The YouTube Ad Budget That Pays You Back Three Months Later

Here’s the pattern I see most often with YouTube.

A business runs ads for a month. The video gets watched. The dashboard shows two conversions, maybe three. The cost looks embarrassing next to Search.

They switch it off and conclude YouTube doesn’t work for their industry.

Then something odd happens about eight weeks later. Branded searches tick up. A few enquiries arrive from people who “have seen you around”. The Search campaign quietly gets cheaper, because more people now type the company name.

Nothing in the reporting connects any of that to the money spent in month one. So it gets filed as luck, or a good quarter.

The payback was real. The reporting window had already closed before it arrived.

That’s not a conspiracy, and it isn’t Google hiding anything. It’s arithmetic, and once you understand it, the whole channel makes more sense.

WHY THE NUMBERS DISAPPOINT IN MONTH ONE

Three things stack up against you at the start, and they’re all structural.

The attribution window is shorter than the buying cycle

Somebody sees your ad, doesn’t click, and converts later. Google records that as a view-through conversion. It only counts inside a lookback window, commonly seven to thirty days.

Now think about how people buy a commercial service. They notice you. They do nothing. Then something breaks, a contract ends, or a budget opens. Only then do they go looking.

If that gap is longer than thirty days, the conversion is never credited to the ad. It shows up as a branded search, or a direct visit, and Search takes the credit.

You aren’t looking at a failed campaign. You’re looking at a report that can’t see that far.

The learning phase eats the first fortnight

Conversion campaigns need volume before they optimise properly. Two to four weeks is the usual allowance before performance settles.

If you run a four-week test, you’ve measured the learning phase and almost nothing else.

The audience doesn’t exist yet

This is the one that does the most damage, and it’s entirely fixable.

In month one, you’re advertising to strangers. There’s no pool of people who’ve watched your video, because you’ve only just started making one.

The conversions in this channel mostly come from the second pass, not the first. Month one builds the audience. Month two starts converting it. Month three is when the numbers look like anything.

Run one month, and you’ve paid for the expensive part and stopped before the part that pays.

WHAT YOU ARE ACTUALLY BUYING

Search captures demand. Somebody wants a thing, types it, and you bid to be there.

YouTube creates demand. Somebody wasn’t thinking about you at all, and now they know you exist.

Those are different products, and it’s a mistake to price them the same way.

Here’s the practical version. A click from Search is worth a lot, because the person is already in motion. A view on YouTube is worth far less on its own. That’s why it costs a fraction as much.

The value comes from volume and from what you do next. Thousands of cheap views, a slice of whom you can identify and speak to again.

You aren’t buying conversions on YouTube. You’re buying an audience you can convert somewhere cheaper.

Businesses that treat it as a conversion channel get disappointed. Businesses that treat it as an audience-building channel with a conversion layer bolted on tend to do well.

THE FORMATS, AND THE TWO THAT MATTER ON A SMALL BUDGET

YouTube sells six formats. Skippable in-stream, non-skippable in-stream, bumper, in-feed, Shorts and Masthead. They’re bought at auction, priced on cost per view or cost per thousand impressions.

You can ignore most of that list.

Skippable in-stream is your default

This is the one people can skip after five seconds, and it’s the workhorse for a small budget.

The reason is the pricing. On a cost-per-view basis, you pay for engaged views, not for the people who skipped. Anyone uninterested costs you nothing.

That makes it unusually forgiving. A mediocre video wastes far less money here than it would in Search.

In-feed is the one people underuse

In-feed ads appear next to search results and suggested videos on YouTube, so the viewer chooses to click.

That makes them strong when your video answers a question people look up. “How much does commercial waste collection cost” is a video somebody goes hunting for. An advert about your company values is not.

Bumper, non-skippable and Masthead

Bumpers are six seconds, unskippable, sold on impressions. They’re a reminder format. Good for staying in front of an audience that knows you. Poor for introductions.

Non-skippable in-stream buys attention you haven’t earned, which suits brands with reach budgets.

Masthead is the banner across the YouTube home page and is priced for national advertisers. If you’re reading this, it isn’t for you.

WHAT IT COSTS

Published ranges put cost per view for skippable ads at roughly one to ten cents. Competitive sectors such as finance and B2B run higher, often ten to thirty cents. CPMs usually sit around five to ten dollars.

Treat those as orientation rather than a quote. Your market and your targeting set your own numbers.

The comparison that matters is against your Search costs, not against other YouTube advertisers. If Search charges you six pounds a click, views at a few pence change what a modest budget can do.

There is a floor to be aware of. Demand Gen campaigns need enough volume to learn, and budgets under about thirty dollars a day tend to underperform for that reason.

That doesn’t mean you need thirty dollars a day to advertise on YouTube. It means something narrower. Go well below it on a conversion-optimised campaign, and you’re asking an algorithm to learn from almost no data.

For a genuinely small budget, a view-focused campaign at a lower daily spend is the more honest option. You’re buying audience, and you know it.

THE SEQUENCE THAT MAKES IT WORK

This is the part page one leaves out, and it’s the part that decides everything.

YouTube works in three steps, in order.

Step one: reach the right strangers

Your targeting has to be tighter than “people interested in business”.

The strongest option is a custom segment built from your own search terms. Our post on custom segments covers that in full. You take the terms that already produced enquiries in Search, then reach those people on YouTube instead.

That gives you strangers with proven intent rather than strangers with a vaguely relevant interest.

Step two: capture everyone who watched

Link your YouTube channel to Google Ads, then build an audience of viewers in Audience Manager.

Base it on a real watch threshold, not an impression. Somebody who watched a meaningful chunk of your video is a different prospect entirely. Somebody whose browser merely loaded it is not.

This audience is the actual asset. Everything in month one exists to fill it.

Step three: convert them somewhere cheaper

Now retarget that audience. Search, Display, Demand Gen, or another video.

These people know your name, so they respond at a rate strangers never will. They also cost less to reach, because the audience is small and specific.

Skip step two, and you’ve built a leaky bucket. You pay to introduce yourself, then let every introduction go cold, then pay to introduce yourself again.

Set up the viewer audience before you launch the first campaign, not after it.

It takes ten minutes, and it’s the difference between an advert and a funnel.

HOW TO BUILD YOUR FIRST CAMPAIGN

The build itself is straightforward once the thinking is done.

Link your YouTube channel to Google Ads first. Without it, you cannot build a viewer audience, and the whole sequence falls apart.
Create the viewer audience before launch, in Audience Manager, based on a meaningful watch threshold rather than an impression.
Choose a video campaign with skippable in-stream, bought on cost per view. Do not start on a conversion objective with a small budget and no data.
Build your targeting from a custom segment made of your own converting search terms. Proven intent beats a broad interest category every time.
Turn optimised targeting off while you are testing, so you are measuring the audience you chose rather than one Google widened for you.
Put the offer in the first five seconds. On a skippable format, the skip button is the real competition, not another advertiser.
Set a flat daily budget and leave it alone for at least six weeks. Changing budget weekly restarts the learning and teaches you nothing.
Build the retargeting campaign in week three, once the viewer audience has people in it.

Two things not to do. Do not run one video against everybody, because you will never know whether the audience or the creative failed. And do not repurpose a two-minute company overview film, because nobody skipped past a cat video to watch your reception area.

MAKING THE VIDEO ITSELF

You don’t need a production company. You need the first five seconds to earn the next ten.

Lead with the problem, not your name. “Your waste collection bill went up in April” holds a viewer. “Welcome to our company” loses one.

Say who it’s for early and plainly. Naming your customer filters out everyone else, and on a cost-per-view basis, being skipped by the wrong person is free.

Keep it short. Fifteen to thirty seconds is plenty for a first test, and short videos are cheap to remake when the first one underperforms.

One idea per video. If you want to say three things, make three videos and find out which one works.

Film it on a phone in decent light if that’s what you have. Production value matters far less than the first line.

What goes in the first five seconds

On a skippable ad, five seconds is what you have bought outright. Everything after that is optional, and the viewer decides.

So the five seconds have one job. Tell the right person that this is about them.

Name the problem out loud. Not the category, the problem. “Your energy contract renews in the autumn” beats “we are an energy consultancy”.

Or name the person. “If you run a care home in the North West” loses everyone else instantly, which is exactly what you want when the wrong viewer costs you nothing.

Avoid three things in the opening. Your company name, a logo animation, and any sentence starting with “we are”. All three spend your only guaranteed seconds on information the viewer has no reason to want yet.

A useful test before you film. Read your first line aloud to somebody who does not work with you and ask what they think it is about. If they say “an advert”, rewrite it.

WHY IT GETS CHEAPER THE LONGER YOU RUN IT

There’s a compounding effect here that Search doesn’t have, and it explains why quitting early is so costly.

Every month you run video, the viewer audience grows. That audience is the cheapest place to advertise you will ever own, because it’s small, specific, and nobody else is bidding for exactly those people.

Month one, you have nothing to retarget. Month six, you have a standing audience of people who have watched you explain something.

Branded search compounds too. As more people know your name, more of them search it directly.

Branded terms are usually the cheapest clicks in any account, and they convert far better than generic ones. So a slice of your Search spend quietly shifts from expensive terms to cheap ones.

That saving lands in the Search campaign’s column, not the video campaign’s. Which is the whole problem with judging channels in isolation.

The businesses that get the most out of YouTube tend to be the ones who kept a modest budget running for a year, rather than the ones who spent a lot for a month.

WHAT ABOUT SHORTS

Shorts deserve their own note, because the advice on them ages badly and most guides treat them as a footnote.

Vertical video sits in a different part of YouTube. People scroll it fast, with a thumb ready to move on, in a mindset closer to social feeds than to search.

That changes what works. A talking-head introduction that survives on desktop dies in a feed. The first second has to earn the second one.

It also changes the economics in your favour. Shorts inventory is usually cheaper to reach than in-stream, so a small budget buys more impressions.

Here’s the sensible way to use them. Treat Shorts as the top of the funnel and in-stream as the middle. Shorts build the viewer audience quickly and cheaply. In-stream and in-feed do the explaining once people know your name.

One practical warning. Do not upload a horizontal video and let it letterbox. A horizontal clip with black bars top and bottom reads as an advert instantly, and people scroll past adverts.

If you’re filming anyway, film once and frame twice. Shoot with enough headroom to crop to vertical, and you get both formats out of one afternoon.

A WORKED EXAMPLE, START TO FINISH

A hypothetical business makes this easier to follow. Take an accountancy practice on nine hundred pounds a month, all of it currently in Search.

Search works for them. It also costs them around eleven pounds per click on terms like “accountant for limited company”, and the budget is gone by the third week.

Month one, they take three hundred pounds out of Search and put it into video.

They film two clips on a phone. One answers “how much should a small limited company pay for an accountant”. The other answers “when should I switch accountants”. Both run about forty seconds.

Targeting comes from a custom segment built on the twenty-odd search terms that produced real enquiries last quarter, plus two distinctive competitor names.

Before launch, they link the channel and build a viewer audience. This takes ten minutes, and it’s the whole point of month one.

What month one looks like: roughly nine thousand views, two enquiries, and a cost per conversion that looks dreadful beside Search. This is the moment most businesses quit.

Month two, they leave the video running and add a small retargeting campaign pointed at the viewer audience. The audience now has a few thousand people in it who watched a meaningful chunk.

Retargeting costs are low. The audience is small, and nobody else bids for exactly those people. Enquiries start arriving that mention “I saw your video”.

Month three is where the judgement happens. Branded search impressions are up against the baseline from month one. Two or three enquiries have arrived through Search from people who saw the video weeks earlier.

Their attributed conversions still understate it, because the view-through window closed before some of those people acted.

So they judge it on the total. Nine hundred pounds spent on video and retargeting across the quarter, against the qualified enquiries they can fairly connect to it, plus the branded search lift.

If that total beats what nine hundred pounds of extra Search would have bought, they keep going. If it doesn’t, they stop. Either way, they learned something real rather than something premature.

THE MISTAKES THAT WASTE THE MOST MONEY

A short list, in the order I see them.

Running for one month and calling it a verdict. You stopped before the retargeting layer existed.
Never building the viewer audience, so every pound introduces you to somebody and then loses them.
Judging the channel on last-click conversions, when the whole point is the conversions it assists.
Uploading a two-minute company overview film and expecting strangers to sit through it.
Targeting a broad interest category instead of building an audience from your own converting search terms.
Changing the budget every week, which restarts the learning and guarantees you never see settled performance.
Letting a horizontal video letterbox into Shorts, which announces itself as an advert before the first word.
Running video while your Search campaigns still lose impression share on the terms that already make you money.

HOW TO MEASURE IT HONESTLY

You need a method that works inside the attribution limits rather than pretending they don’t exist.

Watch branded search volume

This is the clearest signal that YouTube is doing its job. If people are seeing you, more of them start typing your name.

Take a baseline before you launch. Check it monthly in Search Console and in the branded terms in your Search campaigns.

Rising branded searches during a video campaign is the effect, showing up in the only place it can.

Look at assisted conversions, not last click

Your conversion paths will show YouTube appearing early in journeys that closed elsewhere.

That’s the channel doing what it’s for. Judging it on last click is like judging a shop window on how many people bought from the window.

Track view-through conversions, but know their limit

They’re useful, and they’re real. They also stop counting when the window closes, so they undercount by design on a long buying cycle.

Use them as a floor, never as the whole picture.

Compare cost per qualified enquiry over a quarter

At 90 days, add up what the video and retargeting campaigns cost you. Set that against the qualified enquiries you can fairly attribute to them.

Qualified is doing the work in that sentence. Prospecting reaches people earlier, so a share of the response is always curious rather than commercial.

And check what happened to your Search costs. A stronger brand often makes Search cheaper. That saving never appears in the video campaign’s own column.

WHEN YOUTUBE IS THE WRONG CHOICE

Some businesses shouldn’t run this, and the guides selling YouTube management rarely say so.

Skip it if your Search campaigns still lose impression share on your best terms. Capture the demand that already exists before you pay to create more.

Skip it if you can’t fund at least three months. A one-month test on this channel is money set on fire, because you stop before the part that pays.

Skip it if you can’t handle the enquiries you already get. More awareness makes a follow-up problem worse, not better.

Be careful if you sell something genuinely urgent. Emergency plumbing gets searched, not browsed. There’s still a brand case, but it’s a slower one.

And be honest about the video. If nobody in the business will get in front of a camera, and there’s no budget to make something decent, the channel isn’t going to save you.

FREQUENTLY ASKED QUESTIONS

How much do YouTube ads cost for a small business?

Skippable in-stream ads are usually bought on cost per view. Published ranges sit at roughly one to ten cents a view, rising to ten to thirty cents in competitive sectors such as finance and B2B. CPMs commonly land around five to ten dollars.

You only pay for engaged views, so a small budget goes further here than it does in Search.

Do YouTube ads actually work for small businesses?

They work as a way to create demand rather than capture it.

Almost nobody watches an advert from a company they’ve never heard of and buys that afternoon. The return arrives as more branded searches, warmer enquiries and a retargeting audience you convert later.

Judged on first-month last-click conversions, it will nearly always look like a failure. That’s a measurement problem rather than a channel problem.

How long before YouTube ads start working?

Allow two to four weeks for a conversion campaign to leave the learning phase. Then allow another month for the retargeting audience to fill up enough to matter.

Ninety days is a fair point to judge it. Anything shorter measures the setup.

What is a view-through conversion?

It’s recorded when somebody sees your ad, doesn’t click, and converts later.

Google counts it inside a lookback window, commonly between seven and thirty days. A conversion that lands after the window closes is never attributed to the ad.

That single fact explains most of the disappointment in this channel.

How do I retarget people who watched my video?

Link your YouTube channel to Google Ads, then build a viewer audience in Audience Manager.

Base it on a meaningful watch threshold rather than an impression, so you collect people who genuinely watched rather than people whose browser loaded the file.

Then point a retargeting campaign at that audience. That’s where most of the conversions actually come from.

Which YouTube ad format should I use?

Skippable in-stream on cost per view is the sensible default for most small businesses, because the people who skip cost you nothing.

In-feed is worth testing when your video answers a question people search for. Bumpers suit reminding an audience that already knows you. Masthead is priced for national advertisers.

Should I run YouTube ads or Search first?

Search first, almost always.

Search captures people already looking for what you sell, so it’s the cheapest demand in the market to win. If you’re still losing impression share on your best terms, that’s where the next pound belongs.

YouTube earns its place once Search is capped. At that point, you’ve bought all the existing demand you can, and the only way to grow is to create more of it.

The exception is a genuinely new product nobody searches for yet. If there’s no demand to capture, creating it is the only option you have.

Can I run YouTube ads without making a video?

You need a video, but you don’t need a film crew.

A phone, decent light and a clear first line will outperform a polished corporate overview most of the time. The first five seconds decide it, not the production budget.

WHERE TO START THIS WEEK

Don’t launch anything yet. Do the two things that make a launch worth doing.

First, link your YouTube channel to Google Ads and build the viewer audience. It takes ten minutes, it costs nothing, and without it, the channel can’t compound.

Second, take a baseline of your branded search volume. Open Search Console, note the impressions and clicks on your company name for the last 90 days, and write the number down.

That number is how you’ll know whether any of this worked, because it’s the one place the effect reliably shows up.

Then commit to a quarter or don’t start. YouTube rewards the businesses that stay long enough to build an audience, and it punishes the ones who test it for four weeks and read the learning phase as a verdict.

Did You Enjoy This Blog Post?

I hope you enjoyed this blog post, and thank you so much for being here. We also upload videos to our YouTube channel every weekday. Please subscribe so you are one of the first to be notified.

If you enjoyed this blog, you may also like:

SF Digital Studios