ABM metrics: Coverage, engagement, pipeline and win rate on target accounts, with the quarterly report to run

Account-based marketing (ABM) metrics are four layers measured only on your named target accounts: coverage, engagement, pipeline, and revenue with win rate. Report each against a matched control group. For example, suppose target accounts won 40% of closed deals against 18% for non-target accounts, a 2.2x lift. Keep the target list and the date each account became a target in the customer relationship management (CRM) system, so finance can rebuild the numbers.

What are the ABM metrics that matter

Four layers, each measured only on the named list of target accounts. Anything outside the list goes in a separate column for comparison, never in the same total. That rule keeps an ABM report from turning back into a lead report.

Engagement was the outcome the largest share of surveyed programs reported improving, and higher ROI the smallest of the four.

Share of ABM programs reporting improvement, by outcome

Item	Value
Active engagement with selected accounts	90%
Pipeline growth	84%
Revenue growth	77%
Higher ROI than other marketing	72%

Source: Demandbase, ABM Leadership Alliance and Momentum ITSMA benchmark release, 2022

Coverage starts with the list itself. Demandbase describes the number of target accounts as a metric that "tracks the total number of accounts you’ve identified as high-value targets for your ABM program". Coverage is our next step: of those accounts, how many have contacts you can reach.

For engagement, the same source defines an account engagement score as "A composite score that combines various engagement signals (website visits, content downloads, event attendance, email interactions, etc.)". Pick the signals once and keep them fixed, so the quarterly trend means something.

The last layer is a comparison. Demandbase cites TOPO: "opportunities from accounts receiving an account-based approach closed at 53% versus 19% for demand generation". Your own number replaces that one. Here is the arithmetic with illustrative figures.

Coverage rate   = target accounts with mapped buying group / total target accounts
                = 150 / 200 = 75%

Pipeline rate   = target accounts with an open opportunity / total target accounts
                = 40 / 200 = 20%

Win rate (target)     = closed-won / closed (won + lost), target accounts
                      = 12 / 30 = 40%
Win rate (non-target) = 45 / 250 = 18%

Win-rate lift   = 40% / 18% = 2.2x

Count only opportunities that closed in the quarter, won or lost, in both win rates. Open deals belong in the pipeline layer.

What practitioners reported

The Momentum ITSMA and ABM Leadership Alliance benchmark drew on 279 ABM heads and practitioners. Of the programs surveyed, 90% were improving active engagement with selected accounts, 84% were improving pipeline growth and 77% were seeing revenue growth. Another 72% said ABM delivers higher return on investment (ROI) than other types of marketing.

These are shares of respondents reporting an improvement, not the size of the improvement. In our reading, they show what programs claim. The four layers above show what your own program delivered.

Which ABM metrics come from the CRM and which come from platforms

Every ABM number starts in one system and gets copied into others. The rule we use: name that first system for each metric, and report the metric from there. Salesforce's ABM guide puts it plainly: all meaningful ABM metrics flow from the concept of target accounts. So the target list has to live in the CRM, not in a spreadsheet or a platform audience.

The same guide says to track the date on which the account became a target. That way you report only on activity from the period when the account was under ABM. Without that date, pipeline from before targeting gets counted as ABM pipeline.

Demandbase defines win rate by tier as the percentage of won opportunities from accounts in different tiers. Every input there sits in the CRM, which is why we treat it as the anchor number for a finance review.

Moving platform numbers toward the audited column

Platform metrics become auditable when you write a dated event into the CRM. For intent, the Salesforce guide suggests capturing the date that the account first showed intent via such signals. Then time from intent to meeting can be computed from two CRM dates. Our guide on how intent data works covers which signals to store.

Engagement has a second gap. Interactions happen with people, and the guide flags activities logged on duplicate accounts or activities logged on leads that aren’t attached to the right account. Fix lead-to-account matching before you report engagement by account.

A working test for the audit column:

Auditable = a finance analyst can rebuild the number
            from CRM records alone,
            filtered to target accounts,
            limited to dates on or after the target date.
If any input comes only from a vendor dashboard, mark it "platform-reported".

Label platform-reported numbers that way in the quarterly report, and keep them out of the pipeline and revenue lines.

What account engagement is and how it is scored

Account engagement measures how deep your relationship runs across an account's buying group. Raw traffic is a different thing. Demandbase describes engagement analytics as an ABM metric that measures "relationship quality, not quantity" (Demandbase). The same guide lists Key Contact Engagement as a separate signal that "Measures the interactions of specific decision-makers within target accounts" (Demandbase).

Depth means two things: how committed each action is, and how many people in the buying group take it. Forty page views from one intern are worth less than one meeting with three stakeholders. For typical group sizes, see our buying committee benchmarks.

How to build the weighted score

1. Pick signals you can tie to an account: website visits, ad clicks, content downloads, event attendance, sales meetings.
2. Weight each signal by the commitment it takes from the buyer. A meeting costs the buyer an hour. A click costs a second.
3. Multiply by breadth, the number of distinct engaged contacts, so one heavy user cannot carry the account.

Account engagement score (90-day window)
  = sum of (signal count x signal weight), all contacts
    x breadth multiplier

Breadth multiplier = 1 + 0.25 x (distinct engaged contacts - 1), capped at 2.0

Account A: 40 visits x 1 = 40, one contact, multiplier 1.0  -> score 40
Account B: 6 visits x 1 = 6
           3 ad clicks x 2 = 6
           2 meetings x 10 = 20
           subtotal 32, four contacts, multiplier 1.75       -> score 56

Account B outranks Account A with a fraction of the traffic. That is the point of weighting. To set your weights, adapt our lead scoring model template at the account level, and see lead scoring for how the person-level version differs.

Why a high score is not proof of preference or pipeline

Engagement and intent show activity, and activity alone does not tell you who wins. Forrester observes that most B2B marketers track these signals, yet they do not show whether buyers actually prefer the company (Forrester). The gap carries through to the pipeline. In Forrester's Demand, ABM, And Customer Marketing Survey, 2024, 56% of opportunities handed off to sales failed to close (Forrester).

In our reading, the score should be treated as a leading indicator that you test. Each quarter, compare the score band an account sat in against whether it reached a qualified opportunity. If high-score accounts convert no better than mid-score ones, change the weights before you report the score again.

How to compare target accounts with a control group

A rise in win rate on target accounts proves little until you know what the same kind of account did without ABM. A control group supplies that baseline. It is the account-level version of an incrementality test.

Build and freeze the two lists

1. Start from your ideal customer profile, which the Salesforce guide defines as "Your perfect buyer". Pull every account that fits it.
2. For each target account, pick a non-target account that matches on industry, employee band, region and current stage. Keep it out of every ABM program.
3. Freeze both lists on the same day. Record a target date on each target account, as covered earlier, and give each control account the same date as its match.
4. Count only opportunities created after that date, on both sides.

Compare by tier

Run the comparison separately for each tier, because the tiers differ in size and treatment. One-to-few ABM is applied to clusters of 5 to 15 accounts that share business issues or an industry. One-to-many covers hundreds of named accounts. One-to-one is reserved for the most strategic, high-value accounts. In our reading, one-to-many gives you enough accounts for a fair test within a few quarters. One-to-one rarely does, so treat its results as case notes, not proof.

Opportunity rate = accounts with a new opportunity after target date / accounts in group
Win rate         = closed-won opportunities / closed opportunities
Average deal size = closed-won amount / closed-won opportunities
Cycle length     = median days from opportunity created to closed-won
Lift             = (target metric - control metric) / control metric

Worked example

The figures below are illustrative. In this example, a one-to-many tier of 200 target accounts is set against 200 matched control accounts, four quarters after the freeze.

In this example, read the opportunity-rate lift first: 40 out of 200 against 24 out of 200. In this example, the win-rate gap rests on only 30 and 20 closed deals, so a few deals either way would change it. We report it, but we do not claim it until two or three quarters agree. Compare your control-group win rate with business-to-business win rate benchmarks to check that the baseline looks normal.

The same split also works for velocity. The guide lists pipeline velocity on target accounts as a common metric. Add the control column next to it, and the figure shows lift instead of a level. For paid channels inside the program, the incrementality testing method uses the same holdout logic.

A worked example with 60 target accounts and 60 control accounts

This is our own example, built to show the arithmetic. In this example, a one-to-few program ran for four quarters against 60 named accounts, grouped into clusters by industry. In this worked example, a matched control group of 60 accounts received the normal demand program only. The account lists were frozen at the start, so neither group changed during the year.

The four-quarter results

Every formula step

Coverage          = contacts reached / contacts planned
                  = 210 / 300 = 70%

Engagement rate   = engaged accounts / accounts in group
  Target          = 36 / 60 = 60%
  Control         = 15 / 60 = 25%

Opportunity rate  = opportunities / accounts in group
  Target          = 18 / 60 = 30%
  Control         = 9 / 60  = 15%

Pipeline          = opportunities x average opportunity value
  Target          = 18 x $80,000 = $1,440,000
  Control         = 9 x $70,000  = $630,000

Win rate          = closed-won / opportunities
  Target          = 7 / 18 = 38.9%
  Control         = 2 / 9  = 22.2%

Revenue           = closed-won x average opportunity value
  Target          = 7 x $80,000 = $560,000
  Control         = 2 x $70,000 = $140,000

Incremental revenue = target revenue - control revenue
                    = $560,000 - $140,000 = $420,000
  (valid only because both groups hold 60 accounts)

Incremental return  = incremental revenue / program cost
                    = $420,000 / $150,000 = 2.8x

Cost per engaged account = program cost / target engaged accounts
                         = $150,000 / 36 = $4,166.67

What the numbers say

Demandbase suggests measuring "cost per engaged account" to judge an ABM ad budget. We apply it to the whole program cost here, because one-to-few spend is mostly content and people, not media.

In our reading, the subtraction carries the useful part.

Three checks before the result goes into the quarterly report:

1. Equal group sizes. Suppose the control group had 40 accounts: convert both to revenue per account before subtracting.
2. Small counts. Seven wins against two is a thin base. One deal moving either way changes the win rate by several points, so report the counts next to the percentages.
3. Open pipeline. Eleven target opportunities are still open at year end. Carry them into the next report rather than counting them as lost.

To see how cycle length changes this pipeline's value per day, run the target and control figures through the pipeline velocity calculator. The cluster setup behind this example follows our one-to-few ABM playbook.

How to prove ABM ROI

ROI on an ABM program is a finance calculation, and it should survive a finance review. The numerator is the gross profit the target accounts produced above what they would have produced anyway. The denominator is everything the program cost, people included.

The formula

Incremental revenue   = Target-account closed revenue
                        - (Target-account pipeline x Control-group win rate)

Incremental gross profit = Incremental revenue x Gross margin

Full program cost     = Platform fees + Media + Content and creative
                        + Agency fees + Share of marketing and sales headcount
                          spent on target accounts

ABM ROI               = (Incremental gross profit - Full program cost)
                        / Full program cost

Three rules keep this honest:

1. Use the control group as the baseline. The win rate comes from the matched control accounts set up earlier on this page, not from company averages.
2. Use gross profit, not revenue. Revenue overstates the return by the cost of delivering what was sold.
3. Load the headcount. Seller and marketer time on target accounts is a program cost. In our reading, leaving it out is the most common way an ABM ROI figure inflates.

The size of the cost line matters. In Momentum ITSMA's study, 28% of the 2022 marketing budget is dedicated to ABM. A program that holds a share of budget that large needs a cost-loaded ROI, not a story.

Why survey ROI is not your ROI

The practitioner survey figures cited earlier, where most respondents said ABM beat other marketing on ROI, are self-reports. They record what program owners believe. They do not tell you the baseline, the margin or the cost each respondent used. The same study lists the most common challenges as measurement, customization and personalization at scale, the relationship with sales, and resources. In our reading, a belief in higher ROI sits alongside widespread trouble measuring it.

If roughly half do not measure it, survey answers about ROI include many unmeasured ones.

Provider ROI multiples

Agencies and platforms publish ROI multiples. Treat them as marketing claims, not planning inputs. One such page states that its headline figure is corroborated by The ABM Agency’s competitive positioning study of 17 major industry players, which is the provider's own analysis. You cannot see the baseline, the cost loading or which clients were left out.

Plan with your own control-group numbers. For payback timing on the same cost base, run the program cost through the CAC calculator and read it against CAC payback. If finance asks whether target-account revenue was sourced or influenced by marketing, settle the definition first with our sourced versus influenced method.

How long ABM takes to show results

ABM results arrive late because the deals they feed are long. Demandbase puts it plainly: ABM takes time to show results, so a program needs metrics that show progress inside large, complex deals across the whole sales cycle. Revenue is the final test, but it lands after the budget review that decides whether the program continues.

The same source flags a second effect that catches operators off guard. As the focus shifts to high-value accounts, quantity-based metrics such as leads may go down. A team that moves spend from broad lead capture to a named list will often report fewer leads in its first quarter. In our reading, that drop is expected when the list is narrow, and it is not a failure signal on its own. Agree on this before launch with whoever owns the lead target, so the decline does not end the program.

Maturity matters too. In the benchmark report released by the ABM Leadership Alliance and Momentum ITSMA, the organizations seeing the most success with ABM were two times more likely to have mature programs. We read that as a reason to judge a young program on early layers and hold the revenue verdict until it has run long enough to earn one.

Which layer to judge, and when

The answer is to grade each layer of the scorecard only once it has had time to move. Demandbase notes that ABM requires leading indicators to show progress during long sales cycles. The schedule below is our method, not a published benchmark.

Our sales cycle length benchmarks give a reference point, and pipeline velocity shows whether target deals are moving faster before they close.

Write the checkpoint dates into the quarterly report header at launch.

How to report ABM to the executive team each quarter

The executive team needs one page each quarter. It should show the four layers, the target-versus-control gap and one decision. Anything longer turns into a slide review, and nobody gets to the decision.

Name one owner before the first report

Give the report one named ABM owner who signs it. Forrester describes a common failure: teams fail to identify a designated ABM leader and spread the duties across people who already have full plates. The same piece warns against assuming that buying a single tool covers every ABM need. A platform supplies engagement data. It does not own the definitions, the control group or the call to cut a tier.

The owner also holds the shared definitions with sales. Forrester lists misalignment across marketing, sales, and ABM among the problems that limit results. Write the definitions into your marketing and sales SLA so that both teams count the same thing:

• What counts as a target account, and the date it entered the list
• What counts as an engaged account, using the score set earlier on this page
• Which opportunity stage counts as pipeline
• How a win is dated and attributed to a tier

The one-page template

ABM QUARTERLY REPORT  |  Quarter: ____  |  Owner: ____  |  Sales sign-off: ____

1. COVERAGE            Tier 1    Tier 2    Tier 3    Control
   Accounts in tier    ____      ____      ____      ____
   Key contacts mapped ____      ____      ____      ____

2. ENGAGEMENT
   Engaged accounts    ____      ____      ____      ____
   Engaged / total     ____      ____      ____      ____

3. PIPELINE
   Accounts w/ opp     ____      ____      ____      ____
   Pipeline value      ____      ____      ____      ____
   Cost per engaged    ____      ____      ____      n/a

4. WIN RATE
   Won / closed opps   ____      ____      ____      ____

GAP VS CONTROL (target minus control, per layer):  ____

DECISION OF THE QUARTER (one per tier):
   Tier 1: EXPAND / HOLD / CUT   Reason: ____
   Tier 2: EXPAND / HOLD / CUT   Reason: ____
   Tier 3: EXPAND / HOLD / CUT   Reason: ____

DEFINITIONS CHANGED THIS QUARTER: none / list
NEXT QUARTER'S ASK: budget ____  headcount ____  accounts added/removed ____

How to make the decision

Agree on the rule with the executive team before the quarter starts. Here is how we set it. Expand a tier when its gap over control holds or widens on pipeline and the cost per engaged account is flat or falling. Hold when engagement leads control but pipeline has not yet moved, which is the expected pattern early in long cycles. Cut when the gap has stayed at zero for the period you agreed in advance.

The quarterly report is that repetition written down. It also asks the room for something: Forrester notes that the executive teams must demonstrate commitment and discipline alongside sales, product and customer success. A dashboard can supply the numbers in each cell, and the demand gen dashboard template maps them. The signature and the decision line remain the owner's.

What this page does not cover

This page does not cover how to choose or compare ABM platforms, agencies or intent data vendors, and it does not set budget levels for a program. The ROI and win-rate figures in the examples are illustrative arithmetic, not benchmarks to plan against, and nothing here is financial or investment advice.