# Account based marketing for SaaS

> How SaaS ABM really works: account selection, cost per account by tier, orchestration with sales, and the deal size floor below which the maths stops working.

Source: https://saas-marketing.net/guides/account-based-marketing-saas/
Topic: SaaS Demand Generation
Type: guide
Published: 2026-09-11
Last updated: 2026-09-17
Publisher: SaaS Marketing (saas-marketing.net)
License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/guides/account-based-marketing-saas/

## Short answer

Account based marketing for SaaS is a named account program where marketing and sales work one agreed list instead of chasing inbound volume. Loaded cost runs roughly $15,000 to $40,000 per account per year at 1:1, $3,000 to $8,000 at 1:few and $300 to $900 at 1:many. The economics clear only when annual contract value is above about $25,000 and named reps commit to the list in writing. Expect nine to twelve months before pipeline moves.

## Key takeaways

- Annual cost per account should stay under a third of expected gross profit per account per year, or the program cannot pay back.
- Media is about a third of real ABM cost. Research, content, people time and events are the rest and never appear on a platform invoice.
- Below roughly $25,000 ACV only the 1:many tier survives the arithmetic, which makes it outbound with better targeting.
- A named list needs open pipeline coverage of three times its bookings target before anyone should call the program healthy.
- If sales will not name a rep against every account in writing, do not launch. The list is the contract, not the software.
- Hold out 100 fit matched accounts you deliberately never target. That control group is the only clean read on whether ABM worked.

---

Most ABM programs die before the first ad runs. Marketing builds a list of 400 accounts, licenses intent data, ships a few branded boxes, and discovers in month four that no account executive ever agreed to work the list. The arithmetic that decides the outcome takes one afternoon, and almost nobody does it first.

That arithmetic is cost per account per year against expected gross profit per account per year. The software, the direct mail, the executive dinners, all of it sits downstream of whether those two numbers are in the right order. Get them wrong and you will spend a year producing engagement charts that no revenue leader believes.

## What ABM actually costs per account per year

Somewhere between $300 and $40,000, and the tier you choose is the entire story. A 1:1 program is a research and content operation run for one company. A 1:many program is list targeting with a personalised headline on the landing page. Both get called ABM, which is how budgets get approved and then disappointed.

The model below is loaded cost. It includes people time at fully loaded salary, media, content production, gifting and field events, and the share of data and software licences each tier consumes. Vendor pricing pages quote the media line only, and media is about a third of the real number.

Read the middle column as an annual commitment rather than a launch budget. A 1:1 tier of 20 accounts runs $300,000 to $800,000 a year, which in practice is one senior marketer, a researcher's worth of agency time, an events line and a content budget. If that sounds heavy for 20 logos, that is the correct reaction. The tier exists for accounts whose lifetime value clears seven figures.

Costing ABM as media spend. Media is roughly a third of it. Research, content, people time, gifting and events make up the rest, and none of them appear on the platform invoice you used to build the business case.

## Does your ACV clear the floor

Work out expected gross profit per account per year, then compare it with cost per account per year. Expected gross profit is lifetime gross profit per customer multiplied by the annual account win rate you can defend on a named list.

Take a $60,000 ACV product at 80% gross margin with three year average retention. Lifetime gross profit is $144,000. If a well chosen named account converts to a closed deal at 10% in a year, expected gross profit per account per year is $14,400. My working rule: annual cost per account should sit under a third of that figure, because sales compensation, discounting and the accounts that stall will absorb the rest.

| ACV | Lifetime gross profit | Defensible annual win rate | Expected GP per account per year | Ceiling on cost per account | Tier that fits |
|---|---|---|---|---|---|
| $10,000 | $24,000 | 12% | $2,880 | $960 | None. Run demand capture instead |
| $25,000 | $60,000 | 10% | $6,000 | $2,000 | 1:many only |
| $60,000 | $144,000 | 10% | $14,400 | $4,800 | 1:many plus a 1:few core |
| $120,000 | $288,000 | 9% | $25,920 | $8,600 | 1:few with a small 1:1 tier |
| $250,000 | $600,000 | 8% | $48,000 | $16,000 | 1:1 for the top 15 accounts |

The pattern is blunt. Below roughly $25,000 in ACV only the cheapest tier survives the maths, and at that point you are running outbound with better targeting and a nicer name. The [demand generation playbooks by ACV band](/playbooks/demand-generation-by-acv-band/) set out what to run instead at $8,000 or $15,000 deal sizes, and none of it involves per account content.

**17%** Share of the B2B buying journey buyers spend with any vendor sales rep, split across every vendor they consider

That Gartner figure is the strongest argument for funding account coverage from marketing. If sales gets 17% of the journey and has to share it, the other 83% is either your content reaching the account or a competitor's.

## Building the list: fit, intent, and who you already know

Three inputs, in that order of reliability. Fit comes from your closed won data, not from a persona document written in a workshop. Pull the last 40 wins and the last 40 losses and find the attributes that genuinely separate them: headcount band, funding stage, a tech stack signal, the number of people holding the job title your product serves, region.

Intent is the second layer and the most oversold. Third party intent from 6sense, Demandbase or a Bombora feed tells you that someone inside an IP range read about a topic. That is weaker than it sounds. First party signals are cheaper and sharper: three visits to the pricing page from one domain in a week, a docs session, a security page view, a second person from the same company opening the same email thread.

Relationship data is the input most teams skip and it beats both. Former customers who changed jobs, second degree connections held by your executives, portfolio overlap with your investors, users from an account that churned two years ago under leadership that has since been replaced. Score it and weight it heavily.

Then cut. Negative criteria typically remove a third of a raw fit list:

- Signed a competitor in the last 12 months on a multi year contract
- Nobody in the account holds the job title your product serves
- Headcount below the floor where your pricing has a plan
- A region you cannot support for data residency, language or billing
- A churned account where the original reason for churn is still unfixed
- An open partner sourced deal that your program would duplicate and confuse

If negatives remove less than a fifth of the list, your fit criteria were too loose. [Building and tiering the target account list](/playbooks/abm-tiering-target-account-lists/) is a separate job with its own capacity maths, and it is the part of ABM most worth doing slowly.

## The orchestration calendar and who does what

ABM runs in waves, not campaigns. A wave is six weeks against one cluster, and everything in it is scheduled before week one so nobody improvises. The structure below is what I use for a 1:few cluster of roughly 30 accounts.

**A six week ABM wave for one cluster**

Split ownership so there is no ambiguity on a Monday morning. Marketing owns the list, the context pages, the ads, the cluster content and the events. Sales owns acceptance, sequences, calls and the multi threading push. Whoever owns the weekly review owns the argument, so make it a shared meeting with a written agenda. A [sales and marketing SLA](/templates/sales-marketing-sla-template/) is worth an hour here because it forces the acceptance commitment into a document rather than a hallway agreement.

Paid delivery mostly means LinkedIn, and the mechanics deserve their own attention. Matched audiences will drop 20% to 40% of a company list depending on how you upload, and company page follows skew reach toward people who already know you. The [LinkedIn demand generation playbook](/playbooks/linkedin-demand-generation-saas/) covers frequency capping and creative rotation for lists this small. Events are the other half: a 12 person dinner with four target accounts beats a 300 person webinar for this tier, which is why [field marketing](/playbooks/saas-field-marketing-events/) usually ends up as the largest line in a 1:1 budget.

## The stack you need, and what a spreadsheet can fake

Every category below can be bought or improvised. The improvised version is genuinely fine under about 150 accounts, and I would spend the difference on a researcher.

The tell that you have outgrown the spreadsheet is the weekly dedupe review taking more than an hour, or two reps calling the same account in the same week because the matching rule missed a subsidiary domain. Buy tooling when that happens, not before. Clay and Attio have both made the cheap end of this considerably more capable in the last two years, which pushed the buy decision later for most teams under $20M ARR.

## What to measure, and when anything shows up

Account engagement first, meetings second, pipeline third, revenue last. Reporting pipeline in month two is how programs get killed in month five, because you set a clock you cannot beat.

| Months | What genuinely moves | Report this | Do not report this yet |
|---|---|---|---|
| 1 to 3 | Ad reach, rep acceptance rate, context pages delivered | Match rate, acceptance rate, engaged contacts per account | Pipeline, CAC, ROI |
| 4 to 6 | First meetings, second contacts inside accounts | Meeting rate per 100 accounts, multi threading depth | Closed won |
| 7 to 9 | Qualified opportunities from named accounts | Pipeline coverage of the list, opportunity rate versus control | Payback period |
| 10 to 14 | Closed revenue, win rate difference | Win rate on target versus control, ACV difference, cycle length | A clean multi touch attribution number |

Three definitions worth agreeing in writing before launch. Account engagement is the count of distinct contacts in an account with a tracked action in the last 30 days, and three is the threshold that predicts a meeting. Pipeline coverage of the named list is open pipeline from those accounts divided by the bookings target assigned to them, and three times is the planning minimum. Meeting rate is meetings per 100 named accounts per quarter, which for a decent 1:few tier lands between 8 and 18.

Attribution will not be clean, and pretending otherwise costs you credibility with the CFO. Hold out 100 fit matched accounts you deliberately do not target and compare opportunity creation after three quarters. That comparison is the only honest causal read available, and it survives scrutiny in a way that a weighted model does not. [Attribution modelling](/guides/saas-attribution-models/) is still worth setting up for reporting hygiene, but it will not settle the argument about whether ABM caused the deal.

## Where these programs break

Sales never committed. This is the first and largest failure mode, and it is usually visible in week zero when the rep accepts every account without reading the list. Genuine acceptance sounds like an argument about six specific companies.

Capacity got ignored. An account executive can work 25 to 40 named accounts properly alongside their inbound. A team of six therefore supports 150 to 240 accounts in tiers one and two, and a list of 600 means 360 accounts get a marketing touch and no human follow up. Those accounts learn your name and buy from someone who called.

Reporting stayed on lead volume. If the board still reads [cost per lead](/calculators/cost-per-lead/) as the headline number, an ABM program will look like a catastrophe for three quarters while it produces fewer, better conversations. Change the reporting before you change the program, not after.

The honest cost is not the budget line. It is the demand capture you stop funding, and the uncomfortable fact that a share of the accounts you win were going to buy anyway. Every ABM case study quietly includes those deals. Assume 20% to 30% of your wins in year one would have arrived without the program, and your numbers will hold up better under questioning than the vendor deck that inspired them.

## What ABM does not replace

It does not replace inbound, and the teams that treat it as a substitute end up with a pipeline that stops the moment the program pauses. Named account programs cover demand you can see. Everything else in the market still finds you through search, community, referrals and content. The [inbound versus ABM comparison](/comparisons/inbound-vs-abm/) works through where each motion actually wins, and the honest answer for most SaaS companies between $5M and $50M ARR is that they run both, with ABM taking 20% to 35% of the demand budget. Start from the wider [SaaS demand generation](/saas-demand-generation/) picture before you carve that share out.

## Do this in the next two weeks

**Before you spend anything**

Write the plan on one page, with the account count, the cost per account, the expected gross profit per account and the control group named on it. If any of those four numbers is missing, the program is not ready. The [demand generation plan template](/templates/demand-generation-plan-template/) has the structure if you would rather not start from a blank page.

Then go and get the acceptance signatures. ABM is a sales program that marketing funds, and if sales will not commit named reps to a named list, the right move is to keep the money and spend it where it compounds without them.
{/* intent-consolidation-2026-09 */}
## Check whether coordinated account work is justified

ABM is a coordinated operating choice, not a universal contract-value threshold. Evaluate the expected commercial value, number of suitable accounts, buying complexity and cost of the work. Any illustrative price or effort range needs to be replaced with the team's actual scope before it is used in a budget decision.

### Start with one account-level question

Identify the buying situation and the people whose decisions affect it. A list of job titles is not yet a buying-process map. Record which requirement each participant owns and which evidence would help them evaluate the proposal. Keep a distinction between known information and inferred intent.

The account plan should identify a useful offer, an appropriate contact route, the sales owner and the implementation specialist required. More personalization is not automatically better. A small number of accurate, relevant actions can be preferable to a large sequence based on weak assumptions.

### Use tiers to allocate work, not to create prestige

A high-touch account tier should justify the extra research, creative and coordination it consumes. A lower-touch program still needs a relevant audience and honest offer. Review whether the team can perform the promised work before assigning many accounts to a custom-work tier.

| Operating decision | What to verify |
| --- | --- |
| Account selection | Supported workflow, urgency and readiness |
| Stakeholder coverage | Relevant roles and unresolved requirements |
| Offer | A useful next task for the account |
| Coordination | Named marketing, sales and specialist owners |
| Economics | Complete cost and a realistic observation window |
| Stop rule | Evidence that should pause or change the account plan |

Use the [account fit and intent diagnostic](/guides/account-fit-and-intent-are-confused/) and [committee-map diagnostic](/guides/buying-committee-map-is-outdated/) to keep engagement scores from replacing an actual understanding of the account.

## Frequently asked questions

### How much does ABM cost per account for a SaaS company?

Loaded cost is roughly $15,000 to $40,000 per account per year at 1:1, $3,000 to $8,000 at 1:few, and $300 to $900 at 1:many. Those figures include people time, research, content, events and the share of data licences each tier consumes. Media alone is about a third of the total.

### What is the minimum ACV for ABM to make sense in SaaS?

About $25,000 for anything beyond the cheapest tier. At $25,000 ACV with 80% gross margin and three year retention, lifetime gross profit is $60,000, so a 10% annual account win rate produces $6,000 of expected gross profit per account per year. Once sales cost is added, only a 1:many program stays profitable.

### How long does a SaaS ABM program take to produce pipeline?

Engagement lift appears in weeks one to twelve, first meetings with named accounts in months four to six, qualified opportunities in months seven to nine, and closed revenue between months ten and fourteen on a normal mid market sales cycle. Reporting pipeline before month six sets an expectation you cannot meet.

### What is the difference between 1:1, 1:few and 1:many ABM?

1:1 is custom research and content for 10 to 25 named companies. 1:few groups 60 to 150 accounts into clusters that share an industry or use case and gives each cluster its own content set. 1:many targets 500 to 2,000 accounts with list based ads, personalised landing pages and intent alerts routed to reps.

### Do you need 6sense or Demandbase to run ABM?

Not under about 150 accounts. First party signals plus G2 buyer intent cover most of what you need for a fraction of the cost. Intent platforms earn their $40,000 to $120,000 a year once the list passes 500 accounts and manual lead to account matching starts eating hours every week.

### How many accounts should be in a SaaS ABM program?

Cap the list at rep capacity, not ambition. An account executive can genuinely work about 25 to 40 named accounts alongside inbound, so a team of six supports 150 to 240 tier one and tier two accounts. Anything above that belongs in the 1:many tier where nobody pretends there is human follow up.

### How do you measure ABM when attribution is unreliable?

Report account engagement, meeting rate per account, multi threading depth and pipeline coverage of the named list in the first two quarters. Then compare opportunity creation rates between targeted accounts and a held out control group of similar accounts. That comparison survives a board conversation in a way that a multi touch model does not.