The outreach email almost every production studio sends reads like this: we're an AI video studio, we do reels and product video, our turnaround is 48 hours, can we book 20 minutes?
It is ignored for a structural reason. It asks the reader to imagine the output. Nobody buys imagination from a stranger when they already have a supplier, a queue and a deadline.
The alternative is to stop describing the work and send it.
What a spec ad actually is
A spec ad is a finished asset, made for a prospect who has not bought anything. Not a mockup, not a showreel with their logo dropped in — a real replacement for one specific ad they are currently running, built to their product and their offer.
It costs you production time. It costs them nothing, including obligation. That asymmetry is the entire mechanism: you have de-risked the first decision to zero, so the only question left is whether the work is good.
The six steps
1. Build a list you can actually inspect. Not a scraped ten thousand. Two hundred targets, chosen because their existing creative is weak in a way you can name. Agencies reselling content, DTC brands with a single tired ad, SaaS companies running a screen recording as their hero.
2. Find what they are running. Every serious advertiser has an ad library. In the EU and UK, the platforms expose all ad types through the transparency tools; in the US you will lean more on manual collection. Either way you are looking for the same thing: one ad carrying the whole account.
3. Pick the weakest one, and diagnose it. Write one sentence about what is wrong — the hook is a logo reveal, the product does not appear until second eight, there is no price. A diagnosis you can defend is what makes the spec land as expertise rather than luck.
4. Build the replacement. Ten to fifteen seconds. Their product, their offer, a hook in the first second, one clear call to action. Build three hook variants if you can — a grid reads as a testing mindset rather than a one-off favour.
5. Send it with nothing attached. No PDF, no calendar link in the first line, no "reply to book". Something close to:
Subject: a replacement for your [product] ad
I saw you're running [describe the ad] — [one sentence on why it underperforms].
Made you this instead. It's yours, use it whether we work together or not: [link]
If you want three more in this style for your other products, I can have them with you next week.
6. Follow up once, on the work. Not "just bumping this". Reply to your own email with a second variant, or a note that you made one for a different product. The follow-up should contain another asset, because the reason they did not reply is almost never that they were offended.
Why this outperforms a pitch
Four mechanics, and none of them are clever:
- It removes imagination from the buying decision. The prospect is not evaluating a promise, they are evaluating a file.
- It proves the diagnosis. Anyone can say "your creative is tired". Delivering the replacement proves you saw something specific.
- It is asymmetrically generous. A £0 obligation on their side, a real cost on yours. That reads as confidence, not desperation — provided you do not then chase them six times.
- It is impossible to compare on price alone. A quote next to two other quotes is a spreadsheet. A delivered ad next to nothing is a decision.
The numbers, honestly
Spec outbound is not free and it is not instant. What I would expect from a 200-target list run properly, with a genuinely good asset per target:
| Stage | Rough rate |
|---|---|
| Delivered (valid address, landed) | 180 of 200 |
| Opened | 60–70% |
| Replied | 5–10% |
| Took a call | 3–5% |
| Became a paying client | 1–3% |
That is two to six clients from two hundred targets. At a $2,045 monthly retainer, the arithmetic is not subtle — but the cost is production slots, not media spend, and production slots are the thing you already own.
Which is also the constraint. This does not scale to thousands of targets, because you cannot make a thousand genuinely good spec ads. It scales to a list you can hold in your head.
The four ways agencies get it wrong
Generic spec. If the ad could have been sent to any of your two hundred targets, it was not a spec ad, it was a sample. The specificity is the product.
Asking for the meeting in the first line. The asset is the ask. A calendar link in line one converts it back into a pitch.
Sending the whole portfolio. One asset, one diagnosis. A folder of work says "browse"; a single replacement says "I looked at you".
Not knowing the ad library. Half the value is the diagnosis, and the diagnosis comes from actually watching what they run. An agency that skips this step is guessing at a hook in public.
Where NNI sits
We are the production line behind this play. Agencies run the outbound under their own name; we build the spec assets at from $70 a finished video, under their brand, with no supplier marks and no contact with their prospect — ever.
If you are running a list of twenty this week and want the assets built alongside your briefs, that is exactly what the 24–48 hour turnaround is for.
---
Jayesh runs NNI Design Studio, white-label video production for agencies. Rate card at nnidesignstudio.com/pricing; rush delivery via Launch Strike.
Drop your take 0
// no hot takes yet. start the thread 👇