Most media companies I talk to have the same problem, and it is never a talent problem. They have four to nine channels, a real production calendar, an editor bench, and a library of two thousand videos. What they do not have is one person who owns the thumbnail and the title as a single decision. So packaging gets done last, by whoever is still online at 9pm, on a template someone built in 2023.
That is where the money leaks. A media company publishing sixty videos a month is making sixty click decisions a month, and if the packaging is average on all sixty, the whole slate underperforms quietly. Nobody gets fired for it. The graph just flattens.
I run ClickTheory as a YouTube packaging studio, which means we design the thumbnail, the title, and the first-frame promise together, as one unit. This post is about what changes when the client is a media company rather than a solo creator, because the constraints are genuinely different and most agencies pretend they are not.
Why does packaging break down when a media company scales?
Packaging breaks at scale because it is the only part of the pipeline that never got a dedicated owner. Every other role got one. You have a head of production, a post supervisor, a channel manager, maybe a data person. Packaging sits in the gap between the editor and the channel manager, and gaps do not have standards.
The symptoms are predictable once you have seen a few:
- Thumbnails that look like the show, not like the video. Branding wins, curiosity loses.
- Titles written from the episode title in the CMS instead of from the viewer's question.
- Six channels sharing one template, so a viewer who clicks on one cannot tell the others apart.
- No one comparing the packaging of a hit to the packaging of a flop, because nobody owns the comparison.
There is also a structural issue. In a media company, the person who knows the content best is the producer, and the person who knows the design tools best is the editor, and neither of them is measured on click through rate. A packaging agency exists to be the one party in the room whose entire job is that number.
The fastest CTR gain in most media companies is not a new design style. It is one person, inside or outside, who is allowed to say no to a thumbnail.
What a media company actually buys when it hires a packaging agency
You are not buying pictures. You are buying a repeatable decision process that runs sixty times a month without you in the loop.
In practice that means four things. First, a house point of view on what each channel's audience clicks, written down, not vibes. Second, a designer who has already seen your last two hundred thumbnails and knows what has been used to death. Third, titles that were chosen against the thumbnail rather than after it, so the two are not repeating the same information. Fourth, a turnaround you can put in a production schedule.
Our numbers on that last point: 48 to 72 hours per thumbnail, one thumbnail per video, unlimited revisions until it is right, and a rush option at plus $20 when an episode gets pulled forward. A dedicated designer, not a queue you get assigned to randomly. Title consultation is part of the package, not a line item, because a thumbnail priced without the title is priced for the wrong job.
Everything is custom per video. We do not run templates, and we test styles on our own channels before they get near a client slate. If you want the reasoning on the specific style questions we argue about internally, our notes on thumbnail text best practices cover the part people get wrong most often, which is word count.
How do you keep six channels on-brand without making everything look the same?
You separate the parts of packaging that should be consistent from the parts that should not, and you write the line down.
Consistent across a media company's channels: legibility rules, safe margins, how faces are cropped, how much text is allowed, contrast floors on mobile. These are craft standards, and standardising them costs you nothing creatively.
Not consistent: colour, subject framing, emotional register, and the specific curiosity mechanic. A business news channel and a lifestyle channel under the same roof should not share a palette. If a viewer sees your two channels side by side in a suggested feed and cannot tell them apart, you have built a template, not a brand system.
The practical version of this is one style sheet per channel plus one craft standard for the group. Then when a new show launches, you are deciding one thing, the style sheet, instead of relitigating every thumbnail. This is the same discipline that makes packaging work for personal brands transfer badly to media companies if you copy it directly. A single creator can carry a look with their own face. A network cannot, because the face changes every show.
Worth also watching what is actually moving in the format itself. We keep a running read on where thumbnail styles are heading, and the honest summary is that the big shifts happen roughly twice a year, not monthly, so chasing every trend is a good way to spend money on nothing.
Why is the back catalogue the cheapest growth a media company has?
Because you already paid for it. A media company with two thousand published videos is sitting on inventory that costs nothing more to make and is still being served by the algorithm every day.
Here is the reasoning I would apply to a library. Pull the videos with strong average view duration and weak click through rate. Those are videos the audience likes once they arrive, and does not arrive at. That gap is a packaging problem almost every time, and it is the only kind of underperformance you can fix in 72 hours instead of six weeks.
Then be honest about the ones you should leave alone. If the retention is bad too, a new thumbnail brings more people to a video that will disappoint them, which is worse than leaving it. Repackaging is a magnifier, not a repair.
For a media company the arithmetic is unusually good. Thirty old videos repackaged is thirty shots at recovering traffic you have already earned, against the cost of maybe one and a half new productions. I have yet to see a library where that trade was a bad idea. YouTube's own creator resources are reasonable on the fundamentals here, though they stay deliberately vague about the design decisions, which is the part that actually moves the number.
What should a media company pay for thumbnails?
Our pricing is $250 for 8 custom thumbnails, which works out to about $31 each, with unlimited revisions and a dedicated designer. Add-ons are $15 for an A/B variant, $25 for a title pack, and $20 for rush.
I will be direct about why that number is what it is. Cheaper exists, and it is almost always template work sold by volume, where you get a layout with your face dropped into it. More expensive exists too, and above a certain point you are paying for account management rather than design hours.
For a media company, the relevant comparison is not agency versus agency. It is agency versus a full time designer. A designer on staff costs more, takes months to hire, and gives you one taste across every channel. An outside studio gives you a wider reference set and no headcount, but you have to brief properly. Both are legitimate. The honest test is volume: past roughly eighty thumbnails a month with strong internal art direction, in house starts to win. Below that, it usually does not. There is a longer version of that argument in our piece on when to hire a packaging agency, including the signals that mean you are not ready yet.
Whichever way you go, price the decision, not the file.
How do approval chains kill click through rate, and what fixes them?
Approval chains kill CTR by averaging. Every extra reviewer pulls a thumbnail toward the safe middle, because safe is the option nobody has to defend.
I have watched a strong concept go through four rounds of stakeholder notes and come out as a centred logo on a gradient. Nobody chose that. It was the only option that offended no one.
Three things fix it, and none of them are software:
- One decider per channel. Not one approver, one decider. Others advise.
- Reject on a rule, not on taste. "Text is unreadable at 168 pixels wide" is a note. "I do not love the orange" is not.
- Cap the rounds. Two rounds of notes, then the decider picks. Unlimited revisions on our side means the craft can keep improving. It does not mean the decision should stay open forever.
If your organisation cannot do point one, buy the A/B variant instead and let the data settle the argument. That is what the $15 is for.
How does a remote studio fit into a media company's production week?
It fits by working to your publish calendar rather than your meeting calendar. We are based in Dubai on GMT+4, working worldwide and async, which in practice means European teams get overnight turnaround and North American teams wake up to work that happened while they slept.
For a media company that runs a weekly slate, the useful pattern is batching. Send the week's eight videos together with the raw episode files or a few frame grabs, get the set back within 48 to 72 hours per thumbnail, and review as a set rather than one at a time. Reviewing as a set is also how you catch the sameness problem before publication.
We have written up the mechanics of how that collaboration actually runs in working with ClickTheory remotely, and if your team sits in Asia specifically, the timezone notes in our Hong Kong packaging piece are more concrete.
One more thing that matters to media companies: we never show client names. Case discussion happens by niche and subscriber band only, never by brand. If you are a network with talent contracts and PR sensitivities, that is not a nice to have. And if your slate includes long form interview shows, our sister service edits video podcasts under the same working rhythm.
Where would I start with a library of two thousand videos?
I would start with eight. Not eighty.
Pick four upcoming episodes and four back catalogue videos with good retention and weak CTR. Run them as one batch. In two weeks you will know whether the packaging point of view fits your audience, and you will have a like for like comparison against your own historical numbers rather than someone else's case study.
That is deliberately a small commitment. Packaging is a claim about what your audience wants, and claims should be tested cheaply before they become a system. You can see the range of work on our portfolio first if you want a sense of the styles before committing to anything.
If you run a media company and packaging is currently nobody's job, send me your channel list at samy@pixelsamy.com and I will tell you honestly whether an outside studio is the right call for your volume.