Straight answers

A YouTube ads tool for podcasts and long-form videos. It runs your episodes as in-feed ads in your own Google Ads account and steers the budget by one number: what a completed hour of watching costs you.

Not views. Completions — the people who stay to the end.

Three things follow from that: completed hours instead of opens, reach into audiences organic doesn’t get to, and output steady enough to plan around.

Anyone publishing long-form video on YouTube. The common thread is the format, not the label: videos of seven minutes and up, where someone finishing one actually means something.

Two groups make up most of it:

Inside those two, it fits particularly well for:

Three things have to be true: ten or more videos on YouTube, $500–750 a month of ad budget, and promotion you intend to keep running.

No. Connect your Google Ads account, set a budget, apply the Actions Hourcast hands you. Setup takes 10 minutes.

Every step has a guide behind it, and the dashboard flags what’s missing before it costs you anything.

No. It’s a tool, running in your own Google Ads account on your own budget. No retainer, no account handover, nobody between you and your spend.

What the campaign gets steered on.

Google Ads optimizes the opening — the view, the first few seconds. A 45-minute episode has a drop-off point at every second after that, and none of them are in the optimization.

Cost per completed hour isn’t a setting you can switch on in the account; the metric doesn’t exist there. It had to be built outside the platform — and Hourcast is the only engine that runs a campaign on it.

How the engine decides →

Two categories, same mechanics:

Podcasts. Full episodes — interviews, solo shows, panels, roundtables, video podcasts, audio-first uploads with a static visual.

Any long-form video on YouTube. Deep dives, documentaries, talks, keynotes, masterclasses, tutorials, recorded webinars, long interviews, video essays, teardowns.

What unites them isn’t the label. It’s that the value sits at the end of the video, not the start.

Seven minutes, and it’s a hard gate. The engine is built exclusively for episodes of 7+ minutes — shorts and clips are filtered out by design. Try to run one and the engine refuses to optimize it, flags the ad, and tells you why.

That’s by design, not a shortcoming. A 30-second clip gets finished by accident. A 45-minute episode gets finished by a listener.

Why the gate exists →

No. Shorts placement stays off, and anything under 7 minutes is refused.

Clips earn their keep in organic, or as a short pre-roll spot. Here they’re the wrong unit — the ad has to be the thing you want watched to the end.

Yes — and for most shows it’s where the easiest growth sits.

Look at what’s already on your channel. A company that published a strong episode a year ago still has a strong episode: the argument holds, the guest still matters, and almost nobody has heard it. Producing a replacement costs real money. Putting the existing one in front of new people costs a fraction of that.

Publishing the episode is roughly 20% of the work. The other 80% is getting it watched — and most budgets are split the other way round.

The engine doesn’t care how old an upload is. If people finish it, budget flows to it, and a proven episode from last year regularly beats this week’s. Same for creators sitting on a back catalogue, and for anyone building a name: the library is an asset, not an archive.

What paid distribution does →

Yes — in any language on the planet. You set country and language on the campaign, and the engine steers on spend and completions, which mean exactly the same thing in every language.

Ten on YouTube is the minimum. The engine finds winners by comparing episodes against each other — and by running variations of the same episode, different thumbnails and different copy, against each other too. One episode gives it nothing to compare.

Regular new episodes are ideal.

Reach into rooms organic can’t get you into.

Organic keeps recommending you to people who already look like your audience. Paid is the lane where you choose.

All eight reasons →

Predictable, once it’s running. Paid promotion is measurable and plannable: after the engine has settled in, output holds week to week rather than depending on what the feed surfaces.

Organic is the harder one to plan. It reaches people already close to your topic, and the timing isn’t yours to set.

No. Two lanes, one audience — ads run in their own slots, and your organic reach stays as it is.

The reason they don’t collide is who each lane reaches. Organic recommendations go to people already circling your topic: viewers YouTube has watched engage with that kind of content. Paid goes broad, to large audiences that have never met you. Different rooms, different people.

You can also cap how often one person sees you. Keeping frequency low stops the same viewer being shown the episode again and again — the Hourcast guides walk through where that setting lives.

And you don’t have to take it on trust: after a few weeks, switch Hourcast off for one week and watch what organic does.

Paid and organic side by side →

As in-feed ads — video recommendations sitting among the organic ones. In practice the large majority land in the home feed; the rest show up in search results and next to videos people are already watching.

That makes in-feed the one format where the viewer chooses you. Your thumbnail and title sit in the feed, and someone decides to open them. The attention you get was chosen, not taken from someone on their way to something else — and it’s the only ad format on the internet that suits 45 minutes of content.

Apart from a small Ad tag it looks like everything else in the feed, and it links straight to your watch page.

Your episode, recommended →

A pre-roll is forced viewing — someone wanted a different video and got yours. Fine for a 20-second spot, wrong for 45 minutes.

In-feed is the only placement where the viewer chooses your episode before it starts. In-stream and Shorts stay off at campaign level, and a campaign that drifts elsewhere gets detected and pulled out of optimization until you fix it.

About 10 minutes, once. You paste one script into your own account:

1. Copy the script — Hourcast → Settings → Script.
2. Paste it into Google Ads — Tools → Bulk actions → Scripts → +.
3. Authorize — tick Select all, then Continue. Google asks twice.
4. Run it once.
5. Schedule it for 6:00 AM.
6. Confirm — Last data received shows today’s date.

The onboarding walks you through each step when you get to it, and says so if something didn’t land.

Your YouTube Video campaigns only — spend, completions and device signals for every ad. Not Search, Display, Shopping, Performance Max or conversion campaigns.

A full demo account, from the minute the trial starts. Every page is loaded with sample data — the charts, the device breakdown, the campaign groups, the Actions — so you can click through the entire tool and see exactly what you’ll be looking at once it’s your show.

That’s what the trial is for: decide whether this is for you before a dollar goes into ads.

Your first real numbers replace it automatically.

The onboarding walks you through each step, and the dashboard keeps pointing at the one thing left to do. Every step has a how-to behind it — script, campaign, ads, audiences.

And it speaks up: notifications when something needs you, warnings when something isn’t set up the way the engine needs it — a flagged ad, a campaign delivering outside in-feed, data that stopped arriving.

A Google Ads account with billing set up, a channel with at least ten long-form videos, and an episode you’d put in front of strangers. Creating the Google Ads account takes about 15 minutes if you don’t have one.

Day 1: your first numbers — spend, completions, cost per completed hour. Around day 5: budget starts moving toward your winners. From week two: accuracy climbs, weak ads surface as Actions, the effect compounds.

At the minimum budget it’s usually performing by week three or four.

The first weeks, week by week →

Works out the number. What a completed hour costs you — per campaign, per ad, per device, per audience. That number doesn’t exist in Google Ads; Hourcast builds it.

Judges what’s working. New ads and new episodes get their own read before anything is judged on them.

Moves the budget. Toward the episodes earning completions cheapest, inside your campaign groups, never above your total.

Tells you what to change next. Pause this ad, it’s spending without completing. Lower the bid on mobile. This campaign is working — give it room. This one’s gone stale, add fresh content.

Behind all four sits a decision engine that works through a long list of situations before it moves anything. It makes better calls than we did running this optimization by hand.

The four moves in detail →

No. Your ads, your targeting, your copy, your thumbnails — nothing is edited without you. Hourcast works out what should change and puts it in front of you; the change itself happens in your account, when you decide.

Budget moving between the campaigns you grouped together is the one automatic step, and it never goes above your total.

Two things, and both are levers rather than chores.

Content and ads. You pick the episode, write the copy, supply the thumbnail, add and switch off ads. That’s the biggest lever you own: the content earns the completed hours, and a distinct thumbnail per ad variation moves cost per completed hour more than almost any setting in the account.

Device bids. Hourcast works out the number; putting it into the account is yours.

Measuring, testing, moving budget, deciding what to recommend — all of that runs on its own.

No. You paste one script into your own account and authorize it on Google’s standard screen. No admin access changes hands.

A script is the right mechanism for this because it’s open: you can read ours, and you can check exactly what it pulls. It only ever reads video campaign data — nothing else in your account is touched.

The Google Ads API has no supported function for writing device bid adjustments. Hourcast can calculate the number; there’s no endpoint to send it through.

You don’t have to type them in yourself, though. The recommended changes work as input for an AI agent — hand them over and let it make the edits in your account. Hourcast is building its own agent for this step over the coming months.

Budget keeps moving — that part runs on its own. The Actions are where the rest of the gain sits: the ad spending without completing, the device costing several times what the others do.

Left unapplied, those stay as they are.

One hour of your content watched to 100%. Two people finishing your 30-minute episode is one completed hour.

Only a full play counts. 95% isn’t a completed hour.

What you pay for one full hour of your content watched start to finish.

Your 30-minute episode: two people watch to the end, that’s one completed hour. Each completion costs $5. 2 × $5 = $10, so your CpCH is $10.

One metric runs it all →

Cost per 100% completed hour is the main one. That’s what the tool is for, and the number you work down over time.

Total completed time and completions come next. They tell you whether a campaign has enough to learn from — too few completions and the fix is usually fewer ads or more budget, not a different setting.

Views and watch time come last. Useful for reporting. Neither one says whether anyone heard your episode.

Quality. That’s the whole of it.

A view means someone pressed play, maybe for three seconds. Watch time averages everybody: five people watching four seconds and one person finishing the episode average out to a number that looks fine.

A completed hour can’t be flattered that way. It’s one person who heard the whole thing — the setup, the argument, the joke that only lands if you were there for all of it. That’s the deepest attention there is.

Views and watch time measure volume. Completed hours measure quality — and quality is what you’re paying for.

Most people who open your episode won’t finish it, and the ones who leave mostly leave early. That’s the normal shape of paid traffic — the same gap you see between a shop’s visitors and its orders.

Which is why the budget is steered on completions rather than opens. You’re paying for the people who stay, and the engine’s job is to make those cost less.

Google can count views, but it can’t deliver completed hours — its algorithms were built for 25-second spots, not 60-minute episodes where the value is in someone listening to the end.

It’s a capability the platform doesn’t offer, which is why the metric had to be built outside it.

There’s no universal benchmark — the metric is new, and it varies by show and audience. The one that matters is yours: track it and work it down. After a month, your own average is the baseline.

More than you’d expect. A few dollars per completed hour on a TV screen, several times that on a phone — people commit to a big screen and drift on mobile.

Hourcast breaks every number down by device and hands you the bid adjustments. It’s often the single biggest lever in the account.

A standard algorithm optimizes the first few seconds — the view, the decision to stay or skip — not the next 40 minutes. A 42-minute episode has 2,520 seconds where someone can drop off, and it sees almost none of them.

Without CpCH optimization a completed hour commonly costs 10 to 20 times more. At the extreme we’ve seen 50x.

$30 a month. One plan, everything in it — no tiers, no seats, no cut of your ad spend. The trial runs 14 days and doesn’t ask for a card.

$500 a month is the minimum that tells you anything. $750 is what we’d recommend — more room, faster learning.

That’s your own Google Ads spend on YouTube, on top of the $30 subscription. It’s a qualifier rather than a suggestion: below $500 there isn’t enough room for the engine to tell your winners from your losers.

Yours — in your own Google Ads account, billed to your own card. Hourcast directs where that budget goes; it never touches your billing and never holds your money.

No markup on your spend, no percentage of it. The $30 subscription is the only thing Hourcast charges for.

No. Your total is never raised — the engine redistributes inside the limit you set, never above it. Moves are capped, so nothing swings overnight.

New campaigns get protected learning time, and ones that stall get cut before they cost you.

The rules it can never break →

Yes. No long-term contracts. $30 a month, cancel whenever, reactivate anytime.

Two ways, and neither asks you to take our word for it.

Your own curve. From day one the dashboard shows what a completed hour costs you. Watch that number across the first weeks — the whole result of the tool sits in that one line, on your own account.

The A/B test. Two identical campaigns — same settings, same targeting, same episodes. One runs as the Optimized campaign, the other stays untouched as your Control campaign. Then compare what a completed hour cost in each, side by side.

Run the proof yourself.

Control vs. Optimized →

Two months at $750 a month of ad budget is the clean test. One month at $500 is the least that will tell you anything. Both figures are ad spend — the $30 subscription sits on top.

Even at the minimum it’s usually performing by week three or four — and it doesn’t level off there. Every week the engine knows more about your account: which placements, devices and ads deliver for your show. The picture keeps filling in, and the result keeps improving with it.

More budget shortens the early stretch. It doesn’t skip it.

Indirectly, and for some channels more than others. Shows in finance and investing have seen subscribers, comments and organic views rise alongside the paid campaigns.

It’s a secondary effect, and not one we can guarantee: it isn’t measurable from inside the tool, and the budget can’t be steered toward it. Google Ads does report earned actions — the organic views, subscribes and likes that followed a paid ad.

Partly, yes. There are strategies that optimize toward completions — audience and placement work, hand-tuned campaigns, careful reading of the reports. They work.

What they don’t reach is scale and predictability. Done by hand it doesn’t hold across dozens of ads, devices and audiences, and it doesn’t hold week after week.

And the steering is only part of it. The analysis, the budget moves between campaigns, and the Actions telling you what to change next all come with the tool.

The full version →

The script is the easy part. What’s hard is the decision logic behind it: which situations to act on, which to leave alone, and what a change does three weeks later.

What sits inside Hourcast came out of years of running campaigns optimized for content consumption, and hundreds of podcast campaigns across different accounts. All of it is built into the engine, which checks a long list of conditions before it moves anything — and makes better calls today than we would make by hand.

It also learns your account in particular: how your campaigns respond to a change, which moves held up and which didn’t. A script written over a weekend starts from zero on every one of those.

What the engine checks →

Start the trial and look around the demo account before spending a cent on ads. Everything on this page is visible in there — the Actions, the device breakdown, the campaign groups, the A/B test.

Or read how the engine works and what we shipped recently.