Paid social generates demand rather than capturing it. That makes creative the main lever, measurement the main risk, and discipline with budget the main reason campaigns survive past month two.
Search advertising answers a question someone is already asking. Paid social does the opposite: it puts a message in front of people who were not shopping, and it works when the creative earns attention and the offer is worth stopping for. You are not bidding on intent that exists. You are manufacturing it.
That changes what the work looks like. Audiences and budgets matter, but the creative does most of the persuading, and the tracking decides whether the platform can learn anything useful from the results.
The pixel and the Conversions API define what the platform is allowed to optimise towards. If the events fire on the wrong action, or fire twice, or only fire for half the traffic, the algorithm learns the wrong lesson quickly and confidently. We audit the event setup before we touch a budget, because no amount of creative testing repairs a broken conversion signal.
Retargeting depends on the same foundation. People who visited a page, started a form or watched most of a video are the warmest audience you have, and they need their own message and their own exclusions. Showing the same introductory ad to someone who already enquired is a waste on both sides.
In paid social, the ad is the targeting. Different angles, formats and opening lines reach different people, and the platform finds them for you once one variant clearly outperforms. That means testing in batches, with a hypothesis for each, and giving a variant enough spend to produce a readable result before replacing it.
We judge creative on response, not on taste. A rough video that sells usually beats a polished one that does not, and the only reliable way to know which is which is to run both.
If your audience is very small and highly specific, the platform cannot find enough of them to make the spend efficient, and search or direct outreach will do better. If your margin cannot absorb a testing phase, or if there is nobody to follow up on the leads it produces, paid social will lose money competently. If the product has no visual or impulse dimension, it usually struggles here too.
That assessment belongs on the first call, not on your first invoice. If paid search or a funnel fix is the better first move, that is what we will recommend.
Paid social works when there is a visual product or a clear problem to name, when the margin supports testing, and when leads get followed up within the hour.
It is a poor fit for tiny niche audiences, for high-ticket sales with no follow-up capacity, and for anyone who needs certainty before spending anything.
Boosting puts budget behind content built for followers, not for a cold audience. It has no event to optimise towards and no structure to learn from, so it produces reach rather than a measurable cost per acquisition.
A low cost per click means nothing if the people clicking never enquire. When the campaign optimises towards the top of the funnel, it finds the cheapest attention available, which is rarely the attention that buys.
Every significant edit resets what the platform has learned. Editing budgets, audiences and creative in the same week means nothing gets a fair test, and the account stays permanently in a learning state.
Retargeting the people who already bought, or who are already in a sales conversation, spends money on decisions that are already made. Exclusions are as important as the audience itself.
Pixel and server-side events are verified against real conversions before any budget is committed.
We map who is worth reaching, what they currently believe, and which angle makes the offer relevant to them.
Several angles run at once with a clear hypothesis each, at a spend level that produces a readable result.
Winning creative gets more budget, the rest is retired, and retargeting picks up everyone who showed interest.
Enough to gather a meaningful number of conversions, which depends entirely on your cost per acquisition and how quickly you need a decision. We work that number out from your margin and your sales cycle on the first call rather than quoting a figure that suits us.
Less reliable than it was, and pretending otherwise would be dishonest. Server-side events, first-party measurement and a blended view of acquisition cost all help. We treat platform numbers as directional and reconcile them against your CRM.
It usually performs better than static, but it does not need to be expensive. Something shot on a phone that shows the product working can beat a produced brand film. We test both and let the response decide.
There is no honest single answer, because it depends on your margin, your offer and how fast the learning phase passes. What we commit to is a clear read on whether the channel is working, with the evidence behind it, before you keep spending.
Book a call and we will look at your margin, your audience and your follow-up together, and work out whether this channel is worth your budget.