A genealogy video usually starts with something that genuinely stops people, a face in a photograph from 1890, a handwritten line in a parish register, a surprise in a DNA match list, right, and then the thumbnail either buries that moment under a brown parchment texture and a family tree clip art, or it goes the other way and shouts "shocking DNA secret" in red over a stranger's portrait, and neither one works for long. The quiet version disappears in the feed, and the loud version gets clicks from people who leave in a minute and makes the regular audience, the patient, careful researchers who are the real backbone of these channels, trust the channel a little less every time.
Family history is one of the niches I find most interesting to think about at ClickTheory, because the material is so strong and the audience is so particular. The people watching are researchers themselves, many of them very experienced, and they can smell an exaggerated claim from across the room. So the packaging has to do two things at once, carry the real emotional pull of a found ancestor or a broken brick wall, and stay precise enough that a person who has spent twenty years in archives nods instead of rolling their eyes.
What do genealogy viewers actually click on?
Genealogy viewers click on a specific, recognisable research moment, a particular record, a particular photo, a particular kind of mystery, because they want to learn how it was solved and apply it to their own tree. Vague "family secrets" packaging attracts the curious, but the method is what keeps the researchers.
That splits the audience into two groups, and it helps to package for them separately. The first group is the hobbyist researchers, people who already have an account on a genealogy site and a tree with a few hundred names, and they are searching for help with something concrete, how to read a 1700s will, how to find an immigrant ancestor's village, how to interpret a shared centimorgan number. For them the thumbnail should name the problem in their words and show the evidence, simplified so it reads. The second group is the story audience, people who never touched a census record but love a good reveal, and they click on the human side, the face, the twist, the long lost relative. A channel can serve both, but basically each video should lean one way, and the packaging should make the lean obvious.
The mistake I see most, right, is packaging every video for the story audience because those videos can spike, and then the channel slowly loses the researchers who were watching every upload, commenting with tips and coming back for years. Those are the people who make a family history channel sustainable, and they want the method, so the method needs to be visible in the packaging.
How should old family photos be used in a thumbnail?
Use one face from the old photo, cropped tight and enlarged so the eyes read at phone size, rather than the whole group portrait shrunk into a corner. A single ancestor looking out of the frame does more than a whole family arranged on a porch.
Old photographs are a gift and a trap, right, and the gift is that a real face from a century ago has a pull nothing else in the niche can match, because the viewer immediately wonders who this person was and what happened to them. The trap is that most old photos are small, soft and low contrast, so when you drop a whole cabinet card into a thumbnail it turns into a grey rectangle on a phone screen. So we crop hard, usually to one face or two at most, lift the contrast carefully so it still looks like an old photograph and not a filter, and leave the rest of the frame for one strong element, the record that cracked the case, or a short line of text that names the question.
The other thing that matters a lot, right, is where the old photo sits against modern material. A clean split, the ancestor on one side and the descendant or the researcher on the other, tells the story of a connection across time very quickly, and it gives the viewer a living face to read emotion from, which old portraits rarely give you because nobody smiled for long exposures. If you are cutting people out of old scans, the pitfalls are similar to the ones I described in photo cutouts that look clean, not cut out.
Can you restore or colourise an ancestor's photo for a thumbnail?
Yes, carefully, and you should tell viewers when an image has been colourised or restored, because in this niche the photograph is evidence, not decoration. A thumbnail that quietly invents detail in an ancestor's face is the genealogy version of a fake quote.
Restoration tools have become very good, right, and some of them now fill in features that were never there, sharpening a blurred face into something that looks like a real person who may not resemble the real ancestor at all. For a lifestyle channel that might be harmless, but genealogy viewers care about what is actually in the record, and they will notice. The rule I would hold any channel to goes like this, light cleanup, dust, scratches, contrast, is fine. Colourising is fine if the title or the video mentions it. Generating new detail in a face, or creating a "what they might have looked like" portrait, should be clearly labelled, and it should never be presented as the original photograph. The broader version of this argument is in AI face swaps and the thumbnail trust problem, and it applies with extra force when the face belongs to someone's great grandmother.
In family history the photo is the evidence, so if we change it we say so, every single time.
How do you package DNA results without overclaiming?
Package a DNA video around the specific result and what it suggests, and keep words like "proves" and "revealed" out of the thumbnail unless the paper trail genuinely backs them. Ethnicity estimates are estimates, and match lists are clues, so the packaging should treat them that way.
DNA videos are the most clickable content in the niche and also the easiest to overclaim. "My DNA test revealed I'm 40 percent Viking" is the kind of line that gets clicks and then gets a comment section full of experienced genetic genealogists explaining, politely or not, that ancestry estimates shift between companies and between updates and do not mean what that title implies. And that comment section, right, is a signal to every researcher who reads it that the channel is not careful. The better route is to package the real intrigue honestly, a surprise match you cannot place, an estimate that did not fit the family story, a test that finally broke a brick wall once it was combined with records. The curiosity is still there, it is just pointed at something true.
I am not giving advice about DNA testing itself here, that is outside what a packaging studio should be doing, and if a video makes claims about health traits or medical findings from a test, that content needs its own careful review before anyone writes a title for it. Our job is the click, and the click should match what the video actually shows. The general line between curiosity and clickbait is something I wrote about in clickbait vs curiosity thumbnails, and genealogy is almost a perfect case study for it.
Records, documents and the problem of tiny text
A census page or a ship manifest is the heart of a lot of genealogy videos, right, and it is almost unreadable in a thumbnail, because handwritten script in a dense grid collapses to texture at phone feed size. So we never ask the record to be read. We pick one line, one name, one date or one word that matters, enlarge it until it is legible, and let the rest of the page sit soft in the background as atmosphere. A circled name on a passenger list tells the story in half a second, and the viewer understands this is a video about finding someone in that list.
The same applies to family tree screenshots, which are often the first thing a channel reaches for. A full tree is a lot of small boxes, and the eye has nowhere to land, so if the tree is the point we simplify it to the two or three generations involved and highlight the gap or the new branch. History channels deal with a similar problem when they package archive material, and the approach we use for them, explained on our page about history channel thumbnail design, carries straight over to family history.
Living relatives and the consent question
This is the part of genealogy packaging that nobody talks about enough, right, because a lot of family history videos involve people who are still alive, a newly found half sibling, a cousin who matched, an adoptee looking for birth parents, and putting a living person's face or name into a thumbnail is a very different decision from using a photo of someone born in 1850. I would want any channel to confirm that anyone living who appears in the packaging has agreed to it, and where that is unclear the better move is to design around them, using the record, the researcher's own face or a symbolic image instead.
It is not legal advice, and the rules around privacy vary by country, so a channel working with sensitive adoption or donor conception stories should check with someone qualified. From a packaging point of view, though, at the end of the day it is also about trust, because an audience that sees a channel treat living people carefully believes the channel will treat their own stories carefully too. Does that distinction between historical subjects and living ones make sense for your channel? The packaging should reflect it.
Who is a packaging studio not right for in family history?
If your channel is a personal archive, videos made mostly for your own extended family to watch and share, you do not need a studio, and I would honestly tell you to keep it simple. Your relatives will click on anything with grandma's photo on it, and that is lovely just as it is.
The same is true for very small research channels just starting out, where learning to make your own thumbnails is part of learning what your audience wants. The built in Ask Studio feature in YouTube Studio will even propose AI made thumbnail ideas, and playing with those is harmless, as long as nothing it generates passes itself off as a historical photograph. A template pack can also work if you accept that it will look like the other channels using the same template. Where a studio starts paying for itself is when the channel is publishing regularly, has a back catalogue of tutorials and story videos, and is trying to reach researchers who do not know it yet.
What ClickTheory needs from a genealogy channel to start
Send us the video or a summary of the find, the photos and records you are allowed to use at the highest resolution you have, and a note on anything that must not be shown. For every video the work begins with the title, a strategist drafts something like twelve versions modelled on formats with a proven record, the designer turns the best into a few rough frames, and only then does anything get polished, so the title and the image are designed together rather than one after the other.
For a family history channel the starting point is one pack, which costs 250 dollars and covers eight videos, revisions as many times as the research demands, and a finished thumbnail lands between 48 and 72 hours after the brief, not counting weekends. The rest of the pricing is on the services page. Every final can ship with an alternate for YouTube's built in A/B test of titles and thumbnails, and the results go into a playbook for your channel, so over time you learn whether the old face, the record or the researcher earns the click for your audience. If your channel leans more toward long story videos, our page on documentary channel thumbnails and the piece on why quiet packaging wins for documentaries are worth reading next. The thing I want a genealogy channel to take from all this is that you do not have to choose between honest and clickable, because in this niche the honest version, done with enough craft, is the one that keeps getting clicked by the people who matter. So yeah. That's my way of saying it.