Your next paid video campaign may already be sitting in your Instagram grid, wearing a slightly awkward caption and collecting comments from people who keep asking where to buy it.

That does not mean you should download every Reel, slap a logo on it and call it performance creative. Please do not. It means the content you made to earn attention organically may contain the hardest thing to manufacture in an ad: evidence that somebody cared before you paid for the reach.

Google has just made this point rather plainly in its latest guidance for YouTube ads: existing social assets can be a strong starting point, and teams should test deliberately rather than making endless small changes. That is useful advice for a time-poor founder. It also punctures a weirdly expensive habit: treating every channel as if it needs an entirely new creative universe.

Organic engagement is a clue, not a verdict

A post with strong reach is not automatically an ad. A funny behind-the-scenes clip can make people laugh and still be terrible at helping a customer choose you. Likes are applause, not a purchase order.

But organic performance tells you where to investigate. Start by pulling your best 20 posts from the past six months, then ignore vanity metrics for a minute. Look for signals that reveal commercial curiosity:

• saves and shares, especially on explainers or comparison posts;
• comments containing questions about price, availability, fit or delivery;
• visits to the product or booking page;
• direct messages that begin with “I saw your video…”;
• watch-through rate on videos that demonstrate a result, objection or transformation.

The winning asset may not be your most polished one. A trades business might find that a phone-shot video showing the before-and-after of a difficult repair outperforms its glossy brand reel because it makes the value obvious in three seconds. A founder-led skincare clip may work because the founder explains why an ingredient is there without sounding as if they swallowed a brochure.

You are not selecting content because the algorithm liked it. You are selecting it because it contains a proof point worth putting in front of more of the right people.

Find the job the post is doing

Before you repurpose anything, write one sentence: this asset helps a potential customer ___.

Maybe it helps them recognise a problem. Maybe it reduces fear about switching suppliers. Maybe it shows how a product looks on a real human rather than a perfectly lit mannequin. Maybe it gives them a reason to believe your service is quicker, calmer or more considered than the alternative.

That sentence determines the edit.

If the job is to spark recognition, keep the opening tension. “If your invoices are taking Sunday afternoon again, this is probably why” is stronger than leading with your company name. If the job is proof, show the proof earlier: the customer result, the unboxing, the chaos before the system. If the job is confidence, use the clip where someone explains a decision in normal language.

This is where most “repurposing” goes off the rails. Businesses preserve the original post too faithfully. A social video may take 12 seconds to arrive at the point because its followers already know the creator. Paid viewers do not owe you that patience. They need a clear reason to continue almost immediately.

Keep the useful heart; rebuild the runway.

Adapt it for the screen it will occupy

YouTube is not Instagram with a different button. The context is different, the viewing behaviour is different and the technical formats are different. That is not a reason to start from nothing. It is a reason to adapt with intention.

Create a small set of versions from one validated idea. A practical starting set might include a vertical cut for Shorts, a landscape or square version for other placements, and an edit with a different opening hook. Google’s Demand Gen updates have expanded video resizing options, but automatic resizing is not an excuse to ignore framing. A face, product detail or key text can become useless the moment it is cropped badly.

Add captions because plenty of people will encounter the video without sound, but do not turn the screen into a moving subtitle wall. Use on-screen words to carry the thought that matters: “Fits a week of work gear,” “Installed in one afternoon,” “No chemical fragrance.” Let the visual do the rest.

And check your rights. If a customer, creator or freelancer appears in the original social asset, make sure your agreement covers paid use, geography, duration and edits. The sentence “we can repost it” is not the same permission as “we can run it as advertising for six months.” This is boring right up until it is expensive.

Test a clear idea, not forty tiny tweaks

When creative is cheap to make, teams often respond by making a lot of it. Quantity can be useful, but only when you can learn from it.

Pick one variable to test at a time. For the same product demonstration, try a problem-first opening against a result-first opening. Or test a founder speaking directly to camera against the same footage with a customer voiceover. Keep the offer, audience and landing page stable enough that a difference means something.

Google’s current recommendation is similarly restrained: make changes gradually and give the system enough time to measure quality rather than churning through assets for the sake of it. That is sensible. If you replace everything after two days, you have not run a test. You have refreshed your anxiety.

Decide the success measure before launch. For a low-cost product, that may be purchases or cost per purchase. For a higher-consideration service, it could be qualified enquiry starts, booked calls or time spent on a pricing page. “Views” belongs in the story, but it should not be the ending.

Build a content library with paid use in mind

The real win is not one good ad. It is a workflow that keeps producing credible material without sending everyone into a last-minute shoot.

When you plan organic content, tag the purpose of each asset: product demo, customer objection, origin story, social proof, comparison, use case. Save the original clips, clean audio, captions, consent details and performance notes in one place. A simple folder structure will beat a brilliant system nobody updates.

When a post performs, note why. Was it the person on camera? The first line? The contrast between before and after? The fact that it showed a genuine limitation? Those notes make the next shoot better. They also help you avoid copying the surface of a successful post while losing the reason it worked.

This is especially valuable for small teams. You do not need your founder on camera every day, and you do not need a production agency for every promotion. You need a reliable way to spot a useful story, shape it for a paid audience and learn whether it moved a real business metric.

Let social do what it is good at

Organic social is still a place to experiment, listen and be a little more human than a campaign deck usually permits. Paid media is where you can take the clearest of those lessons to a larger, more targeted audience.

They work beautifully together when neither is asked to impersonate the other.

So before you book another expensive shoot, spend an hour with your existing posts. Find the one where a customer’s hesitation visibly disappears. Find the one where the product makes sense without a paragraph beneath it. Find the one people saved because it helped them make a decision.

That is not old content. That is a working hypothesis with a head start.