vpv.

← Notes·2026-09-09·behind-the-video / startup-vlog / documentary

How to film a tech scene documentary from the inside: Da Nang, Vietnam.

How to film a documentary vlog about a city's tech scene while taking part in it: order it by access, shoot proof instead of coverage, and narrate the gaps.

By updated 7 min read

Story pass and shot list: Planned in Vanpelt Studio →

A documentary vlog about a tech scene works when it is ordered by access, from the rooms anyone can walk into to the ones you need an invitation for, and when the person filming is inside the scene rather than touring it. This note is for founders, creators and community organisers who want to film a city's startup world without making a list of venues. The short answer: promise the viewer a count of layers, film each layer as a participant, shoot a few images that prove you were inside, and let the last room be the one that was hardest to get into.

The worked example: a six-minute documentary from Da Nang

The worked example is I Investigated Vietnam's Next Tech Hub, 6 minutes 41 seconds, published in August 2026. I spent that summer in Da Nang building Vanpelt Studio on the road, and the scene around me deserved more than a passing mention. The film moves from open co-working days to a hackathon I placed second in, to an invite-only hacker house, and ends at a demo day where I was asked to pitch. Every step below uses it.

What are the six steps of a tech scene documentary?

  1. 01

    Promise a count of layers

    Tell the viewer how many levels there are in the first 20 seconds. Output: one line, like three levels of tech events you can join.

    Before filming
  2. 02

    Rank the rooms by access

    Open, registration, selection, invitation. Output: a ladder of four or so rooms.

    20 min
  3. 03

    Enter each room as a participant

    Code, compete, pitch. The camera is the second job. Output: a reason to be in the room.

    Each event
  4. 04

    Shoot proof, not coverage

    A wide of the room on arrival, then three or four details only an insider would see. Output: a short proof list per layer.

    10 min per event
  5. 05

    Narrate in the gaps

    Corridors, walks between venues, the flat at night. Output: a diary track that carries what the footage cannot.

    Daily
  6. 06

    Cut from the outline, not the footage

    Lay out the layers as beats before trimming a clip. Output: an edit order that climbs.

    Edit day
The method on one card. A plan for any scene documentary, not a timecode log of this one.

Why the obvious structure fails

Most tech scene videos are a list of places, visited in the order the maker happened to visit them. Nothing connects, so nothing compounds, and the viewer has no reason to stay after the second venue.

Da Nang is an easy place to fall into that trap because there is a lot to list. It is the fourth largest city in Vietnam by population and home to the Da Nang Hi-Tech Park, set up in 2010 as one of the country's three national hi-tech parks. The country around it is moving fast too: in July 2026 the World Bank reclassified Vietnam as an upper-middle-income country, and the government's stated goal is high-income status by 2045. Facts like these make a good voice-over line, and a terrible spine. A viewer does not stay for statistics; they stay to see whether you get further in.

So the film opens with a claim and a count instead. The claim: everyone talks about San Francisco, but Vietnam has been quietly building the next international tech hub, and not a place you pass through for a few weeks like Bali. Then the promise: three different levels of tech events you can join in Da Nang. The count is the contract. The viewer knows how far the video goes and can feel each layer being paid off.

Step 1: rank the rooms by how hard they are to enter

Order the film by access rather than by chronology. Each layer should be harder to get into than the last, which gives the viewer a reason to stay that a list never has.

  1. Open to anyone

    Co-working with the founder group

    A group chat where someone asks every day where people are working. You turn up or you do not. The surface of any scene, and the only layer most videos reach.

  2. Registration required

    Events and the hackathon

    Sessions with a few talks, and a hackathon I joined 72 hours after arriving, without planning to, and placed second in with a teammate I had just met.

  3. Invitation only

    The hacker house

    A residency for builders, where the film arrives early, the doorbell does not work and most of the house is still asleep.

  4. Asked to pitch

    Demo day, in front of the startup minister

    The most closed room in the film, and an invitation I did not have when I started shooting.

The film's spine, as it appears on screen. Rooms ordered by how hard they are to walk into, not by the date I walked into them.

Ordering by access also solves a documentary problem: it makes the filmmaker's presence legitimate. The viewer is not watching someone tour a scene, they are watching someone get further into it, which is a story rather than a survey. Your output for this step is a ladder of rooms, written down before the first event.

Step 2: enter every room as a participant

None of these were shoots. Every one was an event I attended to work, code or compete, with the camera as the second priority. That is the constraint that shaped the film, and it is also the source of its access.

The hackathon is the clearest case. I did not plan to take part; I crashed it, ended up on a team with someone I had just met from Germany, and we built a browser game based on the Japanese shape-in-the-wall shows, where you strike the pose of the gap coming towards you and land on a leaderboard. We placed second. A second-place finish on camera is a better scene than any interview about the scene, because it proves the layer instead of describing it.

The general rule: you cannot cover an event you are taking part in, and you should not try. Your output is a reason to be in each room that has nothing to do with filming it.

Step 3: shoot proof, not coverage

Proof is the handful of images that show you were inside: things a visitor with a press pass would not see. Get a wide of each room the moment you arrive, because it is the one shot you can take without missing anything, then collect the details.

In the co-working layer, the proof is a conversation. I asked two people to re-explain on camera what they had just been discussing ("for the sake of having a cool shot"), and one of them mentioned hitting a million monthly impressions. That number did more for the layer than any establishing shot. At the hacker house, the proof is the awkwardness: ringing a broken doorbell, being waved in, sitting down in a house of builders who were mostly asleep an hour before an event. I kept that because it shows the room is not open.

Ask before you film people at private events, and tell them where the video will go. YouTube's privacy guidelines let anyone ask an uploader to remove a video of them, and then file a privacy complaint if that fails, so consent on the day is cheaper than a takedown later. Your output is three or four proof shots per layer, with the people in them aware of the camera.

How the 6 minutes 41 seconds split across the layers

The claim and the promise of three levels 20 s
Level one: open co-working 65 s
Level two: registered events and the hackathon 145 s
The longest layer, because it holds the only competition in the film.
Level three: the invite-only hacker house 95 s
The pitch at demo day 76 s
Not in the three promised levels. It arrives as the payoff.
Approximate seconds per section of the published 6:41 cut, read from the caption timestamps. A map of the edit, not of how long each event lasted.

The turn: the best room was not in the plan

The video promises three levels and delivers four. After the hacker house, I say on camera that there has been a crazy turn of events: I have been invited to pitch my product, so I throw some slides together and go. That demo day, with a speaker describing a proposed visa for digital nomads, is the top of the film, and I did not know it existed when the outline was written.

This is why the structure is a ladder rather than a schedule. A schedule breaks when something unplanned happens. A ladder has room at the top for the one invitation you could not have arranged. The obvious alternative, opening on the demo day and working backwards, gives a stronger first ten seconds and leaves the film nowhere to climb for the next six minutes.

My own limit shows here too. The hackathon footage is the thinnest of the four layers for the obvious reason: I was competing. What saved it was narration recorded in the gaps (corridors, walks between venues, the flat at night), which is why the voice-over sounds like a diary rather than a report.

Step 4: cut from the outline, not the footage

When you cut from the footage, you get the order you shot in. When you cut from an outline, you get the order the story needs, and you find out early which shot is missing. I laid the layers out as beats in a beat sheet before trimming a clip, then graded the footage against it with a story check, which is how the thin hackathon layer was caught while it could still be narrated around.

Scene documentary checklist

  • A count of layers promised in the first 20 seconds
  • Rooms ranked by access: open, registration, invitation
  • A reason to be in each room that is not the camera
  • Wide of every room on arrival
  • Three or four proof shots per layer
  • Consent from the people you film at private events
  • Narration recorded in the gaps, every day
  • Space at the top of the ladder for an invitation you do not have yet
  • Outline laid out as beats before the first trim
Copy it before your first event. A shoot plan, not a guarantee of access.

If it went wrong, it is one of these

  • It plays like a tour. Layers are ordered by date. Re-rank them by access and cut to that order.
  • A layer has no footage. You were too busy taking part. Keep it and carry it with narration and one proof shot; do not drop the layer.
  • It feels like an advert for the city. Statistics are doing the work of scenes. Move them to one voice-over line and put a person on screen instead.
  • The ending is weaker than the opening. The strongest room was used as the hook. Put it back at the top of the ladder.
  • Someone asks to be removed after publishing. Consent was assumed. Ask on the day and say where the video will run.

What carries over to client work

This is the shape a city, an accelerator or an ecosystem programme needs and rarely has: a founder's view of a scene, told with a structure instead of a montage. The access came from taking part, not from a press pass, which is the uncomfortable lesson for organisations commissioning this kind of film. A crew flown in for two days gets the co-working day. Someone already inside the residency gets the residency. The same summer produced three videos from a two-hour pickleball tournament and the surf movie vlog, both from the same inside position.

The whole recipe on one line

Promise a count, rank the rooms by access, enter each one as a participant, shoot proof instead of coverage and leave the top of the ladder open. If your programme has founders in it, the documentary is already half shot; the video work is finding the order.

Sources and further reading

Want a video like this made for your startup?

Video by the studio →Start a conversation →