Account-Level vs. User-Level Campaign Targeting in B2B SaaS
Account and user-level signals answer different questions, and mixing them breaks both.
The gap between account-level and user-level data in B2B SaaS is structural. A single account produces many users, and each person's behavior, looked at alone, only tells part of the story. That gap is where most targeting decisions go wrong, and closing it, deliberately, is what separates campaigns that actually move the product from campaigns that just move a dashboard.
Account and User as Different Units of Measurement
An account and a user are not two sizes of the same thing. They measure different parts of the business. The account carries the commercial facts: who holds budget authority, whether the contract renews, whether the deal expands. The user carries the product fact: whether a specific person, on a specific day, does the thing the product was built to make possible.
One account can have a user who created it, a different user who added the payment method, another who uploaded the first batch of content, and a fourth who actually published something. Put together, a picture forms where looked at separately, none of them does.
This split also explains why attribution built around one contact fails so often. Tracking one of them and calling it the account's journey misses most of what actually happened. Account-level attribution exists to put back together what individual-user tracking tends to scatter.
What breaks when campaigns collapse this distinction
Campaigns fail in specific, repeatable ways when the targeting unit doesn't match the goal. Two patterns show up constantly, and a team can diagnose both once it knows to look for them.
The first: user-level targeting with no account context. A power user on a top-tier plan and a brand-new user in their first two weeks can produce the exact same aggregate adoption number. Send them the same message, and neither one moves, because the message was built for an average person who doesn't exist.
The second failure runs the other direction. Account-level targeting without any way to resolve down to the right person sends the message to the correct company, but it lands on whoever happens to sit in the CRM record. That contact might be a billing admin with no influence over product adoption, while the person whose behavior actually needs to change never sees the message.
There's a third problem that compounds both of the above: the denominator. Tools like Userlens address this by evaluating each person independently inside an approved campaign; they skip anyone the evidence doesn't support instead of messaging an entire cohort regardless of fit. That independent judgment at the moment of sending is what keeps dead accounts from masking where the real gap sits.
Three targeting levels that already exist
Account-based advertising has settled into three distinct levels of targeting, and each one sees a different slice of the account-user relationship. None of them is better on its own. Each fits a different job.
Level 1 is company-level, built on IP matching. Platforms running at this level include Demandbase (which also supports native people-based targeting), 6sense, Rollworks, The Trade Desk, BidTheatre, and Madison Logic.
Level 2 is person-level, built on identity matching. A contact list gets matched against advertising identity graphs, and ads reach named individuals across their devices. This level fits mid-market and enterprise deals where several stakeholders weigh in before a decision. It also allows person-level engagement tracking, something Level 1 simply can't offer.
Level 3 is intent-triggered. They tell you an account is researching. They don't tell you who inside that account is doing the research.
Account-level signals versus user-level behavior
Account-level signals and user-level behavior aren't competing versions of the same measurement. They answer two different questions. Account-level signals answer: is this account ready to expand, engage, or churn? User-level behavior answers: which person, right now, needs guidance to take the next step?
Account-level signals worth tracking include plan tier, seats used against seats licensed, which feature modules are turned on, days since the account last showed any activity, and whatever expansion or at-risk flags the database already carries. First use marks a beginning. It doesn't prove the habit stuck.
A single campaign aimed at that blended rate won't move either group, because it wasn't built for either group specifically.
The strongest ABM programs map both dimensions at once. That threshold exists to prevent single-threaded deals, where the whole relationship runs through one person who can leave, get reassigned, or simply go quiet.
What account-level targeting requires to work
Account-level targeting should lead when the goal is pipeline generation, expansion into new teams inside an existing account, or reaching a buying committee where several stakeholders each weigh in before a decision gets made.
Making it work takes real operational discipline. It also means measuring success at the account level: pipeline velocity, net revenue retention, contract value.
A three-tier structure keeps budget honest: 10 to 20 Tier 1 accounts get hyper-personalized treatment, 50 to 100 Tier 2 accounts get signal-triggered escalation, and 200 to 500 Tier 3 accounts sit on programmatic monitoring. That structure stops the common mistake of treating every account the same regardless of size or stage.
The buying committee itself usually runs six to ten stakeholders per account. Account-level automation exists precisely so you can reach that wider group without adding headcount for every account in the program.
Timing depends on intent signals converging, not firing on one alone. Response times in 2026 reflect that urgency once signals do converge: Tier 1 multi-signal convergence triggers a rep Slack notification within one hour and personalized outreach within 24 hours, while Tier 2 signals launch automated email sequences within four hours.
Clean data has to sit underneath this, or none of it works. If the CRM holds one version of the account list and the ad platform targets a different one, reporting can't connect the two, and the program turns into coordinated spam. Deduplication, filling in missing firmographic data, and mapping every contact to its account have to happen before any workflow gets built on top.
What signals justify leading with user-level targeting
User-level targeting should lead when the goal is feature adoption, activating a specific role inside an account that already converted, or walking an individual through a step their own behavior shows they haven't taken yet.
The first decision is the denominator. If you measure against every registered user, almost any campaign looks like it's underperforming. Measuring against active, eligible users, the people for whom the feature actually applies given their plan, role, or past behavior, produces the real gap.
First use paired with repeat use is the minimum signal worth acting on. A user who triggers a feature once and never comes back hasn't adopted anything. A campaign built to fire on first use and then stop watching has optimized for the wrong moment.
Person-level targeting at the ad level lets user-level work run alongside an ABM program. A CFO sees a message about cost reduction. An IT director sees one about integration. An end user sees one about the specific task the product makes faster. All three sit inside the same account, reached by the same campaign, but with three different messages.
The evidence needed to justify a message at the user level differs from the evidence needed at the account level. An adoption agent that reads both database state and behavior over time can write that grounded message to the one person at the one moment it matters, rather than broadcasting identical content to a whole segment and hoping the aggregate numbers justify it afterward.
Combining both levels: how state and behavior connect account readiness to individual guidance
The most precise campaigns run on both dimensions at once. Current database state, what's true about the account right now: plan, configuration, seats, utilization, sets the context. Behavioral history, what a specific user has and hasn't done over time, sets the action. Together they produce guidance grounded in both facts at the same time, not one or the other.
The mechanism is straightforward. Account-level state defines who's even eligible and what context applies. User-level behavior defines what the message should recommend, and whether the evidence backing it is strong enough to send anything. A weak signal should produce silence, not a generic message sent out of habit.
Picture an account where usage of one module crosses a threshold the data shows predicts interest in a second module. That crossing is an account-level event, it triggers the campaign. But the message itself goes to the specific user whose role makes them the right person to act on it, not to the account record, and not to whoever happens to be the primary CRM contact.
Most lifecycle campaigns break down exactly at this seam. Account-level intent fires a message to whoever sits in the CRM. User-level segmentation fires a message to whoever matches a behavioral cohort. Neither one checks whether that particular person is the right one, at that particular account, at that particular moment. ABM automation's real value is coordinating across that seam: running person-level campaigns against account-defined audiences, so named accounts get consistent presence across channels while the individuals inside them get messages calibrated to their role and where they stand in the product.
Measuring the combination correctly means keeping account-level outcomes, expansion ARR, net revenue retention, account-level feature coverage, separate from user-level outcomes: first use, repeat use, role-specific activation rate within the eligible cohort. And it means resisting the urge to report a behavior change as proof of causal lift when it might just be coincidence sitting next to a campaign.
How Userlens approaches the account-and-user problem
Tools built for this problem split into two camps: descriptive instruments that hand a team a chart or a segment and leave the judgment to a human, and prescriptive instruments that act on the combination of account state and individual behavior to produce one grounded message for one specific person.
Userlens, through its product Lumi, was built for product and PLG teams at B2B SaaS companies that need to apply founder-level attention to individual users at account scale. It connects backend database state, accounts, people, roles, plans, configuration, what's true right now, with product behavioral data, what a person has and hasn't done over time. A human sets a campaign's goal, audience, guidance, and limits once. Inside that approved campaign, Lumi writes and sends a grounded personal message only when the evidence supports it, then checks product data afterward to see whether behavior actually changed. Frequency caps, quiet hours, never-contact flags, and skip-if-unsure logic get enforced at the moment of sending, built into the structure.
Account-based advertising platforms, the IP-based and identity-based tools described earlier, Demandbase, 6sense (a full ABM and revenue intelligence platform where paid media is one piece), Rollworks, and comparable person-level platforms, handle the paid media side of this problem. What they don't do is reach into product usage. They can't tell an eligible user who's never touched a feature apart from one who uses it daily, because that data lives inside the product, not in an ad platform.
ABM automation platforms coordinate advertising, outreach, and content against the same named accounts, cutting out manual repetition in building audiences, sequencing messages, and reporting results. As the team at Hey Sid puts it, automation scales coordination, not volume: real ABM automation delivers the same personalized message consistently across channels to every stakeholder who matters, rather than just pushing more messages to more people. These platforms handle multi-channel coordination and account-level intent well. They typically stop short of product usage data, so they can't confirm whether the specific user targeted actually completed the adoption step the campaign was built to drive.
Putting the framework into practice: choosing the targeting unit for a given campaign goal
The targeting unit should follow from the campaign's goal. Not from whatever tool happens to be open, and not from a platform's default setting.
When the goal is pipeline generation or engaging a buying committee before a sale closes, lead with account-level targeting. Use intent signals to time activation, map the committee by role, and measure success at the account level: pipeline velocity, influenced pipeline value. Person-level identity targeting inside those accounts handles the role-specific messaging once the account itself has defined who's eligible.
When the goal is feature adoption or activation inside accounts that already converted, lead with user-level behavioral evidence. Let account state define eligibility and context, and send a message only when both together justify it. Measure against the eligible cohort, active users in the relevant plan or role who haven't completed the target action yet, not against every registered user or every seat on the account. Track first-use rate alongside repeat-use rate; optimizing for first use alone optimizes for the wrong moment.
When both goals run at once, expansion into a new team inside an account that already uses the product, use account-level state to spot the expansion signal and user-level behavior to find the right person inside that new team to reach first. This is the combination case described throughout this piece: account state says the account is ready, behavioral data says who to contact, and the message gets grounded in both, rather than broadcast to the account or nurtured generically toward a behavioral segment.
Before any campaign launches, five questions are worth running through every time: Does the targeting unit match the campaign goal? Is the denominator the eligible cohort, not total signups? Does account state factor into eligibility? Does a weak behavioral signal produce no message? Are account-level and user-level outcomes being measured separately? Running that checklist consistently is what keeps the gap between account and user from quietly swallowing a campaign's results.
