How long should the indoor TV loop be? Sizing the playlist by the customer journey

Someone signs up for an indoor TV, uploads five videos, leaves it running and, a few weeks later, concludes that "the screen doesn't sell". Almost always the problem is not the content of each spot — it is the size and arrangement of the grid. Three mistakes explain most cases:

1. Loop far too short. Three 20-second spots make one minute of schedule. Someone sitting down for a coffee sees the same sequence five times and stops looking on the second.
2. Loop far too long. A fifteen-minute grid in an operation where the customer stays two minutes: the main offer is only seen by whoever happens to walk in at the right moment.
3. Wrong mix. An ads-only grid is tiring and sends the customer back to their phone; a grid of news and trivia fills the time and sells nothing.

This article solves all three — and it is the direct counterpart to the article on scheduling indoor TV by time of day: that one covers when each grid goes on air; this one covers how long it runs and how it is composed. Together they make the full schedule.

Measure the customer journey before building any playlist

The sentence that organises everything came from the episode behind this article, and it starts by ruling out two very common playlists (translated from the Brazilian Portuguese original):

"The TV schedule is not a music video for the waiter and it is not for you. It is based on the customer journey. The customer stays 5 minutes, so we are going to build a schedule of 7, 8 minutes, 9 on loop."

In other words: grid duration is not the owner's preference or a catalogue number. It is a consequence of a figure that only exists in your store — how long your customer stays. And that figure can be measured in one afternoon, with no tools at all:

• Stopwatch and sample. Time the arrival and departure of 10 to 15 customers at two or three different hours. The average is enough: you are looking for an order of magnitude (2 minutes? 8? 25?), not second-level precision.
• Measure at different times. Dwell time changes by shift. The morning counter, with a queue and people in a hurry, has nothing to do with the afternoon crowd who sit down to work. If the two numbers differ a lot, you have two grids.
• If the store already has cameras, use the footage. Reviewing a recording beats standing there with a stopwatch during the rush. This is our own suggestion as a measurement method: in the episode, cameras come up as a way for the manager to follow footfall remotely and launch a promotion on the spot, not as a stopwatch.
• Measure per screen position, not per store. People walking past the window, people moving down an aisle and people waiting at the checkout have different dwell times at the same address. It is the same segmentation logic as the article on indoor TV by store zone, which maps dwell time per zone — the difference is that there the segmentation is by space and here it is by time.

The maths: dwell time + margin = loop duration

With dwell time measured, the grid falls out of a one-line calculation:

And here is the part almost everyone gets wrong, including people who publish on the topic: the loop has to be longer than dwell time, never shorter or equal. The reasoning is mechanical. If the customer stays 5 minutes and the grid runs 4, at the four-minute mark the schedule restarts while the person is still in front of the screen. They recognise the spot, conclude they have seen everything, and tune out — and the spots that would have come next, including the offer that matters, stop existing for them.

The table below is a starting point — and it is a practical JMV recommendation, drawn from our customers' operations and from what was said in the episode. It is not research data, not an institute study, and it is meant to be calibrated against your own measurement:

The third row is our own extrapolation from the method: in a very long wait there is no grid that never repeats, so the design question becomes how many times the key spot comes back, not whether it comes back. The rule that holds across all three rows is the same: dwell time + margin.

Third-party numbers disagree — and that is the proof of the thesis

Search for how long a digital signage playlist should run and you will find ranges that contradict each other. Adxter, which holds the featured result on Google for that query in Brazil, recommends a 3 to 5 minute playlist for an average viewer of 30 to 60 seconds. Progic states it is rare for a piece of content to hold the screen for more than 20 or 30 seconds. And Google's own AI summary, queried three times on the same day (3 September 2026) with the same question, returned a different range each time: 1 to 3 minutes in one, 8 to 10 in another, and 10 to 15 minutes for a waiting room in a third against 15 to 30 minutes in a fourth pass.

Same question, same day, different answers. That does not weaken the topic — it is precisely the argument of this article. Someone else's range is a reference, not a truth, and none of them knows your store. What works is measuring your own dwell time and calculating your own grid.

"But isn't frequency better than length?"

This is the serious objection, and people do argue for it: instead of a long grid, a lean loop, with the customer seeing the same offer two or three times on purpose. It is the logic of treating the screen as ad inventory, on the out-of-home model: what counts is impact per hour, with an audience that turns over constantly.

The answer is that a store's indoor TV is not ad inventory — it is a conversation with the customer who is there right now. On the street, the audience refreshes every minute and repeating is the only way to reach more people. Inside your store, the same person stands in front of the screen for several minutes; repeating in front of them expands no reach at all, it just burns the time you had to show them something else.

And there is a second argument that usually gets used backwards. People say a long grid is exhausting for those working in the space all day. It is the opposite. The customer spends a few minutes there; the people who watch the entire schedule for an entire shift are your staff. Do the maths: a 3-minute loop repeats about 20 times an hour in front of the same employee — 160 times across an eight-hour shift. A 10-minute loop repeats 6 times an hour. Short loops are what wear the team out; a long grid is exactly what protects the people who work there. That is why so many store screens end up switched off or swapped for broadcast TV — the subject of the article on the switched-off indoor TV and the sales it costs.

One honest caveat, so this does not become a straw man: repeating the key spot inside a long grid is legitimate and solves what the frequency logic wants to solve. The offer of the day can appear three times in 12 minutes, interleaved with different content. What you do not do is shorten the whole grid to get that repetition.

How long each spot can run

Loop duration and spot duration are two different windows, and confusing them is what creates most of the mess around this topic:

• The seconds window belongs to the individual spot. The customer glances at it while doing something else, and either gets it instantly or never gets it.
• The minutes window belongs to the whole dwell time, and that is what sizes the loop.

For the spot, the practical rule this blog already published in digital menu boards for restaurants still holds and is the easiest to apply: if the spot cannot be read in five seconds from three metres away, it holds too much information. Market references for display time land between 10 and 30 seconds per spot — consistent with the five-second rule, because five seconds is the time to read, not the time to display.

With both measurements in hand, the number of spots works itself out: an 8-minute grid with 20-second spots holds about 24 spots. Count by time, never by quantity.

Playlist mix: what is yours and what is filler

Sizing the time solves half the problem. The other half is what goes into that time — and the episode is explicit about the two extremes that do not work. One: "there is no point in signing up for indoor TV just to run news". The other: "if you start cramming in nothing but ads and boring stuff, the guy is going to stay on his phone".

The sequence that works alternates attention and selling:

1. Attention spot. News, product trivia, the story of the item, how it is made. It keeps the customer's eyes from wandering. Automatically updated news covers this without anyone producing anything — as long as it is an ingredient, in the right dose, and not the main course.
2. Product spot. The real item, with a real photo, what is in it, the price. This is the content only you have.
3. Offer spot. Combo, deal of the day, add-on — the mechanics are detailed in upselling triggers at the point of sale.
4. Call to action. What to do now: order at the counter, point your camera at the QR code, join the loyalty programme.

And the rule that closes the section, from the episode: "your content is the soul of your business… your playlist can never be made only from our content". In proportion terms: most of the grid has to be your own content, and supporting content goes between your spots, not in place of them. We do not give an exact percentage here — the episode gives none, and an invented number would be worse than the rule in words.

Producing the spots takes no photographer and no editor: a real product photo taken on a phone is enough, and the platform generates the video. The routes are in indoor TV with artificial intelligence and building the schedule without editing video. If your operation is a coffee shop, the list of what to show is in this article's sibling, digital menu board for coffee shops — both came from the same episode: that one covers the content, this one covers the time and the proportions.

When to change the grid mid-day

Duration sets the size of the grid; the operation decides when it changes. Two triggers are immediate and do not wait for a calendar:

• Stock. Slow-moving product goes into the highlight slot right away; a sold-out item leaves the grid right away. Keeping an offer on screen that the store cannot fill is worse than having no screen.
• Footfall. Traffic below expectation on a given afternoon is a reason for a flash promotion — and the screen is the only channel in the store that changes in minutes.

This only works because the schedule lives in the cloud, and here the comparison with a USB stick becomes concrete. In the words of the episode: "you are not going to take a photo every hour and swap the USB stick every hour. That routine destroys you… that is why so many people just put broadcast TV on and leave it running". A screen that requires someone to physically go there does not get updated — and a grid that never changes becomes wallpaper in two weeks.

Planned changes — which grid runs at 7am, which at 11am, which in the evening — are the entire subject of the article on scheduling indoor TV by time of day. It is worth reading both together: each part of the day can have its own grid, with its own duration, because the dwell time of an 8am customer is rarely the dwell time of an 8pm one.

Mistakes that kill the grid

• A loop shorter than dwell time. The sign error — the most common of all, and the one this article exists to correct.
• Broadcast TV or a video platform running on the screen. That is handing your point of sale over to someone else's communication — and, in the case of broadcast TV, to your competitor's advertising.
• A grid that never changes. It works for two weeks. After that, the regular customer and the staff know it by heart, and the screen drops out of attention.
• A spot with too much text. If you have to stop and read it, it is no longer a screen spot.
• A screen where nobody stops. As the episode puts it, "there is no point in it being hidden in a corner". A perfect grid on a badly placed screen sells nothing.
• A playlist built to please the owner. Or the staff. The grid belongs to the customer journey; if it was assembled around the taste of the people who work there, it is not a schedule, it is background radio.

How to do this in JMV Indoor

Building and sizing the grid in the panel is the easy part — the analogy is from the episode itself: "you know your Spotify playlist? It is exactly the same — video 1, 2, 3, 4". In practice:

1. Order the spots in the sequence you designed (attention → product → offer → call to action), like a music playlist.
2. Add up the durations until you reach the number you calculated. That is how loop size is controlled: by the sum of the spots, not by a "grid duration" field.
3. One playlist per screen. Each screen can run its own grid — the window display does not have to match the checkout — all controlled from the same panel.
4. No per-video charge. There is no cap on playlist items; what limits you is the storage quota of your plan and the number of screens contracted. Current pricing is always on the plans table.
5. Everything from a phone or computer, from anywhere — which is what makes the stock trigger workable.

Questions about the product? Talk to us through the contact form or on WhatsApp at +55 11 4063-8923.

Start the free 30-day JMV Indoor trial →