How We Built an ABM Campaign That Generated $3M in Pipeline for AirOps

Six months. $3M in pipeline. $500K closed-won. That is what Ivan Falco delivered running an ABM campaign for AirOps at ColdIQ.

This was not a spray-and-pray outbound motion. It was a tightly coordinated account-based campaign where every target account received tailored messaging across multiple channels, where every touchpoint was personalized to the specific person and their buying stage, and where every signal fed back into the system to sharpen the next one.

ABM gets talked about a lot. The execution is where it falls apart for almost everyone. The gap between "we should do ABM" and "we generated $3M in pipeline from ABM" is entirely in the details.

Here is the exact seven-step process Ivan followed, broken down so you can replicate it.

1. Start With Your Best Customers

The first step had nothing to do with prospecting. Ivan started by analyzing AirOps' existing customer base to find patterns in who was already buying and generating the most revenue.

He pulled the full customer list and looked at three dimensions:

→ Average contract value: which customers were paying the most

→ Retention and expansion: which customers renewed and expanded over time

→ Sales cycle length: which customers closed fastest with the least friction

The intersection of these three signals reveals your ideal account profile. Not a theoretical ICP built from assumptions, but a data-backed profile built from real revenue.

For AirOps, the pattern was clear. Mid-market SaaS companies with content-heavy go-to-market motions were closing faster, paying more, and staying longer. They had marketing teams producing high volumes of content and needed AI workflows to scale production without adding headcount.

This step matters because it defines everything downstream. If you start with a weak ICP, every dollar spent on ads, every hour spent on personalization, and every email sent is pointed at the wrong accounts. Getting this right is what separates campaigns that generate $3M in pipeline from campaigns that generate noise.

Clay was the backbone of this analysis. Ivan pulled firmographic and revenue data into Clay tables, ran enrichment across the customer base, and used AI-generated columns to tag common attributes. The output was a scored profile of exactly what a high-value AirOps customer looks like.

2. Build a Lookalike Target List

With the ideal profile defined, the next step was finding more accounts that matched it.

Ivan mapped the total addressable market for AirOps and then filtered aggressively. The goal was not a large list. It was a precise list of accounts that closely resembled the best existing customers.

The filtering criteria included:

→ Company size between 50 and 500 employees

→ SaaS or technology vertical

→ Content marketing as a core growth channel (visible from blog volume, social activity, and job postings for content roles)

→ Series A through Series C funding stage

→ Based in North America or Western Europe

From a TAM of roughly 15,000 companies, Ivan narrowed the target list to approximately 800 accounts. That is a 95% reduction. Every account on that list was worth the extra effort of full personalization because they closely matched the profile that was already generating revenue.

This is where lookalike modeling becomes critical. You are not guessing which companies might be a fit. You are finding companies that share the same characteristics as your proven winners.

You can find companies that match your best customers here, for free:

Lookalike Finder Tool

Clay handled the list building and enrichment. Ivan used Clay's integration with multiple data providers to pull in firmographic data, technographic signals, hiring patterns, and funding history for each account. The enrichment layer is what made the filtering possible at this level of granularity.

3. Map the Buying Committee

Targeting the right accounts is half the equation. Targeting the right people inside those accounts is the other half.

Ivan broke down each target account into buying groups. For AirOps, the typical buying committee included:

→ VP of Marketing or CMO: the budget holder who cares about content output and team efficiency

→ Head of Content or Content Director: the day-to-day operator who feels the pain of scaling content production

→ Head of Growth or Demand Gen: the person connecting content efforts to pipeline metrics

→ VP of Engineering or CTO: the technical evaluator who needs to approve AI workflow integrations

Each persona has different priorities, different objections, and different language. The VP of Marketing cares about output volume and team productivity. The Head of Content cares about workflow friction and quality control. The CTO cares about security, integrations, and implementation complexity.

Ivan researched each persona deeply. He read their LinkedIn posts, reviewed their company's content strategy, studied their tech stack, and looked at what they were hiring for. This research informed the messaging for each persona so that every touchpoint spoke directly to their specific concerns.

Writing one generic message and sending it to four different personas is not ABM. It is outbound with a nicer label. The entire point is crafting distinct messaging for each role.

You can identify the decision-makers at your target accounts here, for free:

People Finder Tool

4. Personalize Every Touchpoint

With the account list built and personas mapped, Ivan personalized every channel for each account and role.

This went beyond inserting a first name and company name into a template. Each touchpoint was crafted to reflect something specific about the account:

→ LinkedIn ads referenced the prospect's industry and the specific challenge their role faces

→ Landing pages were tailored by vertical, showing case studies and messaging relevant to the prospect's company type

→ Email sequences referenced the account's content strategy, their tech stack, recent company news, or a specific initiative visible from their website or LinkedIn activity

For the email layer, Ivan used Clay to generate AI-powered personalization variables. Clay's Claygent feature scraped each company's website, summarized their content strategy, and identified specific pain points that AirOps could solve. These variables were then injected into email templates so each message felt hand-written.

Instantly handled the email sending, managing deliverability, warmup, rotation across sending accounts, and sequence timing. The combination of Clay for enrichment and Instantly for execution meant Ivan could deliver hand-crafted-feeling emails at scale without spending hours manually writing each one.

Expandi managed the LinkedIn outreach layer. Connection requests and follow-up messages were coordinated with the email sequences so prospects received touchpoints across both channels without overlap or awkward timing.

The personalization had to be specific enough that the prospect could not tell it was automated. If someone reads your email and thinks "this could have been sent to anyone," the personalization failed.

Fibbler tracked which target accounts engaged with the LinkedIn ad campaigns, giving Ivan visibility into which accounts were warming up before outbound even started. This data helped prioritize which accounts to push harder on and which needed more nurturing.

5. Sync Channels Into One System

Running ads, outbound, content, and LinkedIn outreach in parallel is not ABM unless those channels are feeding each other. The syncing is what makes the system compound.

Here is how Ivan connected the channels:

→ When a target account engaged with a LinkedIn ad, that signal triggered a more aggressive outbound cadence for that account

→ When a prospect opened an email but did not reply, they were added to a retargeting audience for LinkedIn ads

→ When someone from a target account visited the AirOps website, that visit was logged and used to personalize the next outreach touchpoint

→ Content published on LinkedIn was designed to address the same pain points being referenced in the email sequences

Every channel generated data that informed the others. The ad engagement data told outbound which accounts were warming up. The outbound reply data told the ad team which messaging resonated. The content engagement data revealed which topics drove the most interest from target accounts.

Attio served as the CRM layer, tracking every interaction across channels in one place. Every email open, ad click, website visit, and LinkedIn engagement was logged against the account record. This gave Ivan a complete picture of where each account stood in the buying journey.

Usermaven handled analytics and attribution, showing which channel combinations were driving the most pipeline. This was critical for understanding which sequences of touchpoints were converting and which were wasted effort.

Without channel syncing, you are running four separate campaigns that happen to target the same accounts. With syncing, you are running one coordinated campaign across four surfaces. The difference in conversion rates is significant.

6. Adjust Messaging by Buying Stage

Not every account in the target list is at the same stage. Some have never heard of you. Some have engaged with your content for weeks. Some are actively evaluating solutions. The messaging needs to reflect where they are.

Ivan segmented the 800 target accounts into three stages based on engagement data:

→ Cold accounts: no engagement signals. These received awareness-focused messaging that introduced the problem AirOps solves, with no hard ask. The goal was to get them to engage with content or visit the website.

→ Warm accounts: engaged with ads, opened emails, or visited the website. These received messaging that referenced their engagement and introduced specific use cases. The CTA shifted from "learn more" to "see how companies like yours are using this."

→ Hot accounts: multiple engagement signals across channels. These received direct outreach with case studies matched to their industry and company size, with a clear meeting request.

The messaging at each stage was different in tone, content, and ask. Sending a meeting request to a cold account wastes the touchpoint. Sending awareness content to a hot account wastes the momentum.

Clay powered the segmentation logic. Engagement signals from Fibbler, Attio, and Usermaven were pulled into Clay and used to score each account's engagement level. Based on the score, accounts were automatically routed into the appropriate messaging track.

This dynamic routing is what makes ABM campaigns improve over time. As accounts move from cold to warm to hot, the messaging evolves with them. The campaign gets smarter the longer it runs because every interaction adds data to the scoring model.

7. Iterate Relentlessly

The campaign did not start generating $3M in pipeline on day one. The first month was about establishing baselines. Which ad creatives performed. Which email subject lines got opens. Which personas responded to which messaging angles.

Ivan reviewed performance weekly and made adjustments:

→ Ad creatives that underperformed were replaced within 7 days

→ Email sequences with low reply rates were rewritten with different angles

→ Personas that were not engaging were deprioritized in favor of those that were

→ Channel allocation shifted based on where the pipeline was coming from

By month three, the campaign was operating at a fundamentally different level than where it started. The messaging was sharper because weeks of engagement data had revealed what resonated. The targeting was tighter because low-fit accounts had been pruned. The channel mix was optimized because attribution data showed exactly which combinations were converting.

The final six-month results:

→ $3M in qualified pipeline

→ $500K in closed-won revenue

→ Consistent month-over-month improvement in conversion rates

The $500K in closed revenue from $3M in pipeline represents a 16.7% conversion rate, which is strong for enterprise B2B. The remaining pipeline was still in various stages of the sales cycle at the six-month mark, meaning the total closed revenue from this campaign will continue to grow.

8. Conclusion

ABM is not a channel. It is a coordination system. The $3M in pipeline that Ivan generated for AirOps did not come from running better ads or writing better emails in isolation. It came from connecting every channel, every touchpoint, and every data signal into one system where each piece reinforced the others.

The seven steps are straightforward. Start with your best customers. Build a lookalike list. Map the buying committee. Personalize every touchpoint. Sync all channels. Adjust by buying stage. Iterate on everything.

The difficulty is in the execution. Each step requires infrastructure, data, and coordination across tools and teams. Clay sits at the center handling enrichment and segmentation. Instantly and Expandi handle outbound execution across email and LinkedIn. Fibbler, Usermaven, and Attio provide the engagement tracking and CRM infrastructure.

The system compounds over time. Month one sets baselines. Month three has sharp messaging and pruned targeting. Month six has $3M in pipeline. The teams that commit to the full process and iterate through the early noise are the ones that see these results.

If your outbound is running in isolation from your ads and content, you are leaving pipeline on the table. The opportunity is in connecting the motions.