Most B2B teams choose their first paid channel by asking where their buyers are. The better question is whether those buyers type the problem into a search box often enough to fund a quarter.
Paid channel selection for B2B comes down to one decision: are you buying demand that already exists, or paying to create it? Search buys existing demand and only works when enough people are searching the terms that describe your product. Audience platforms buy attention from people defined by who they are, which works when the buyer is identifiable but silent. You can answer which one you need in an afternoon, before any money moves.
This is written for a marketing leader at a mid-market or enterprise B2B company who has budget approved, one channel's worth of money, and pressure to pick.
The short answer: run four tests before you fund a channel
Size the search demand for your category and turn it into a lead count. Read the live result page for your main phrase and see whose product is actually there. Check whether the person doing the searching can sign anything. Then, if your buyer set is a countable list rather than a market, price what it costs to stay visible to that list.
Three of those four tests are free. All of them are cheaper than finding out with a live campaign eight weeks later.
Ready for paid ads that pay off?
Book your free auditTest 1: Turn search volume into a lead count, not a vibe
A category head term with a few thousand monthly searches sounds like a market. Do the arithmetic and it usually is not.
Take an illustrative example. Say your main phrase and its close family show 2,000 average monthly searches in the United States. You win 60% impression share, your ads pull a 6% clickthrough rate, and your page converts 3% of clicks into leads. That is roughly 72 clicks and two leads a month. Half of those will be qualified. You are now running a channel that produces one good conversation every month.
That number matters beyond pipeline. Google's own documentation says it can take "up to around 50 conversion events or 3 conversion cycles" for an automated bid strategy to calibrate toward a new objective (Google Ads Help). At two conversions a month, the system never finishes learning and neither do you.
Read the volume number carefully. Google states that average monthly searches are averaged over a 12-month period, cover a keyword and its close variants, are only reported for exact matches, and that search volume statistics are rounded (About Keyword Planner forecasts). The same page defines top of page bid low and high ranges as approximations of the 20th and 80th percentiles of what advertisers have historically paid. So you are planning against a rounded average and a bid range, which is fine as long as you treat it as a range and run the lead arithmetic at both ends.
Test 2: Read the live result page before you fund the keyword
Search volume tells you how many people typed something. It says nothing about what they meant. Open an incognito window, search your main phrase in your target country, and look at what the page is selling.
The failure pattern we see most often is a category phrase that belongs to a different kind of product. A company sells a service, and the phrase that describes the problem returns hardware, tools, or consumer versions of the same words. Every click is real, the intent behind it is not yours, and your cost per qualified lead ends up three times what the plan assumed because you are paying a price set by advertisers selling something else entirely.
Two things to record while you are there:
→ Who is buying ads on this phrase. If several direct competitors are present, the demand is commercial and priced. Keyword Planner's Competition column describes this as the number of advertisers showing on a keyword relative to all keywords across Google, rated low, medium or high.
→ What the top organic results are. If they are all definitional articles and no purchase pages, people searching the term are learning, not buying, and a demo request is the wrong thing to ask them for.
An empty result page is information too. Nobody advertising on a phrase with real volume usually means one of two things: the query does not convert, or nobody has tried. Assume the first until you have evidence for the second.
Test 3: Ask whether the person searching can sign anything
Search finds the person with the problem. In committee purchases, that person is often a researcher gathering options for someone else, and the person who signs never searches for your category at all.
This is the part teams skip, and it shows up later as a pile of leads that all need internal permission. If you map your last ten closed deals and the initiator was a mid-level specialist while the decision sat two levels up, search will reach the specialist and stop. That is a fine start if you have content that helps the specialist sell internally. It is a bad start if your only offer is a sales call.
When the signing buyer is definable by role and company but does not search, paid social is the correct first channel. You are buying access to a person rather than a query, which is the one thing a professional network platform genuinely sells.
Test 4: If your buyer set is countable, price the coverage
Some B2B companies do not have a market. They have a list. When the total addressable set is a few hundred or a few thousand named accounts, the channel question changes shape. It is no longer "which platform has our audience," it is "what does it cost to stay in front of these specific accounts for a quarter."
Two practical constraints govern that answer.
→ Platforms have floors. LinkedIn's company list targeting documentation states an uploaded list needs at least 300 rows, and that the minimum ad set audience size is 300 member accounts (LinkedIn Marketing Solutions Help). A 150-account list is not a campaign yet, it is an outbound project with ads layered on later.
→ Platform audience estimates are soft. Company size and revenue fields are often self-reported and the counts you see are ranges, which is why we treat firmographic targeting data as directional rather than exact when sizing a program.
Coverage math is simple and nobody does it. Take your matched audience size, decide how many times a month you want each person to see you, and multiply by the platform's cost per thousand impressions. That gives you a monthly number you can defend or reject. Most teams discover the list is cheap to cover and expensive to convert, which is a much better problem than the reverse.
The decision rule
Demand exists and the result page shows your category. Start with search. One campaign on non-brand commercial terms, tight ad groups, a page built for that query, and brand defense added only if competitors actually appear on your name. Our B2B Google Ads work almost always starts here when the arithmetic in Test 1 clears.
Demand is thin, or the result page belongs to a different product. Start on an audience platform and accept that the first flight buys evidence rather than pipeline. Fund the creative, not the keyword list.
The buyer is neither searching nor definable. You do not have a channel problem. Go back to the readiness checks and fix the offer, the page, or the conversion definition first. Paid media amplifies whatever is already true.
Fund one channel to a verdict, not three to an opinion
The most common mistake after the channel decision is hedging it. A team picks search and social and a review site, splits the budget three ways, and six weeks later has three campaigns that all look mildly disappointing and none that produced enough data to defend or kill.
Size the first flight by what a verdict costs. Six to eight weeks, one channel, enough spend to clear the learning math above, and a written definition of what qualified means before launch. If you cannot fund one channel to that standard, you cannot fund two, and the sequencing question belongs in annual planning rather than in this month's test.
Once both channels are live and producing, the question becomes reallocation, which is a genuinely different exercise with its own rules.
Two numbers to write down before you open an account
First, the lead count your search volume actually supports at a realistic impression share, clickthrough rate and page conversion rate, calculated at both ends of the bid range. Second, the monthly cost of covering your target account list at a frequency you chose on purpose.
Whichever number survives contact with your budget is your first channel. If neither does, the honest answer is that paid is not the next move, and we will tell you that. That is also what happens on a free paid media audit with us, before anyone signs anything.
