Expensive clicks, deliberately spent.

Because you are buying precision, and the auction prices it. Dreamdata measured a LinkedIn CPC of EUR 5.98 against Meta's EUR 1.60, which looks damning until you compare what each click represents. Nowhere else can you reliably reach a director of operations at a 500-person manufacturer. The right comparison is cost per qualified opportunity, or cost per company influenced — measured that way, the same benchmark data puts LinkedIn return on ad spend at 121%. If your reporting stops at cost per click, the channel will always look like a mistake.

Enough to reach your audience repeatedly rather than once, which for most B2B companies starts in the low thousands per month. The constraint is audience size rather than ambition: a tightly defined list of 5,000 companies does not need a large budget, it needs consistency and fresh creative. Spreading a small budget across several broad audiences is the reliable way to waste it. If your addressable audience is genuinely small, we will often recommend starting with one campaign, one offer and a disciplined test, and only widening once the economics are proven.

Both, chosen per offer rather than by preference. Native forms remove friction and pre-fill from the profile, so they win on volume — which is right for a guide or benchmark aimed at a cold audience. Landing pages ask more and therefore qualify harder, which is right for demos and pricing conversations where you would rather have five serious enquiries than fifty curious ones. With a 6.6% median conversion rate across landing pages, a page only earns the extra step if it is genuinely good, so we test the two against pipeline rather than lead count.

Three changes usually do it. Tighten the targeting using company lists and firmographic filters and switch expansion settings off, because that setting is the most common source of stray leads. Make the offer self-selecting: a benchmark for practitioners attracts practitioners, while a free voucher attracts anyone. Then feed the CRM back into the platform, so its optimisation learns from qualified opportunities rather than from form fills. That last step is the one most advertisers skip, and it is the difference between a platform optimising toward your buyers and toward whoever clicks most.

It is the platform's natural shape, provided you accept that you are buying influence across a committee rather than leads from individuals. Dreamdata's benchmarks describe 88 touchpoints across four channels with ten stakeholders in the average journey, so targeting one decision-maker and waiting for a form fill misreads how these deals happen. We run matched company lists, address the roles that influence as well as the one that signs, and report influenced companies so sales and marketing are arguing about the same list of companies rather than different numbers.

Leading indicators appear in weeks; the pipeline read takes a quarter. Within three to four weeks you can see whether the targeting is right, whether creative is holding attention and whether lead quality has improved in the CRM. Whether it produced revenue depends on your sales cycle, and pretending otherwise leads to campaigns being killed a month before they would have worked. We agree up front which indicators we act on monthly — frequency, engagement, qualified lead rate, opportunity creation — and which decisions wait for the deals to mature.

It varies by segment more than most people expect, which is why we test rather than recommend a format up front. Document ads consistently punch above their weight for practical, information-led offers because people can consume them without leaving the feed. Thought leader ads work when the person genuinely has standing and something to say, and fall flat when they are a logo with a face attached. Video earns attention for product explanation, single image remains the efficient workhorse, and conversation formats work briefly then wear out.

Long cycles are the case for the channel, not against it, but they change how you measure and what you ask for. If a deal takes nine months, an ad campaign judged on this month's demo bookings will always look like a failure. We shift the ask earlier — useful content, subscriptions, engagement — then retarget from that engagement, and we push CRM stages back into the platform so it optimises toward the opportunities that eventually mature. Reporting shows influenced accounts and pipeline created, which is the honest picture for a long cycle.

Ask what they will report on. An agency that leads with cost per lead is optimising for something your sales team does not care about; one that talks about qualified opportunities, influenced companies and CRM feedback has run the channel before. Ask who does the work day to day, how often creative is refreshed, and whether you keep the ad account and audiences afterwards. Ask to see the first ninety days written down. Any agency comfortable putting that in writing before you sign is worth a second conversation.

We can, and B2B advertisers usually benefit because the channels are measured against one definition of a qualified opportunity instead of three. Our paid search, paid social and analytics teams share the same tracking, so LinkedIn stops being judged in isolation against cheaper clicks elsewhere. It is equally workable to buy LinkedIn alone while another agency runs your search — we will share the measurement setup so the numbers still reconcile.