Arcane Marketing logoArcaneApply
← All posts

Two firms pitch the same prospect. One has been in that buyer's field of view for a year: quoted in a newsletter they read, mentioned by a peer at a conference, already a name in the room. The other arrives through a well-targeted search ad in the week the buyer starts looking. Forrester's Buyers' Journey Survey, 2025 found that 68% of B2B buyers already have a front-runner vendor in mind at the very start of their purchasing process, and 80% of the time that front-runner wins. The second firm is paying for the right to be compared. That gap is the real brief for a B2B growth marketing agency working with a professional services firm, and it is almost never the brief written on the scope of work.

Service businesses have a specific version of the problem. What you sell is people and judgement. The buyer is a committee. The contract runs a year or longer. Nobody can sample the thing before they buy it. Each of those facts breaks the default playbook of a lead volume target, a cost per lead ceiling, and a monthly dashboard that shows both.

Preference is set before anyone fills in your form

The same Forrester work reports that only 19% of B2B marketing leaders believe buyers have clear vendor preferences at the beginning of the purchase process. So roughly two thirds of buyers show up with a favourite, and about one in five marketers is planning for that. The programmes get designed for an open-minded buyer who, most of the time, is not the buyer who turns up.

The practical consequence is a budgeting one. Spend that only intercepts late intent competes for shortlist slots you did not already own, against incumbents who have been accumulating preference for months. It still converts, because some buyers genuinely start cold and some incumbents are weak. It just cannot be the whole programme, and when it is, the symptom is exactly what you would expect: respectable cost per lead, flat win rate, long stalls after the first call.

The work that builds front-runner status for a service firm is unglamorous and slow to show up in a dashboard:

  • Named-practitioner content with a point of view, published somewhere the buyer already reads
  • Podcasts, panels and partner audiences where your buyer is already paying attention
  • A referral and alumni motion that is actually operated, with owners and a cadence, instead of being left to goodwill
  • Retargeting and email that keep you present through a quiet six months rather than pushing for a meeting every fortnight

The decision has about thirteen people in it

There is a second structural reason lead-level thinking underperforms here. Forrester counts an average of 13 people inside the buyer's organisation and nine from outside involved in making a purchase decision, with 73% of purchases involving three or more departments.

Read that as a content requirement. The person who books the call is rarely the person who approves the spend, and almost never the only one with a veto. Your champion has to carry your argument into rooms you will never sit in, against a procurement question, a legal question, and an internal alternative where someone's team does it themselves. If the only asset you have given them is a proposal PDF and a good call, they are carrying your case from memory.

So build for internal selling. A one-page business case the champion can forward without editing. A short piece aimed squarely at the finance objection. Something that addresses the build-versus-buy argument head on, which for service firms is usually a hiring question more than a vendor question. If that comparison is live in your category, the in-house versus agency cost breakdown is the shape of argument your champion needs in hand.

What a B2B growth marketing agency cannot see in the ad platform

Here is where most reporting conversations go wrong, and the platforms are reasonably candid about it. Google says data-driven attribution is the default attribution model for most conversion actions, and that the model works by comparing the paths of customers who convert to the paths of customers who don't to identify patterns among those ad interactions that lead to conversions. To do that properly it needs volume. Google recommends having at least 200 conversions and 2,000 ad interactions in supported networks within a 30-day period.

Now apply that to a service firm optimising toward qualified calls rather than newsletter signups. Check your own thirty-day count on that money event against the volume Google recommends before you trust what the model is telling you. The model still runs. You are simply reading a model that is working with thin data on the event you care about, which is a different thing from reading a measurement.

LinkedIn is equally explicit about mechanics. For a conversion action associated with an ad set, the most recent ad set that has had an ad interaction in the conversion window will get credited with a conversion, and for a conversion action associated with multiple ad sets, every ad set that has had an interaction in the conversion window will get credited. Those are credit-assignment rules. None of this establishes what your advertising caused, and no window length or model choice turns it into that. Treat in-platform numbers as optimisation signal and stop asking them to settle revenue arguments. If your current reports are mostly platform exports, reading agency reporting like an operator is the more useful next step than another attribution model.

Send the outcome back rather than reading the dashboard harder

The constructive move is to feed downstream reality into the systems doing the bidding. Google's enhanced conversions for leads is an upgraded version of offline conversion import that uses user-provided data, such as email addresses, to supplement imported offline conversion data to improve accuracy and bidding performance; the hashed customer data is compared to hashed customer data of signed-in Google Accounts and attributed to ad events to help measure conversions driven by your campaign.

That is the actual machinery behind the phrase everyone uses about optimising to pipeline instead of form fills. It requires your CRM to record stage changes reliably, which in service firms is usually the binding constraint rather than the tagging. If your sales team updates opportunity stages when they remember to, no import will help, and that is a sales operations fix before it is a marketing one.

Model it on your own numbers before you buy anything

Assumptions first, and all of them are yours to replace. Assume $20,000 a month of media, $400 to book a qualified call, a 60% show rate, a 25% conversion from held call to proposal, a 30% proposal close rate, and $60,000 of first-year contract value per client.

$20,000 at $400 is 50 booked calls. At 60% showing up, 30 held calls. At 25%, 7.5 proposals, so call it seven. At a 30% close rate, roughly two clients, with the third rounding in and out month to month. Two at $60,000 is $120,000 of first-year revenue against $20,000 of media, about six times media spend before fees, salaries or delivery cost.

Now add time. Assume five months on average from first click to signature. The two deals signing this month were created by media that ran five months ago, at a different budget, with different creative. Comparing this month's spend to this month's closes is comparing two unrelated cohorts, and the comparison will swing wildly on two deals either way. That is the arithmetic reason quarterly cohorts beat monthly ROAS for firms like this.

Run the model and the weakest number usually announces itself. A 60% show rate puts the fix in confirmation and reminder sequences. Seven proposals converting at 30% puts it in the proposal and the follow-up, not the ad account. A $400 booked call with a 10% close rate downstream means the targeting is buying the wrong conversations. Measurement being the bottleneck is the common case, and marketers agree: HubSpot reports that measuring marketing ROI is the number one challenge marketers face in 2026, cited by 33% of respondents, with 73% saying their budget receives more scrutiny now than in the past. Any growth programme worth paying for ships the reporting layer as a deliverable, because the reporting layer is the thing the owner has to defend. Worth asking about before you sign: the vetting questions that get at how a partner reports, and which numbers actually predict revenue when they do.

If you take one operational thing from this, make it the cohort. Pick the month a lead first touched you, follow that group to closed revenue, and review it quarterly next to the spend that created it. It is slower and less satisfying than a live dashboard, it will be the only report in the building that survives a sceptical finance question, and it quietly settles the brand-versus-activation argument by showing which cohorts actually matured. Build that, then argue about channels. If you want a second pair of eyes on the model before you commit budget to it, apply here.

Sources

Stuck above the $10k ceiling?

We diagnose your account in 24 hours and tell you straight whether it's a creative, audience, or sales problem.

Talk to us