Performance marketing borrowed its vocabulary from ecommerce, where a click becomes a purchase in the same session and the arithmetic closes within a day. Business-to-business selling does not work that way. A qualified enquiry may become revenue nine months later, after a procurement review and three stakeholders who never filled in a form, which means a performance agency is really being asked to optimise against a proxy. Choosing one well is mostly a question of whether you agree on the proxy, and whether you both trust the way it is measured.
What performance means when the cycle is long
In practice a B2B performance agency works to one of three targets. Cost per lead is the easiest to measure and the easiest to game, since form-fill volume can be manufactured with a low-value offer nobody in sales wants to call. Cost per qualified opportunity is better, because it requires the sales team to accept the enquiry and forces both sides to agree a definition, though it introduces a lag of weeks and a dependency on your own team's discipline. Pipeline or revenue influenced is the most honest and the most contested, since attribution across a long cycle involves judgement calls that reasonable people dispute. Decide which of the three the contract is written against before you compare agencies, because they will optimise towards whatever you chose to count.
Measurement is the product
Ask each finalist to draw the chain from an ad impression to a closed deal inside your systems, naming the platform, the analytics property, the form, the CRM object and the field that stores the source. If they cannot draw it, they are not a performance agency whatever the website says. Then ask what they do about the parts that cannot be tracked: the buyer who reads three pages on a phone, mentions your name in a meeting and arrives eight weeks later through a branded search. A serious answer acknowledges the gap and proposes something practical, a self-reported source field on the form, a lift test, a hold-out region. An answer claiming complete attribution is either naive or selling.
Channel mix and where the money really goes
Most B2B performance programmes concentrate on paid search, advertising on professional networks, and retargeting, and the economics differ sharply between them. Paid search captures existing intent and is priced accordingly, with legal, insurance and software among the most expensive categories in the country. Professional network advertising reaches people who are not searching yet, costs more per click, and often produces better-fitting accounts. Google explains that its auction ranks ads on the bid together with auction-time quality including expected click-through rate, ad relevance and landing page experience, summarises its view of ad quality in a Quality Score advertisers can monitor in the account, and states that higher quality ads can often lead to lower costs per click. The landing page is therefore a performance lever, not a design afterthought, so ask who writes and builds those pages and whether that sits inside the fee.
Fee models, and why per-lead pricing misfires here
Pay per lead is common in consumer categories such as home services, where a job's value is known and the decision happens on one phone call, and it translates badly to B2B. When a single account is worth many times the average and the cycle runs for months, a fixed price per enquiry pushes the agency towards the cheapest leads it can find rather than the accounts you actually want. Flat retainers against a defined scope avoid that distortion and are easier to audit, while hybrid models with a modest performance component can work if the bonus is tied to qualified opportunities rather than raw form fills. Whatever the model, keep ad accounts, analytics and conversion tags in entities you own, because the historical conversion data feeds automated bidding and is expensive to lose.
Questions people ask about b2b performance marketing agency
What should a B2B performance agency be held to?
Whatever you can both measure honestly and your sales team will accept. Cost per qualified opportunity is usually the best compromise between measurability and meaning. Agree the definition of qualified in writing before the contract starts, not after the first disappointing month.
Is attribution in B2B ever reliable?
Not fully. Multi-stakeholder buying, long cycles and untracked research make complete attribution impossible. The practical approach combines platform data, a self-reported source field on your forms and periodic hold-out tests, and treats any agency claiming perfect attribution as a warning sign.
How long before a B2B performance programme can be judged?
Leading indicators such as qualified enquiries within a quarter, and revenue only after a full sales cycle has completed. Judging on closed revenue before that has elapsed will make you cut a programme that was working, or keep one that was not.
Should the agency own the ad accounts?
No. Accounts, tags and audiences belong in entities your company owns, with the agency granted access. Conversion history trains the platforms' automated bidding, so handing it over means any future agency starts effectively blind.