Someone has told you to run ads, and the advice stopped there. Not what to run, not how much to spend, not how you would know whether it worked. Paid advertising is the part of marketing most surrounded by confident people who skip the part where they explain themselves, and it is easy to spend a month's budget learning something you could have read in an afternoon.
This guide is that afternoon. It covers what paid actually buys you, when it is worth doing and when it is not, the order to think about objectives in, how paid and organic feed each other, and how to reason about a minimum budget from arithmetic rather than from a number someone invented. It stays deliberately platform-agnostic until the end, where the ad managers get a short tour.
What paid actually buys
Strip away the language and paid advertising buys you two things that organic cannot give you on the timescale you usually need them.
The first is reach you cannot earn fast enough. Organic distribution compounds, but slowly, and it is rationed by the platform. If you have something that needs to be in front of a lot of the right people this month (a launch, an offer with an end date, a product nobody knows exists yet), organic will not build that audience in time. Paid does, because you are buying the distribution outright instead of waiting for it to be granted.
The second is testing speed. Organic tells you how a piece did with the audience the platform chose to show it to, which is a moving target you do not control. Paid holds distribution roughly constant, so the difference between two creatives is much closer to being about the creatives. That makes paid the fastest, cleanest way to learn what actually resonates, often faster than the thing you are ostensibly buying, which is sales. Which posts deserve budget goes deep on using paid as a testing instrument rather than an amplifier.
Notice what is not on that list: paid does not make bad content work. It makes content reach more people, which for weak content just means more people ignore it, faster. Money is an amplifier, and amplifying a flat signal gives you a louder flat signal.
When it is worth it, and when it is not
Paid is worth it when you have a specific thing to move and a way to tell whether you moved it.
It is worth it when you have a proven creative and distribution is the only thing missing: a post that earned attention from the people who saw it but never reached enough of them. It is worth it when you have an offer with a deadline and organic cannot build the audience before the deadline arrives. It is worth it when you want to learn quickly which message, audience or format works, and you are willing to treat the spend as tuition.
It is not worth it in a few common situations, and recognising them saves real money. It is not worth it when you have nothing proven to put behind it: paying to discover that untested creative does not land is the most expensive way to run a test you could have run organically first. It is not worth it when there is nowhere for the attention to go: budget on awareness with no destination and no follow-up is fine only if you have decided, in advance, that awareness is the whole goal. And it is not worth it when you cannot measure the result you care about, because a campaign you cannot judge is a campaign you will keep running out of hope.
The honest version of the decision is: do not start with "should we run ads". Start with "what one thing are we trying to move, and how will we know". If you cannot answer the second half, the answer to the first is not yet.
The objective ladder
Every ad platform asks you to pick an objective before it will let you build anything, and the choice matters more than any other setting, because the system optimises towards whatever you chose. Pick the wrong objective and the platform will faithfully deliver the wrong thing.
The objectives form a ladder, roughly from cheap and far from a sale to expensive and close to one:
- Awareness: getting known by as many relevant people as possible. Cheapest per person reached, furthest from revenue. Judge it on reach and recall, never on sales.
- Traffic and engagement: getting people to click through, watch, or interact. A middle rung: closer to intent than awareness, still not a purchase.
- Leads: capturing a contact detail, a sign-up, a form completion. Now you have a person you can follow up with.
- Sales: the actual transaction, or the equivalent action that is the point of the whole exercise.
The mistake almost everyone makes is judging a rung by the metric of a higher one. Awareness content judged on sales looks like a failure and gets killed, when awareness was never its job. Sales campaigns judged on reach look weak and get scaled back, when they were the most profitable thing running. Match the metric to the rung you chose, and decide which rung you are buying before you look at the numbers. Meta describes how objectives map to optimisation in its own Ads Help Center, and the specifics change, so confirm the current objective set there rather than trusting any list (including this one) as permanent.
How organic and paid feed each other
Treating organic and paid as separate budgets run by separate people is the most common structural waste in the whole discipline. They are one system, and each makes the other cheaper.
Organic is your test bed. It shows you, for free, which ideas earn attention from a warm audience. The pieces that over-performed organically are your best-qualified candidates for paid, because one expensive question (does the creative hold attention) has already been answered without spending anything. Structuring a Meta campaign from a proven organic winner is the whole method for that hand-off, including the important caveat that an organic win proves the creative and nothing else.
Paid, in turn, is your discovery engine for organic. The genuinely interesting result is when the two disagree: a post that flopped organically and flies when boosted was good and simply never got shown, which is a discovery you cannot make any other way and which should change how much you trust organic reach as a quality signal.
So the loop is: make organically, watch what breaks pattern, fund the winners, read what paid teaches you back into what you make next. Neither half works as well alone.
A sensible minimum budget, from arithmetic
The question everyone wants answered, "how much do I need to start", has no fixed answer, but it does have a floor you can calculate rather than guess.
The floor exists because ad delivery systems need a certain volume of results before they stabilise. On Meta, a new ad set spends its early days in a learning phase while the system explores, and results before it settles are noisy rather than meaningful. Meta documents the learning phase and the volume it looks for in its own Ads Help Center; the figure it has published is around 50 optimisation events within seven days per ad set, and because that specific number can change, you should confirm it there rather than treating it as fixed.
That gives you a division you can do before spending anything:
daily budget × 7 ÷ estimated cost per result = results per week per ad setIf that comes out well under the volume the platform needs, your budget is spread too thin for the structure to ever produce a readable answer. The fix is fewer ad sets with more budget each, not more patience.
The practical takeaway: your minimum is whatever makes at least one ad set clear the platform's volume bar, with your own estimated cost per result as the input. Start there, with one thing to test, and expand only once you can read the result.
The ad managers at a glance
You run paid social through a platform's ad manager, and they are not equal in depth.
Meta Ads Manager is the deepest and most widely used, covering Facebook and Instagram from one place, with the richest targeting, testing and reporting of the group, which is also why it has the steepest learning curve. It is the platform this library covers in most detail: the deep Meta Ads Manager series walks through campaign structure, building audiences, the pixel and events, budgets and bidding, creative and placements, and reading your results, and set up your first Meta ads campaign ties them together into one end-to-end build. Meta's own Meta for Business hub is the authoritative reference for all of it.
The others exist and matter for the right audience. TikTok Ads Manager runs paid on TikTok and is the place to reach a younger, sound-on, short-video audience. LinkedIn Campaign Manager runs paid on LinkedIn and is where professional and B2B targeting lives, at a higher cost per result that its audience often justifies. Google Ads, including YouTube, covers search intent and the largest video platform. Each has its own documentation and its own quirks; the reasoning in this guide (objective first, test small, judge on the right rung, calculate your floor) transfers to all of them even though the buttons differ.
Where Acumin fits
Acumin does not run your ads, and it is honest about the line. What it does is the work either side of the campaign.
Studio is where you make the creative (the video or image you will actually put money behind) so that the strongest candidate from your organic testing becomes a finished ad. The Content Strategy cockpit holds the plan the paid budget serves, so a campaign is funding something you decided on purpose rather than a boost you reached for in a panic. And Snapshot reads your public organic numbers, which is how you find the over-performing pieces worth funding in the first place.
One honest limit, stated the same way it is stated in the product: reading Meta ad results back automatically is pending Meta's app review. Until that lands, Acumin drafts and prepares while you read your results in Meta's own reporting and send from there. Acumin drafts; you send.
Do this today
Write one sentence: the single thing you want a first campaign to move, and the number that would tell you it worked. Then do the division above with a daily budget you can sustain for a month and your best estimate of cost per result. If one ad set clears the platform's volume bar, you have a starting budget. If it does not, you either need a larger daily figure or a cheaper objective. And knowing that now is worth more than the campaign you were about to launch without it.