How to build fake YouTube analytics for a thumbnail

The line you don’t cross

A thumbnail is a promise about the video behind it. On a skit, a parody or an explainer built on example figures, a simulated analytics screen makes a promise the video keeps. On a video that claims to show your own results — “how I made $38K in a month” — invented figures make a promise it can’t, and when that video sells a course, coaching or a channel method, it’s exactly what our acceptable use policy prohibits. Build thumbnails for the first kind of video.

Use Analytics, not the dashboard

The dashboard is made to be read rather than glanced at: a subscriber count, four small tiles, a list — and no chart, because like the real app it keeps the charts one tap away. The Analytics screen is the one to build a thumbnail from. Each metric gets a card to itself: the figure in the largest type on the screen, a badge beside it, the gap from usual spelled out beneath, and a daily chart at the bottom with its axis and dates. One number and one line is about as much as a thumbnail can carry, and that card is exactly that.

The exception is a subscriber milestone. The dashboard’s headline prints the total in full, every digit with its commas, which is what a milestone gag wants.

The Realtime panel, lower down on Analytics, is a second option: Last 48 hours, a single figure and twenty-four bars in your Accent colour. It reads as right now rather than this month, which suits a skit about a video taking off overnight, and it always follows whichever metric you chart.

Choose the metric the video is about

Charted metric decides which card the row starts on, and that first card is the only one fully on screen: the next peeks in at the right edge, and four page dots underneath count the set. Set it to what the video is about — Estimated revenue for a money skit, Subscribers for a growth joke, Views for the upload that took off.

Then set that metric’s verdict. More than usual puts a green up-arrow badge beside the figure and prints the gap in green beneath it. Same as usual gets a green tick and prints in grey. Less than usual gets a grey badge with a down arrow and prints in grey. The green line is the only coloured text on the card, so on a good-news thumbnail it’s the second thing anyone reads.

Mind the length of the figure as well. The count cards shorten — 4.3M, 284.5K — but the revenue card prints every digit and the cents, so $38,420.75 is ten characters against 4.3M’s four. A money thumbnail needs a tighter crop around the figure to read at the same size, and the badge beside it should stay in the crop: it’s the verdict.

Gap vs usual sets the figure in that line, in the charted metric’s own units: the default 1,340,000 views prints as 1.3M more than usual. Change the charted metric and you need to retype the gap in the new units. Switch to Estimated revenue and leave 1,340,000 in place, and the revenue card reads $1,340,000.00 less than usual while the Views card claims 149.8M more; type a gap in money instead — 6,000 against $38,420.75 — and the cards fall back into proportion.

Shape the curve

Views trend is twelve bars in the editor. The app turns them into twenty-eight daily readings, adding a weekly rhythm and a small wobble that come out the same every time, so the line looks jagged like a real daily chart rather than smooth like a drawing of one. It scales the curve so the days add up to the card’s figure, then rounds the top of the axis to a number a person would have picked: 400.0K for the default views, not whatever the peak happened to be.

Drag the bars up or down, with a mouse or a finger, to set the shape; the trend up and randomize buttons under them give you a starting point. For a thumbnail, pick one idea and make it obvious:

• The spike. A low start, one or two tall bars, then a slow settle: the upload that took off. The default is this shape, jumping at the third and fourth bars.
• The climb. Bars rising from left to right. Trend up gives you one.
• The cliff. Tall early bars falling to a floor, for the video about it all going wrong. The badges won’t turn red for you, so the line has to carry that on its own.

Two constraints come with it. Every card uses this same shape, scaled to its own total, so a curve drawn for views also has to make sense for revenue. And the chart’s only time labels are Chart starts and Chart ends — 24 Aug and 20 Sep by default — so change them to dates that suit the story, four weeks apart for Last 28 days. The Realtime panel under the cards is drawn from the last three bars, so ending the curve on a rise puts a rise there too.

Colour and contrast

• Accent colours the chart line, the realtime bars, the avatar and the active tab — #8b5cf6, a violet, by default. Choose it against whatever the thumbnail puts beside the screen, and not in the colour of your thumbnail text, or the two will compete.
• Background has to stay dark. The app’s type is white, so a light background washes it out; the default is #0b0b0d.
• The greens are fixed. The more-than-usual line and badges are always the same green, and the analytics screens never use red on their own: a shortfall is grey. If the story needs a warning colour, the chart line is the place for it, through Accent.

Get the image

For a thumbnail with no phone in it, build the screen in the editor, tap Analytics in the preview — it switches screens the way the app does — and use Export PNG (3x). You get the screen on its own, 960 pixels wide, and unless you’re signed in on Pro it carries the SIMULATED · POZER.COM mark exactly where the preview shows it; the field-by-field guide to the template covers the export in more detail. For a thumbnail with a hand holding a phone, install the app and photograph the phone itself, or take a screenshot on it for a flat image with the phone’s own status bar. On the free version the same mark is on that screen too.

Crop to 16:9 and test it small

YouTube recommends thumbnails at 1,280 × 720. A 16:9 band across the full width of the export is 960 × 540, which you scale up by a third to reach 1,280 wide:

Full-width 16:9 crop of the export = 960 × 540 px
Scale to 1,280 × 720 = × 1.33

A tighter crop means scaling further, and the type holds up better than the thin gridlines and axis labels do. That full-width band is also shorter than the card, so a crop that fills the frame can’t hold all of it. Keep the figure, the verdict line and the peak of the curve — and if your export carries the mark, it stays in: it’s what tells a viewer the screen is simulated, and cutting it off defeats the point of it. When everything won’t fit a crop, don’t fill the frame with the screen at all: set the whole card on the thumbnail’s own background, beside a face or a title.

Then look at it the way viewers will. Shrink the finished thumbnail to the width it gets in a phone’s feed and check two things: the figure still reads, and the line still clearly goes up or down. If either fails, crop tighter or take something out. Everything above starts from the defaults on the YouTube Studio-style template page.

What gives a thumbnail away

1. A figure that disagrees with the title: $38,420.75 on the card and “$50K” in the title text.
2. A curve that argues with its verdict: a line sliding downhill under “1.3M more than usual”.
3. Chart dates that don’t fit the story: a video about last week under a chart that ended a month ago.
4. A light background washing out the white type.
5. A crop scaled up until the gridlines blur.
6. A figure cropped away from its badge, which leaves half the message behind.
7. A thumbnail promising a real result that the video doesn’t show.