-curve claim is the speaker's opinion, and the recommendation is not tested by any source shown

At 28.00 s the video says the gap is "between the, what I would call in the Moore's crossing the chasm technology lifecycle adoption curve, the innovators versus the everyone else right now," and at 39.74 s it says "I don't even think we've reached the laggards of the market yet." Nothing on the slides places these groups on a diffusion curve. The groups are top-decile and middle-decile customers of one vendor, and a middle-decile OpenAI enterprise customer has already adopted the product. Calling the top decile "innovators" and the rest "early adopters" is a framing choice by the speaker, and the sources I read neither support nor contradict it. The description's phrase "frontier firms use skills about 6x more than typical firms" is consistent with frame 29. The recommendation at 44.92 s to codify skills as "first class artifacts in your code base" is advice, and no frame shows a skill file, a repository or a measured saving. The clip points viewers to the full YouTube video, which I did not fetch, so any evidence for the advice that lives there is unchecked.

## Key Takeaways

- **Verified:** The 19% (frontier) and 3% (typical) skills figures, and the 21% and 9% plugin figures, appear on the slide in frame 29 and are credited there to OpenAI's "OAI Signals" release of Aug. 12, 2026. A secondary report of the release also states the roughly six-fold skills gap.
- **Verified:** The 8.3x output-token gap, with frontier firms defined as the top 10% and typical firms as the middle band, matches BankInfoSecurity's report of the OpenAI release, which also gives 2.6x for January.
- **Correction:** The video says at 20.68 s that frontier firms "grew output tokens 17.1x since April 2025 and typical firms 2.1x." The chart axis in frame 22 reads "x Apr 2025 typical", so 17.1x is a multiple of the typical firm's April 2025 level, not frontier firms' own growth.
- **Partial correction:** The video opens by attributing all its statistics to "OpenAI's own data." The "top 1% outspends the top 10% by ~8X" chart is an a16z deck panel labeled "Median Panelist AI Vendor Spend", not an OpenAI usage chart, and the video does not say so.
- **Clarification:** The skills percentages are shares of weekly active enterprise users within each group of firms, and the groups are chosen by token volume, so the data does not show that skills cause heavier use.
- **Unverified:** I could not open OpenAI's own report page (HTTP 403), and I could not read the footnote identifying the data behind the ~8X spend chart, so neither is checked at the source.
- **Context:** The Moore adoption-curve labels and the advice to treat skills as code artifacts are the speaker's opinion. No frame shows a skill file, a repository or a measured effect on token spend.

## Resources

- [OpenAI data finds top enterprise users adopt reusable AI skills six times as often, RuntimeWire, 13 Aug 2026](https://runtimewire.com/article/openai-enterprise-ai-skills-plugins-frontier-firms) establishes the roughly six-fold skills gap, the 12 August 2026 OpenAI publication date and the caveat about customer-only data. It names OpenAI's "How enterprises put AI to work" page as its source.
- [Enterprise AI Token Spend Shifts From Chat to Agents, BankInfoSecurity, 12 Aug 2026](https://www.bankinfosecurity.com/enterprise-ai-token-spend-shifts-from-chat-to-agents-a-32539) establishes the frontier and typical firm definitions, the 8.3x June figure against 2.6x in January, and OpenAI's caveat that token volume is a proxy.
- [Frontier firms are pulling further ahead in enterprise AI use, says OpenAI, T2 Online](https://t2online.in/tech/tech-news/frontier-firms-are-pulling-further-ahead-in-enterprise-ai-use--says-openai/2007224) appeared in a search result that summarized the 19% and 3% figures. I did not open it, so I use it only as a pointer.
- OpenAI, "How enterprises put AI to work" (https://openai.com/index/how-enterprises-put-ai-to-work) is the primary source named by the secondary reports. It returned HTTP 403 to my fetch, so I do not cite any text from it.
