Paid ads, run in volume
The fastest way to find ads that convert is to run many variations at once and let the data decide which survive. One ad is an opinion; twenty live variations tell you in days what one ad takes months to say. This guide covers how to produce that volume, where to run it, and the daily discipline that turns it into winners.
Why do most ad accounts stall?
Because they test too little, too slowly. A handful of ads on a six-week production calendar cannot generate enough data to learn from.
The usual pattern: weeks of production for two or three "perfect" ads, a launch, then a quiet month of watching numbers that never become conclusive. The ads are usually fine. The account is failing because nobody is asking the market enough questions at once to get an answer.
Markets get discovered in volume. Each variation is a question, which hook, which format, which audience, and the platforms answer quickly when you ask enough of them at the same time. The volume is the strategy.
One caveat before spending anything: ads pointed at a broken funnel just measure the leak faster. If positioning or pricing is the real blocker, fix the go-to-market plan first.
How many ad variations should you test?
Enough that every variation answers a distinct question. Corvan's working number is twenty per sprint, with hooks, formats, and audiences varied deliberately.
The number matters less than the structure. Twenty near-identical ads teach you nothing; twenty deliberate variations map the territory. Vary one axis at a time so a winner tells you why it won:
- Hooks: the first line or first second, the biggest lever you have
- Formats: static, carousel, story, short video
- Audiences: which segment of your ICP each variation targets
Speed comes from deciding before producing. Lock the decisions into a one-page brief, build variation frameworks you reuse every sprint, and production stops being the bottleneck. That system is how Corvan's ad operations get twenty variations written, designed, and live within one hour of an approved brief.
Which ad platforms should you start with?
Meta and Google first, because that is where the volume is. Then TikTok, LinkedIn, or wherever your buyers spend their attention.
Platform choice is an ICP question. Meta and Google earn the first slots in most plans because their reach and data volume make them the fastest places to learn. After that, follow your buyers: TikTok, LinkedIn, or a niche platform your category lives on.
Whatever the mix, run it as one operation: one team trafficking all of it, one dashboard reporting all of it. Fragmented platform silos are how budget quietly disappears.
- Meta and Google first: reach and data volume
- Then the platforms your ICP uses
- One dashboard across every platform, from day one
How often should you optimize campaigns?
Daily. Cut losers as they show themselves, scale winners the same day, and let budget flow continuously toward what converts.
Volume only pays if you act on what it tells you. The rhythm is daily cuts and doubles: look at the numbers every day, cut what is losing, put the freed budget behind what is winning. Waiting for the monthly review means funding losers for weeks longer than you had to.
"Set and forget" is the expensive failure mode. Platform automation is a tool. Nothing should run on autopilot, because the platforms optimize for their own goals. Decide your kill criteria before launch, in writing, the same way you would for any channel, so cutting an ad is a rule being applied and nobody has to argue about it.
What should your ad budget be?
Enough to feed your variations with real data. The exact floor depends on your market and platforms, so set it deliberately, not by feel.
There is no universal minimum, but there is a working test: can every variation you run collect enough data to be judged? Spread too little spend across too many ads and nothing reaches significance; you learn nothing and call the channel dead. Better to fund fewer questions properly than many questions badly.
Three budget rules that hold up:
- Size the budget to the test, not the other way around
- Keep spend flowing directly to the platforms, never marked up
- If the money is better spent on strategy first, spend it there
Corvan sets the floor with each client during a free go-to-market consultation, because it differs by market. Be wary of anyone who quotes a universal number before hearing about your market.
What should ad reporting look like?
A live dashboard you can open any time, plus one written report a week in plain language: what was cut, what was scaled, what it cost, what it returned.
Numbers on their own leave the reader guessing. A dashboard shows spend and results per ad; the weekly report explains the decisions behind the numbers, written by the team that made them. Together they give you sentences you can repeat to your board without translating.
- Dashboard: spend and results per ad, live from day one
- Weekly written report: cut, scaled, cost, returned
- The decisions behind the numbers, explained
If someone else runs your ads, this rhythm is the accountability layer: you should never have to book a meeting to find out what is happening. See how we run ads for the full weekly cadence.
The results speak for themselves
+250 clients in the first month
What we did. Meta Reels and Stories with supercar creatives, from October. Every ad carried a 15% code, so every booking traced back to the ad that won it.
What it returned. Over 250 clients in month one. Months later they are still running the ads we built: the screenshots here are recent.
Twenty ads from one shoot session, live within the hour of the approved brief.
Read the case study →“Corvan helped us showcase our new car to our customers. We wanted something that highlighted the luxury of the car, but that also worked for our Meta ads. They delivered quickly and professionally. It worked so well that we are still running these ads today!”
Want this run for you?
Corvan's ad operations are this guide at production speed: twenty variations live within one hour of an approved brief, cut and scaled daily, reported plainly every week.
Questions people ask about paid ads
How fast can new ads go live?
Hours, once the decisions are locked in a brief. Corvan traffics twenty variations within one hour of an approved brief. The speed comes from the system: variation frameworks reused every sprint, and a brief format that settles the decisions before production starts.
Who should own the ad accounts?
You, always. Whoever runs your ads should work inside your accounts, with your spend going directly to the platforms. If you part ways, every campaign and every learning stays with you.
How fast do paid ads show results?
Ads produce signal in the first week and reliable winners within a month, provided you run enough variations to generate real data and cut the losers as they show themselves.
Is there a minimum ad budget?
The floor is whatever feeds your variations with real data, and it depends on your market and platforms. Set it deliberately before launch. If the budget is better spent on strategy first, whoever runs your ads should say so.
One level deeper, one question at a time
Short posts that go past this guide, with sources. A new one most days.
Is TikTok worth it for B2B paid ads?
When TikTok is worth a funded test for a B2B company, which products suit a short video feed, how to budget it next to Meta, and how to judge it on pipeline.
Read more →
Does video outperform static images in Meta ads?
Motion's data on 578,750 Meta creatives ranks text-only ads first for hit rate. When video earns its cost, and what Facebook's full-screen test changes.
Read more →
What attribution window should you use for Meta ads?
Meta's default attribution window since March 2026, when a 1-day click window fits better, and why changing the setting moves your reports but not your results.
Read more →Brief us this week. Be live within the hour.
Bring your market to a 30-minute scoping call. We set the budget floor together and you approve every brief before it runs.