Hollywood-grade is the reader's test: the files hold no budget, crew or quality data, so this catalogue records what the frames show, not how well anything was made. Across 74 B2B tech demo videos from 62 companies, uploaded between 14 January 2026 and 3 August 2026, an interface joins the cinematic picture in 61 of 74, and 58 carry three or more media families. Live-action or stock picture is in the build of 46, CGI in 26; the picture comes before the interface in 35 of the 51 that carry both; and a voiceover carries 54. OSHIMA, Zoho, HPE and Illumina are among the videos catalogued. The set is 74 videos, 6 under the 80 this research series normally requires, and every share is printed with its count.
ADVIDS Research · 74 B2B tech demo videos coded · 62 companies · 666 sampled frames · 769 timed segments · published 14 September 2026 · 21 min read
Top takeaways
An interface joins the cinematic picture in 61 of 74 demos
An interface joins the cinematic picture in 61 of 74 B2B tech demo videos uploaded January to August 2026, and 58 carry three or more media families; the 16 that pair the picture with one family only come from Workday, Bio-Rad, Cegid, Dell Technologies, etc.
Live-action picture outnumbers CGI, 46 to 26
Live-action or stock picture is in the build of 46 of 74 B2B tech demo videos from 2026 and 3D CGI in 26; the 8 that use both come from Cavli Wireless, Dell Technologies, Illumina, Jimi IoT, etc.
The picture opens and the interface follows in 35 of 51
In 35 of the 51 B2B tech demo videos from 2026 with both a cinematic and an interface segment, the picture comes first; the 13 that open on the interface come from G-Stacker, Rillion, RMS, Sana, etc.
The interface cuts in full frame in 48 of 61
The interface cuts in full frame in 48 of the 61 B2B tech demo videos from 2026 that show one, and enters the shot on a device or as an overlay in 34; the 13 overlays come from SAP, Cavli Wireless, Cegid, Contentstack, etc.
The interface outruns the cinematic picture in 42 of 74
The interface covers more runtime than the cinematic picture in 42 of 74 B2B tech demo videos from 2026, a median 33 per cent against 24 per cent; the 16 in which the picture covers half the runtime come from Lenze, OSHIMA, PNY, ProLabs, etc.
A voiceover carries 54 of 74 mixed-media demos
A voiceover carries 54 of 74 B2B tech demo videos from 2026; the 20 that run on music alone come from New Relic, Omron, Setmore, SevenRooms, etc., and 11 of those put the words on screen as kinetic type.
Dark and mixed frames hold 54 of 74; blue leads the palette
Dark or mixed frames hold 54 of 74 B2B tech demo videos from 2026, at a median brightness of 115 out of 255, and blue leads 29 of the 49 chromatic palettes; the 20 light-frame videos come from SCATI, ScienceLogic, StackAI, Sword Health, etc.
Questions this report answers
What is mixed media animation in a B2B tech demo video?
Here it is cinematic picture, meaning live action, stock or CGI, mixed with an interface, motion graphics or 2D animation, or kinetic type. Of 74 B2B tech demo videos from 62 companies uploaded January to August 2026, 58 carry three or more of those four families, OSHIMA and Zoom among them.
Do B2B tech demo videos use live action or CGI?
Live action more often: it is in the build codes of 46 of 74 B2B tech demo videos uploaded January to August 2026, against 26 for 3D CGI and 8 for both. CGI with no live action carries 18, including all three Intel Foundry videos.
How does the product interface appear in a cinematic tech video?
Mostly cut in full frame. Of the 61 B2B tech demo videos from 2026 with an interface segment, 48 cut to it in full frame and 34 place it inside a shot at least once, as with the over-shoulder laptop in the Zoho Cliq video.
What order do cinematic B2B demo videos run in?
Picture first, then product: cinematic picture is the first medium in 40 of 74 B2B tech demo videos from 2026, and it comes before the interface in 35 of the 51 that carry both. A title or logo card closes 55 of them.
Do cinematic B2B product videos use a voiceover?
Yes: 54 of 74 B2B tech demo videos uploaded January to August 2026 carry a voiceover. The other 20, Crestron and Symantec among them, run on music, and 11 of those carry their words as on-screen kinetic type.
How was this analysis done, and what does Hollywood-grade mean here?
ADVIDS Research screened 188 visually coded videos against a rule written before intake; 74 from 62 companies qualified, 6 under the series' usual 80, and were read from frame-coded segments. Hollywood-grade is the reader's test: the files hold no budget or quality data, so nothing here rates a mix.
How to read this report
The set is 74 product, launch, demo and explainer videos from 62 companies selling technology to businesses, uploaded to YouTube between 14 January 2026 and 3 August 2026; each mixes cinematic picture with an interface, motion graphics or 2D, or kinetic type. It is a catalogue of what the set does, with no time periods. N of M counts videos: M is 74 unless a smaller base is named, such as the 61 with an interface segment. A pattern is a finding only when it holds half its base, and runtime shares are medians among the videos carrying the medium.
Pattern 01 · Build · Combination · 74 B2B tech demo videos, January to August 2026
1. An interface joins the cinematic picture in 61 of 74 demos
An interface appears beside the cinematic picture in 61 of 74 B2B tech demo videos uploaded between 14 January 2026 and 3 August 2026, motion graphics or 2D animation in 56 and kinetic type in 44. 58 of the 74 carry three or more media families and 29 carry all four, the largest single combination, ahead of cinematic picture + interface + motion graphics or 2D at 18. 16 pair the picture with one family only. The OSHIMA cutting-machine video, illustrated here, carries all four. Mixed media in this set is the starting point; the choice is which media, not whether.
Companies in this group include SAP, Crestron, Bloomfire, Cadence, etc. The two-family videos come from Workday, Bio-Rad, Cegid, Dell Technologies, etc. Without SAP, the largest company in the pattern, the interface share moves 1.0 points.
Global Payments goes the other way. Its February 2026 Genius for restaurants video runs 0:50 of live-action restaurant stock, a point-of-sale product collage and kinetic-type slogans, with no interface segment at all. Where the product is hardware at a counter, the device and the type carry it.
Source: ADVIDS Research coded set, 74 videos from 62 companies, uploaded 14 January 2026 to 3 August 2026
The 7 combinations add to 74; each video sits in exactly one.
Pattern 02 · Build · Cinematic source · 74 B2B tech demo videos, January to August 2026
2. Live-action picture outnumbers CGI, 46 to 26
Live-action or stock picture is in the build codes of 46 of 74 B2B tech demo videos uploaded January to August 2026, and 3D CGI in 26; 8 carry both. By segment, camera b-roll or stock appears in 43 and a product or hardware shot in 34, and the format is listed as cinematic for 36. Build codes are medium confidence for 15 of the 46 live-action videos and 10 of the 26 CGI videos. The Lenze motion-control video, illustrated here, puts its pick-and-place arm in front of a camera. The camera supplies the picture in more of these videos than the renderer does.
Companies in this group include AFL, Amplitude, Cavli Wireless, Emburse, etc. CGI with no live action comes from Intel, ElevenLabs, ESSERT Robotics, GitLab, etc. Without SAP, the live-action share moves 2.2 points.
Intel goes the other way. Its March 2026 Omni-MIM video runs 0:41 on 3D renders of capacitor stacks and chip layers, with no live-action build code and no camera segment. Where the product sits inside a chip, there is nothing for a camera to capture.
Source: ADVIDS Research coded set, 74 videos from 62 companies, uploaded 14 January 2026 to 3 August 2026
The four groups add to 74. The 10 with neither build code carry their cinematic picture in segments or the format list only.
Pattern 03 · Timing and frame · Order · 74 B2B tech demo videos, January to August 2026
3. The picture opens and the interface follows in 35 of 51
Cinematic picture is the first medium in 40 of 74 B2B tech demo videos uploaded January to August 2026, kinetic type in 15, the interface in 13 and motion graphics in 6. Among the 51 videos with both a cinematic and an interface segment, the picture comes first in 35, and in those the first interface segment starts a median 16 seconds after the first cinematic one. The shared opening run is cinematic picture → interface → cinematic picture, in 14 videos. A title or logo card opens 25 and closes 55, and a direct cut between a cinematic and an interface segment appears in 40. The Jimi IoT tracker video, illustrated here, runs three product shots before its app appears. The picture sets the scene, and the product screen answers it.
Companies in this group include IBM, Epicor, Genpact, HCLTech, etc. Videos that open on the interface come from G-Stacker, Rillion, RMS, Sana, etc. Without Crestron, the picture-first share moves 1.3 points.
Celtra goes the other way. Its May 2026 AI Creative Flows video opens on its node canvas at 0:02 and reaches its only cinematic segment, a grid of ad variations, at 0:22, in a video that runs 0:39. Where the product is the canvas, the canvas is the opening.
Source: ADVIDS Research coded set, 74 videos from 62 companies, uploaded 14 January 2026 to 3 August 2026
The four add to 74. Cards, presenters and abstract backgrounds are set aside, so the first medium is the first segment that is one of the four families.
Pattern 04 · Interface · Interface entry · 74 B2B tech demo videos, January to August 2026
4. The interface cuts in full frame in 48 of 61
The interface cuts in full frame in 48 of the 61 B2B tech demo videos from January to August 2026 that carry an interface segment, and enters the shot at least once in 34: on a device in 25 and as an overlay on the picture in 13. 21 of the 61 do both, and 27 never put the interface inside a shot. Across all 74, the build codes list an animated interface mockup for 28 and captured product screens for 18. The Zoho Cliq video, illustrated here, opens on its interface on an over-shoulder laptop. Entry is read from each segment's frame-coded label, so a device the label does not name counts as full frame. Full frame is how the interface arrives; a device in the shot is how it joins the picture.
Companies in this group include HPE, AIMS360, Amplience, Arm, etc. The interface sits on a device in the shot for Illumina, Jimi IoT, MaintainX, Ofelia, etc. Without SAP, the full-frame share moves 1.5 points.
Rillion keeps the two apart. Its March 2026 Rillion Capture video cuts between invoice screens in full frame and a live-action shot of a woman at a laptop, and the interface never enters that shot. Where the screen carries fields to read, it gets the whole frame.
Source: ADVIDS Research coded set, 74 videos from 62 companies, uploaded 14 January 2026 to 3 August 2026
The three add to 61. Inside the shot means on a device the segment label names, or overlaid on the picture.
Pattern 05 · Timing and frame · Runtime share · 74 B2B tech demo videos, January to August 2026
5. The interface outruns the cinematic picture in 42 of 74
The interface covers more runtime than the cinematic picture in 42 of 74 B2B tech demo videos uploaded January to August 2026. Among the videos that carry each, the interface covers a median 33 per cent of runtime (n=61), cinematic picture 24 per cent (n=64), kinetic type 20 per cent, motion graphics or 2D 16 per cent and title or logo cards 14 per cent. Inside the picture, product or hardware shots cover a median 29 per cent and camera b-roll or stock 20 per cent. The picture covers half the runtime or more in 16 videos and the interface in 10. The HPE Data Fabric video, illustrated here, gives its dashboards more runtime than its office picture. The picture frames the demo, and the screens fill it.
Companies in this group include Contentstack, Google Cloud, Infobip, Jodoo, etc. The picture covers half the runtime or more for Lenze, OSHIMA, PNY, ProLabs, etc. Without SAP, the share moves 1.1 points.
ESSERT Robotics goes the other way. Its February 2026 cleanroom automation video runs 2:36, robot hardware covers 87 per cent of it and the interface 5 per cent. Where the product is a machine, the machine takes the runtime.
Source: ADVIDS Research coded set, 74 videos from 62 companies, uploaded 14 January 2026 to 3 August 2026
Each bar is the median among the videos that carry that medium, so the bars do not add to 100.
Pattern 06 · Delivery · Delivery · 74 B2B tech demo videos, January to August 2026
6. A voiceover carries 54 of 74 mixed-media demos
A voiceover carries 54 of 74 B2B tech demo videos uploaded January to August 2026, and 20 run on music with no voiceover; 11 of those 20 put words on screen as kinetic-type segments. A presenter appears on camera in 10, captions are burned into the picture in 5, and 35 carry 120 or more spoken words; the median voiceover video speaks 148 words. The Illumina Insight Ecosystem video, illustrated here, is one of the voiceover videos; in the lab shot shown, no words are on screen and the voice carries them. Word counts come from auto or manual captions, and a music track with vocals can register words. The picture shows, and the voice explains.
Companies in this group include Doxis, Global Payments, Rackspace Technology, Samsung, etc. Music-only videos come from New Relic, Omron, Setmore, SevenRooms, etc. Without SAP, the voiceover share moves 1.6 points.
Symantec goes the other way. Its March 2026 CBX video runs 1:14 on music with no voiceover, over 3D renders of amplifiers and gauges labelled Protection and Uptime, and closes on a kinetic-type tagline. Where the idea is a metaphor, the labels on the objects do the talking.
Source: ADVIDS Research coded set, 74 videos from 62 companies, uploaded 14 January 2026 to 3 August 2026
The three add to 74.
Pattern 07 · Timing and frame · Frame · 74 B2B tech demo videos, January to August 2026
7. Dark and mixed frames hold 54 of 74; blue leads the palette
Dark or mixed frames hold 54 of 74 B2B tech demo videos uploaded January to August 2026: 29 mixed, 25 dark and 20 light, at a median brightness of 115 on a 0 to 255 scale. Blue is the lead hue in 29 of the 49 videos with a chromatic palette, ahead of warm and green at 8 each; 25 carry no chromatic lead. The AFL splicer teaser, illustrated here, runs at a mean brightness of 11 with blue as its lead hue. The measure is luminance and palette across all nine sampled frames, not the product interface's theme. A dark, blue ground is the stage the picture and the screens share.
Companies in this group include Showpad, SSI Schaefer, SuperOffice, Symantec, etc. Light-frame videos come from SCATI, ScienceLogic, StackAI, Sword Health, etc. Without Intel, the dark-or-mixed share moves 1.2 points.
SAP goes the other way. Its May 2026 AI Agent Hub overview runs 1:44 on light frames at a mean brightness of 171, with an aerial city opening and warehouse picture around Joule chat panels and agent dashboards. Where the product lives in a light interface, the picture around it stays light too.
Source: ADVIDS Research coded set, 74 videos from 62 companies, uploaded 14 January 2026 to 3 August 2026
The five add to 74. Lead hue is the first chromatic colour in the computed palette; greys and near-blacks are set aside.
Benchmarks
Descriptive figures for placing your own B2B tech demo video against the set.
Length, timing and picture
| Measure | 74 videos | What it means |
|---|---|---|
| Runtime, median | 1:14 | Half the videos run shorter than this |
| Runtime, middle half | 0:57 to 1:44 | The runtime span of the middle 50 per cent of videos |
| First cinematic segment, median | 0:07 (n=64) | When the first camera, stock, CGI or product segment starts, among videos that carry one |
| First interface segment, median | 0:17 (n=61) | When the first interface segment starts, among videos that carry one |
| First kinetic-type segment, median | 0:07 (n=34) | When the first kinetic-type segment starts, among videos that carry one |
| First brand logo card, median | 1:00 (n=49) | When the first logo card appears, among videos that carry one |
| Media switches per video, median | 3 | Changes of media family from one segment to the next, cards and presenters set aside |
| Scene changes per minute, median | 8.3 | Coded segment boundaries per minute of runtime |
| Words spoken, median | 104 | Words in the captions; music-only videos with no captions count as zero |
| Frame brightness, median | 115 | Mean luminance across the nine sampled frames, 0 to 255 |
What few mixed-media demos do
| Coded value | Videos | Base | Companies |
|---|---|---|---|
| Interface overlaid on the picture | 13 | 74 | SAP, Cavli Wireless, Cegid, Contentstack, etc. |
| Interface is the first medium | 13 | 74 | G-Stacker, Rillion, RMS, Sana, etc. |
| Presenter on camera | 10 | 74 | Bloomfire, Emburse, G-Stacker, Jodoo, etc. |
| Interface covers half of runtime or more | 10 | 74 | Bloomfire, HPE, Sana, SAP, etc. |
| Captions burned into the picture | 5 | 74 | Cegid, G-Stacker, Ofelia, OSHIMA, etc. |
Each of these holds under a quarter of the set, so none is a pattern of the catalogue.
What to do with this
- Pattern 01 If an interface sits beside the picture in 61 of 74 B2B tech demos, brief the screens and the picture as one scope rather than two separate jobs.
- Pattern 02 If live action is in the build of 46 of 74 and CGI in 26, decide whether the product can stand in front of a camera before pricing renders.
- Pattern 03 If the picture comes before the interface in 35 of 51, write where the first product screen lands into the video script; in those videos it lands a median 16 seconds after the picture starts.
- Pattern 04 If 34 of 61 put the interface inside a shot at least once, list which screens must be captured or composited at the shot's angle before production is scheduled.
- Pattern 05 If the interface outruns the picture in 42 of 74, plan more screen time than picture time and have the screens ready first.
- Pattern 06 If a voiceover carries 54 of 74, book and record the voice before picture lock, and write on-screen type as the carrier only for a music-only cut.
- Pattern 07 If 54 of 74 run dark or mixed frames, set one ground for picture and product screens in the creative brief rather than grading them apart.
Methodology
| Vertical | Videos |
|---|---|
| retail | 4 |
| fintech | 4 |
| semiconductors | 4 |
| enterprise | 4 |
| cx | 3 |
| logistics | 2 |
| hospitality | 2 |
| other verticals | 51 |
| Coding scheme | Values | Definition |
|---|---|---|
| Cinematic picture | cinematic format · live_action · stock · 3d_cgi · cinematic_broll · office and crowd stock · product_shot · robot_hardware · cgi_environment · facility_render · manufacturing | Coded from frames: camera or rendered picture. Live action and stock count only when the format is not talking_head. A product_shot segment can be a render or a mockup the coder framed as a product. |
| Interface | ui_mockup · screen_recording · screencast · dashboard_ui · chat_ui · terminal_ui · handheld_device_ui · notification_alert · screen_recording_generic | Coded from frames: the product's screens, animated as a mockup or captured from the product. |
| Motion graphics or 2D | motion_graphics · 2d_animation · diagram_network · diagram_flowchart · data_chart · illustration motifs | Coded from frames: drawn graphics, diagrams, charts and illustration. |
| Kinetic type | kinetic_type as format, build code or segment | Coded from frames: animated words as a medium. Title cards and logo cards do not count. |
| Interface entry | device · overlay · full frame | Read from the frame-coded label of each interface segment: device when the segment is a handheld-device motif or its label names a laptop, phone, tablet, monitor, panel or over-shoulder view; overlay when the label names an overlay, hologram or floating panel; full frame otherwise. |
| Order | first medium · first cinematic and interface timecodes · closing segment | Computed over timed segments. The first medium is the first segment that is not a card, presenter or abstract background. |
| Runtime share | per cent of runtime | Computed: the union of a family's timed segments over runtime. A family present only in format or build codes has no segment and no share. |
| Delivery | voiceover · music only · presenter · captions burned in · word count | Voiceover or music only comes from audio analysis; presenters and burned-in captions are coded from frames; word counts come from auto or manual captions. |
| Frame | light · dark · mixed · brightness 0 to 255 · lead hue | Computed, not coded: mean luminance and palette across the nine sampled frames. It covers every frame, not the product interface alone. |
| Not coded | cast · surface · altitude | Who appears, how much of the product surface shows, and what each video argues. The library records none of them exactly, and this report does not code them from transcripts. |
Cite this report
Jai Ghosh, Advids. The Cinematic Report: We Analyzed 74 B2B Tech Demos for Hollywood-Grade Mixed Media Animation. 14 September 2026. https://advids.co/blog/cinematic-mixed-media-b2b-tech-demos
Author & editor bio
About Advids
Advids produces video for AI, SaaS and DeepTech companies. ADVIDS Research publishes coded studies of what those companies ship, with the dataset attached so every count can be checked.
Image credits and licensing
Video frames on this page are reproduced for analysis and commentary. Copyright in each frame remains with the company credited in its caption; to license a frame, contact that rights holder. Advids grants no rights over them.
Charts and images credited ADVIDS Research, and the coded dataset behind them, are © 2026 Advids and licensed CC BY 4.0: reuse them with attribution to Advids and a link to this page.