Marketing
Account-Based Marketing: Build a Focused Account Strategy
Build a focused account-based marketing campaign with a small portfolio, coordinated owners, permission boundaries and useful review measures.

What you will learn
- Distinguish an account-portfolio campaign from a single-account plan.
- Allocate a bounded campaign across selected accounts using existing fit evidence.
- Coordinate useful resources and follow-up while separating engagement from buying intent.
Account-based marketing coordinates marketing and sales around a deliberately selected group of organizations and their buying situations. Start with existing fit evidence, choose the accounts your team can responsibly support and give each campaign a useful customer question. ABM is not a generic blast with company names inserted into the greeting.
Understand the account-based choice
Salesforce's ABM teaching describes a strategy that coordinates teams around selected accounts, not a standalone tool or feature. You can use that organizing idea without assuming that a particular platform supplies account research, consent or campaign results automatically.
A broad acquisition program may begin with an audience and a reusable resource. An account-focused campaign starts with named organizations and asks what relevant work would help their particular evaluation. The choice should follow the economics and complexity of the buying situation, not the appeal of a more sophisticated label.
Keep the artifacts separate. An account plan connects one customer's objectives and initiatives over time. A relationship map records people, roles and evidence. This lesson owns the portfolio decision and coordinated campaign, using those records as inputs rather than recreating them.
Select a portfolio your team can support
Start with an approved fit hypothesis and permitted evidence. Record why each candidate belongs in the current review, the specific problem you believe is relevant and what is still unknown. A recognizable brand or a possible relationship path is not proof of an active need.
For an original fictional exercise, a document-review service considers five organizations. Cedar, Birch and Reed have relevant problem evidence and a suitable next learning step. Hill requires a capability the service does not supply. Slate is a familiar name without current problem evidence. The team selects the first three, excludes Hill for this offer and defers Slate pending evidence.
Selection does not authorize contact. Check the actual relationship, channel rules, recipient preferences and purpose before outreach. A teammate knowing someone at an account is not permission to use their name or make an introduction. Keep missing access visible instead of turning it into a speculative automated sequence.
Build a coordinated campaign board
The team has twelve planning hours for this fictional experiment, including preparation and review. The amounts below are invented capacity allocations, not benchmarks or a prediction of required effort.
A three-account campaign within a twelve-hour planning budget
| Account and useful question | Resource and coordinated owners | Permission boundary and next evidence | Planned effort |
|---|---|---|---|
| Cedar: can two teams review the same current document? | Marketing prepares an approved workflow example; the account owner checks its relevance with the existing contact. | Ask whether a short review is useful. A resource open is not an accepted evaluation. | 4 hours |
| Birch: which customer input is needed before review starts? | Marketing prepares a clear input checklist; sales and delivery verify its limits together. | Use only the requested follow-up route. Evidence is a corrected prerequisite question, not an inferred budget. | 3 hours |
| Reed: does the current process need an external service at all? | The account owner and content owner prepare a neutral comparison with the existing method. | Respect a sufficient status quo or refusal. Evidence may justify stopping, not another campaign. | 5 hours |
Four plus three plus five equals twelve hours. A new request requires narrowing, deferring or increasing the budget through the responsible owner; it is not free because the account is attractive. Hill and Slate are not silently added to the campaign's denominator or credited with a result.
For each row, add source/date, account owner, resource version, dependency, proposed action, approved channel and review trigger. Record which decisions marketing and sales accept. One person can coordinate follow-through without having authority to approve another team's claims or a customer's participation.
Measure the intended next step
Choose measures that match the question: a relevant reply, a corrected requirement, an accepted evaluation or a decision not to proceed. Keep delivery, opening, response, accepted opportunity and customer outcomes separate. A document visit may be anonymous, repeated or unrelated to a buying decision.
Use the campaign naming lesson when an approved link needs consistent attribution. Tagged traffic can support a record of where a visit came from; it does not establish that the campaign caused a sale. Do not add channel counts together when they may describe the same people or account.
Compare actual effort with the planned allocation and record missing coverage. If one account provides a useful answer and another remains silent, preserve both outcomes. A small selected portfolio cannot establish market-wide demand, and silence does not prove that an account has no problem.
Copyable AI prompt
Try this with Aurora AI in Finta, Codex, or Claude. Use only account information and source material you are authorized to share.
Using [approved account-fit criteria], [authorized account evidence], [relationship and channel restrictions], [approved resources] and [planning capacity], draft an account-portfolio campaign board. Separate selected, deferred and excluded accounts with reasons. For each selected account state one useful question, resource/version, marketing and sales owners, permission boundary, effort, meaningful next evidence and stop/review rule. Reconcile the total effort. Mark unknown need, authority and attribution. Do not contact people, infer consent, request introductions, launch campaigns or change records. Draft for human review.
Review the portfolio and its stop rules
At the agreed checkpoint, decide whether to continue, narrow, pause or stop each experiment. Recheck the evidence when a requirement, relationship or offer changes. An account leaving the portfolio is a decision to record, not an embarrassment to hide by renaming the campaign.
Use network-led prospecting for the separate question of a credible, permission-sensitive relationship path. Keep the campaign focused on a few justified next steps, with the customer free to correct the premise or decline.
Sources and method. By Finta Editorial Team. Sources checked November 15, 2026. The examples, worksheets and practice cases are original fictional teaching materials, not customer findings or product-test results.