Back to Scaler Labs
18 MIN READNov 15, 2026

Scaler Labs #1: LinkedIn ABM with Multiple Entry Points, from Onboarding to Week One

The diary of a B2B launch for a client from our operation in Brazil, in the regulated financial market. How the campaign was built around multiple entry points and reach, and what the first week showed. Documented by Scaler, a B2B marketing agency serving companies in South Florida.

Experiment cardWorked
Hypothesis
With an audience of 362 target accounts, a campaign optimized for reach, with every message angle shown to everyone, covers most accounts with well-distributed delivery and starts to reveal which door each account comes in through.
Period
Nov 6 to Nov 13, 2026 (week one of the campaign)
What we did
Launched a LinkedIn Ads ABM campaign with four angles in even rotation, a brand awareness objective, a split by company size, a skills-based ad set, a launch webinar and a short form for the sales angle.
Metric
Companies reached, delivery concentration, impressions per day, CTR by angle, webinar sign-ups and leads.
Result
About 340 companies with impressions out of a list of 362, with the top 10 taking about a quarter of delivery. About 1,070 impressions per day, 38 clicks in 8 days, 100 webinar sign-ups and the first lead, now being qualified.
Learning
Optimizing for reach spread delivery across the target accounts, and clicks came anyway. The difference between angles does not show up in average CTR. The reading that matters is at the account level, outside LinkedIn.

Summary

This is the diary of a real project, told from onboarding to the first week of results. The client comes from our operation in Brazil: a B2B software company that sells to the regulated financial market. It launched an application built on AI agents for compliance and anti-money laundering (AML) teams. Scaler Marketing designed and runs the launch campaign on LinkedIn Ads, as account-based marketing (ABM), for 362 target accounts.

Instead of a funnel, the campaign uses a multiple entry points structure: four message angles shown at the same time to the whole audience, optimized for reach.

What the first week showed (Nov 6 to Nov 13, 2026):

  • About 340 companies received impressions, out of a list of 362 target accounts.
  • Delivery was well distributed: the 10 companies with the most impressions took about a quarter of the total.
  • The campaign delivered about 1,070 impressions per day and generated 38 clicks in 8 days, without optimizing for clicks.
  • The launch webinar had 100 sign-ups.
  • The short three-field form generated the first lead for the new application, which is now being qualified.
  • What we cannot claim yet: whether each angle attracts a different type of account, and whether the entry signal improves the sales conversation.

Why the funnel does not describe a complex B2B sale

The classic paid media playbook follows a sequence: educate at the top, nurture in the middle, convert at the bottom. It works reasonably well for individual, fast purchases. In complex sales, with high ticket sizes, cycles that last months and a buying committee, that straight line does not happen.

Market data helps size the problem. Dreamdata (2026) reports an average journey of 272 days, with 81% of it happening before the deal enters the pipeline. Forrester (2026) reports an average of 13 internal people and 9 external influencers in each decision. 6sense (2025) reports that 95% of purchases go to a vendor that was already on the buyer's list on day one. These numbers come from companies that sell platforms to this market, so we treat them as context, not proof.

In practice, inside a single account, one person has not thought about the problem yet, another knows it exists and wants to measure it, a third needs proof before moving forward and a fourth is ready to talk. A funnel forces all of them through the same steps, in the same order.

Scaler's thesis is that the winning brand builds authority with the whole committee before there is mapped demand, and recognizes interest through the signals each account gives. This campaign is a test of that thesis, documented in the open.

The client and the challenge

The client is a B2B software company that sells to the regulated financial market in Brazil. Its main offer is a risk management hub that monitors the client portfolio after onboarding. Within it, there was already a line focused on anti-money laundering. The launch is a new application, built on AI agents, that joins the same offer as another line.

The audience is small and very specific: compliance, AML, risk and internal controls professionals, across a few hundred institutions. The sales cycle is long and the decision is made by a committee.

With an audience this size, the central question was not "how many clicks can we get", but "how many of the right accounts can we reach, and which topic is each one interested in". Instead of a separate campaign for the launch, we built a new campaign under the main offer, with both lines as divisions.

The six steps of the method

Every Scaler client goes through the same process: map, publish, distribute, read, engage and learn. The sections below show what we did in each one, in this campaign.

Step 1, map: accounts, job titles and the company size problem

Before opening the ads manager, we defined the ideal customer, the target accounts, the committee roles and the pains of each role.

The accounts. The base is a list of 362 accounts recognized by LinkedIn, across six high-fit segments: banks, credit cooperatives, payments and crypto, credit and receivables funds, insurance and pension, and asset managers and broker-dealers. Individual credit cooperatives have their own LinkedIn pages, so we can treat them as independent accounts rather than part of a larger system.

The job titles. On top of the list, we applied 68 job titles in compliance, AML, risk, internal controls and audit. We excluded active clients, people who had already said they were not interested, out-of-profile titles and employees of competitors and data vendors.

Company size. When we measured the audience by company size, a strong imbalance showed up:

Company sizeShare of audience members
More than 10,000 employees68%
5,001 to 10,0005%
501 to 5,00017%
Up to 5009%

Companies with more than 5,000 employees account for about 73% of the people in the audience, even though they are few accounts. Without addressing this, the large banks would take almost the entire budget.

Who job titles miss. Many relevant people have a generic title, such as "manager" or "specialist". LinkedIn does not allow targeting by free profile text, but it does allow targeting by skills, which it infers from the title and description. A check with 9 skills from the KYC and AML world, manager level and above, added about 50% more people. It also added noise: close to 40% were in sales and customer service, the typical branch manager who took mandatory AML training.

Step 2, publish: one angle for each awareness level

Instead of organizing the campaign by funnel stages, we organized it by angles. Each angle speaks to a different level of buyer awareness.

AngleFor whomWhat the person gets
EducationalPeople still understanding the problem and what regulation requiresTechnical white paper or product page
DiagnosticPeople who recognize the problem and want to measure it in their own operationA free prompt to use in a corporate AI tool and an invitation to calculate hours spent on manual searches
ProofPeople who need evidence before moving forwardA client case, with no ads yet: the ad set stays paused until there is an approved case
SalesPeople ready to talkNative demo form with three fields

The rule is to show every angle to everyone. There is no required sequence. People show their level through the angle they engage with, and that engagement is the most useful signal in the campaign.

Each angle has two to four creatives, in square and horizontal formats. The copy follows a tone guide: open with the problem, use the language of people who work in compliance and promise nothing the product does not deliver yet.

On the proof angle, one decision worth recording: the new application's only client is recent and the case is not mature yet. We chose not to force it. A weak case published now would be worse than none.

The launch also included a live LinkedIn webinar on Nov 12. It did not go into the proof angle, because it opens with the problem and is not social proof. Placed there, it would distort the reading of how each account came in.

Step 3, distribute: the setup decisions

The thesis only holds if the campaign setup does not work against it. These were the main decisions, made when the campaign was built.

Optimize for reach. If the goal is for the right people, at the right accounts, to see the message, the delivery criterion has to be reach. Clicks are a consequence of the message and of demand. In a small audience, optimizing for clicks makes the platform pay a lot for the few people likely to click and leaves the rest of the accounts uncovered. The educational, diagnostic and proof angles, plus the company size and skills ad sets, run with a brand awareness objective.

One ad set per angle. If all angles sat in one ad set, LinkedIn would concentrate delivery on the angle with the most clicks and, in practice, close the other entry points.

Even ad rotation. The algorithm does not pick a winner, because the question is not which creative performs best, but which door each account comes in through.

Split by company size, to control budget. The angle ad sets target companies with up to 5,000 employees. A separate ad set targets the larger ones, with a locked budget cap and ads from every angle. We considered three size tiers and decided against it for now: with the current budget, each ad set would have too little money to generate data.

A separate skills ad set. LinkedIn does not accept "job titles OR (skills AND seniority)" in the same ad set. Merging everything would mean dropping seniority, and the skills audience would grow from about 3,000 to about 18,000 people, almost all out of profile. The solution was a dedicated ad set, with skills and manager-level seniority and above, excluding people already reached through job titles and the sales, marketing, HR and customer service functions.

Sales in a separate campaign. On LinkedIn, the objective is set at the campaign level, and every ad set inherits it. The sales angle uses a native form, with no site visit, so it runs in a lead generation campaign. The new application's form has only three fields: name, work email and phone. On the AML line, the sales form includes qualification questions. Running both versions side by side shows the effect of friction.

Budget logic. The new application received about 64% of the budget through year end, and the AML line about 36%. The AML line runs at the minimum, but it does not disappear: each of its angles runs at the lowest daily budget the platform accepts, so no entry point closes. The launch's active angles receive the same amount, because the thesis does not favor any angle. Each ad set has a cap calculated to last through year end.

Step 4, read: what the first week showed

The window runs from Nov 6, when the campaign went live, to Nov 13. The AML line ran the whole week. The new application started on Nov 9, so it had four and a half days. This is a first-week reading, with small samples.

MetricWeek one
Companies with impressionsAbout 340, out of a list of 362
Share of the top 10 companies by impressionsAbout a quarter of the total
Impressions per dayAbout 1,070
Clicks38 in 8 days
CTR of the three reach anglesAbout 0.43% each
Webinar sign-ups100
Sales leads1, being qualified

Coverage. LinkedIn's company demographics report shows about 340 companies with impressions in the week. The caveat: LinkedIn attributes the impression to the member's current employer, so some of these companies may be outside the list. The number still needs to be matched against the list before it becomes an official metric.

Concentration. The 10 companies with the most impressions account for about a quarter of the total, far from the alert set in the plan (10 accounts with more than half). The top company received about 9%.

Clicks without optimizing for clicks. The campaign does not ask LinkedIn for clicks, and they came anyway: 38 in the week. Our reading: when the budget spreads across the right audience, part of it clicks.

Angles. The three brand awareness angles had virtually the same CTR, around 0.43% each, with 7, 7 and 3 clicks. That is a small sample. It does not weaken the thesis, because multiple entry points do not depend on one angle beating the others. They depend on knowing which door each account came in through, and that answer is at the account level, not in the ad set CTR. The sales angle had a CTR close to 1.9%, but it runs in another campaign, with another objective, and is not comparable with the others.

Large companies. The ad set for companies with more than 5,000 employees had the lowest CPM of all, less than a third of the angle average, and one of the lowest CTRs, 0.15%. Many people, cheap delivery, little reaction. Since the cap is locked, this does not take budget away from smaller companies.

Skills. The ad set had the highest CPM among the brand awareness ad sets, about 27% above the angle average, and a CTR of 0.12%, with one click in the week. We already treated this ad set as a bet, not a validated learning. For now, the data goes against it.

Webinar. The event ad set had the most expensive CPM of the week, about twice the angle average, and the best CTR outside the sales angle, 0.81%. The live session had 100 sign-ups.

Short form. The three-field form generated one lead, which is now being qualified. The form with qualification questions, on the AML line, has not generated any. The ad with the short form also had a CTR about 3.6 times higher. One lead proves nothing, but it is the first sales conversion.

Step 5, engage: from signal to sales conversation

If the angle indicates the awareness level, it also guides the sales approach. An account that engaged through the diagnostic angle calls for a different conversation than one that engaged through the educational angle.

To capture that signal, measurement was standardized from day one:

  • The angle goes in the URL, in the utm_term parameter. Every visit is tagged with its entry point.
  • The creative and the offer line go in utm_content.
  • LinkedIn's automatic parameters identify the campaign, ad set and ad. They are what separates clicks from the job title audience, the skills audience and large companies.
  • Native forms carry the UTMs in hidden fields, so the lead reaches the CRM (HubSpot) with the same tagging as a site visit.
  • An account-level attribution tool (Factors.ai) matches all of this against the target account list and shows which companies are engaging, even without filling out a form.

The goal is for each target account to build a history over the weeks, such as "came in through diagnostic, came back through proof", and for the sales development team to use it in their approach.

The first lead already shows why this matters. The person attended the webinar one day and, the next day, converted on the sales ad. With last-click attribution, all the credit would go to the form, and the role of the event would be invisible. The lead is being qualified, and the next step is to follow whether it becomes a meeting.

A limit showed up here. LinkedIn's company report shows who saw the ads, but not who reacted: every row shows zero clicks, even with 38 clicks in the campaign. The most important part of the thesis, knowing which account engaged and through which angle, depends on the attribution tool, outside the media platform.

Step 6, learn: what worked, what has not paid off yet and what changed

What worked:

  • Optimizing for reach. Delivery spread across the target accounts, and clicks came anyway.
  • Splitting by company size. Concentration stayed under control.
  • Standardizing UTMs around the angle from day one. It is easy to read where each visit comes in.
  • Keeping the AML line alive at the minimum. The thesis is tested on both lines at the same time.
  • The short form, with the first sales lead.

What did not work or has not paid off yet:

  • The skills audience. CPM about 27% above the angle average, a 0.12% CTR and one click in the week. It is the first candidate to be cut in the December review, if the job function breakdown confirms the noise.
  • Large companies engage little. Delivery is cheap, but CTR stayed at 0.15%. We still need to see whether those impressions turn into engaged accounts in the attribution tool.
  • The angles have not separated yet. CTR was the same across them. Without the account-level reading, the central question of the thesis is still open.
  • The proof angle is paused. Without a mature case, one of the four entry points has not opened yet for the new application.
  • The form with qualification questions, on the AML line, generated no leads in the week.
  • The account-level reading depends on another tool. LinkedIn shows who saw the ads, but not who reacted. The account-level match is not closed yet.

What changed for week two. The delivery history showed that Saturdays and Sundays took about a quarter of the budget, with a CPM about 28% higher and a cost per click about 45% higher. Engagement rate barely changed: people who see the ads on weekends react at the same rate, they just cost more to reach. The campaign now pauses from Friday at 6 pm to Monday at 6 am, through the end of December, as a test. Since the budget is a lifetime budget with an end date, what is not spent on weekends moves to weekdays. This creates a new variable: from now on, any week-over-week comparison has to account for the missing weekends.

Hypotheses being tested: where we stand

HypothesisHow we measure itStatus in week one
Multiple entry points: each angle attracts a different type of accountEngagement by angle and which accounts came in through which angleInconclusive. CTR was the same across angles; the account-level reading is missing
Optimizing for reach covers the target accounts with distributed deliveryAccounts with impressions and delivery concentrationSupported. About 340 companies reached and the top 10 with about a quarter of delivery
Splitting by company size keeps large companies from taking the budgetDemographics by company sizeWorking for the budget
The skills audience expands coverage with acceptable qualityDistribution by job function and seniority, and CTR compared with the job title audienceAgainst, for now
A three-field form generates more leads than one with qualification questionsLeads, form conversion and meetingsFirst signal in favor: 1 lead versus 0
Knowing the entry angle improves the sales approachMeetings from engaged accounts, by angleNo data yet

Coming in the next edition

The account-level reading in the attribution tool (which companies engaged and through which angle), the distribution of impressions by job function and seniority, the first paused weekend and the path of the first lead to a meeting.

Frequently asked questions

What is ABM with multiple entry points?

It is a way to organize paid media in account-based marketing where every message angle (educational, diagnostic, proof and sales) is shown at the same time to the whole target account audience, instead of following a funnel sequence. Each person shows their awareness level through the angle they engage with, and that signal guides the sales approach.

Why optimize for reach in a LinkedIn ABM campaign?

Because in a small, specific audience, the goal is for the right people at the right accounts to see the message. Optimizing for clicks makes the platform concentrate delivery on the few people likely to click. In this campaign, optimizing for reach delivered impressions to about 340 companies in the first week, with well-distributed delivery, and clicks came anyway.

Does LinkedIn show which companies engaged with the ads?

Not fully. The company demographics report shows how many impressions each company received, but it does not include clicks or engagement by company. To know which accounts engaged, and through which angle, you need to match UTMs with an account-level attribution tool.

Do short forms generate lower-quality leads?

We cannot answer that yet. In this campaign, the three-field form generated the first lead, which is now being qualified, while the form with qualification questions had none. Quality will be measured by how many leads turn into meetings.

Does this structure work for any B2B company?

It makes the most sense for companies with complex sales, a buying committee, a defined list of target accounts and a sales team to follow up on engagement. For companies that need high lead volume in the short term, or sell transactionally, other models tend to work better.

Where does Scaler work?

Scaler Marketing works with B2B companies with complex sales in South Florida and, through our operation in Brazil, across Brazil. This project comes from our Brazilian operation: a client that sells to financial institutions across the country.


This is the first Scaler Labs entry on this campaign. The next ones bring the numbers from the following weeks, including the ones that are not good.

If your company sells to a buying committee in South Florida and you want to know whether this method fits your accounts, .

Commercial Diagnostic

We analyze accounts already engaging with your content and deliver high-intent signals before sales outreach.