Next Best Action Strategy
A Next Best Action Strategy is the graph of nodes built in Strategy Builder that decides which recommendations qualify for the current record and in what order.
Definition
A Next Best Action Strategy is the graph of nodes built in Strategy Builder that decides which recommendations qualify for the current record and in what order. The strategy lives as a metadata record in the org, is versioned, and is referenced by one or more Einstein Next Best Action components on Lightning pages. When the page loads, the platform runs the strategy against the current record context, evaluates each node, and returns the top recommendations to the component.
The strategy is where business logic lives in a Next Best Action deployment. Recommendations describe what could be offered. The strategy decides which ones to offer right now, on this record, to this user, based on rules, predictive scores, and external inputs. A strategy can be as simple as load-and-prioritize (show every recommendation, top three by static priority) or as complex as a multi-branch graph mixing Einstein predictions, related-record lookups, and CRM Analytics scores. Strategy complexity is a design choice tied to the team's appetite for maintenance.
The nodes, the canvas, and what good strategies look like
The five core node types
Strategy Builder offers a small set of node types that compose into any strategy. Load nodes bring recommendations or other records into the stream. Branch nodes filter the stream by a condition. Map nodes set or transform per-recommendation values. Filter nodes drop records that fail a predicate. Prioritize nodes sort by a numeric score and return the top N. Predict nodes call a CRM Analytics model and append the score. Recommendation Limit nodes enforce a hard cap on how many recommendations survive to the output. Every strategy is a graph of these few types.
The record context and how it flows through
The strategy runs in the context of a target record (a contact, an opportunity, a case). That record's fields are available as merge variables at every node. Branches can compare {!Target.AnnualRevenue} or {!Target.Status} to literals. Map nodes can set per-recommendation values based on the target. The context is read-only from the strategy's perspective: nothing the strategy does changes the target record. Side effects happen when the user accepts a recommendation and the assigned flow runs.
Branching on predictive scores
The most common reason to add complexity to a strategy is plugging in a predictive score. The simplest path is to pre-compute the score on the record (Einstein Lead Score, an Einstein Prediction Builder output) and branch on the field. The more flexible path is to add a Predict node that calls a CRM Analytics or Einstein Discovery model inline, takes the returned score, and branches on it. The inline approach is more dynamic; the field approach is faster at page-load time. Mixing both in one strategy is common.
Prioritization, ties, and why top-N matters
The Prioritize node sorts surviving recommendations by a numeric expression. The expression can use recommendation fields (Priority__c), target fields (AnnualRevenue), or computed values from upstream nodes. Ties are broken by the order recommendations entered the strategy, which is sometimes surprising. The Recommendation Limit node enforces the top-N cap. Three is the practical maximum for a single component; users start ignoring the fourth and beyond. Strategies that return more than three signals over-engineering rather than thoughtful selection.
Versioning, activation, and rollback
Strategies are versioned. A new version is a draft until activated. Activating replaces the previous version atomically. Rolling back is a matter of activating an earlier version. This makes strategy iteration safe and reversible. The cost is discipline: changes need a deliberate version bump, not in-place edits to the active version. Strategies in managed packages versioning works the standard way; subscriber orgs receive the packaged strategy and cannot edit it without upgrade.
Testing strategies before they hit production
Strategy Builder has a Test tab where any strategy can be evaluated against a sample record ID. The test shows which recommendations survived, which were filtered, and which prioritized to the top. The output is the same the live component would produce. Running the test against a representative set of records (high-value account, low-value account, edge cases) catches branch logic mistakes before users see them. Production debugging is harder because no logs persist by default; test in the canvas first.
Strategies that age badly
Strategies degrade when the underlying assumptions change. A branch on Industry equals Healthcare ignores the new Healthcare-Adjacent industries the company expanded into. A score threshold of 70 was right when the average score was 60 and is wrong when the average is 80. Schedule a quarterly review of every active strategy, walking each branch against current record distributions. Stale strategies surface stale recommendations, which trains users to ignore the component.
How to design a Next Best Action strategy that holds up over time
Designing a strategy is less about Strategy Builder mechanics and more about deciding what the recommendations should mean. The steps below cover the design and the build.
- Define the audience and surface
Pick the user role and the page where the strategy will run (sales console opportunity page, service console case page, community portal). Different audiences need different strategies; do not try to serve all in one.
- Inventory the relevant recommendations
List the Recommendation records that could matter for this audience. Group them into thematic buckets (cross-sell, retention, escalation, education). The bucketing helps later when designing branches.
- Draft the branch logic on paper first
Before opening Strategy Builder, sketch the decision tree. When does each bucket qualify? What scores or fields drive the decision? Reviewing the sketch with the business owner avoids rework once the canvas is open.
- Build, version, test on the canvas
Translate the sketch into Strategy Builder nodes. Save as a new version. Use the Test tab against five to ten representative record IDs. Iterate until the output matches the intent.
- Activate and build the reaction report
Activate the strategy version. Build a Recommendation Reaction report grouped by recommendation. Review acceptance rate weekly and refine.
Brings recommendations or related records into the strategy. Most strategies start with a Load Recommendations node.
Routes the stream based on a condition. The if-then-else of Strategy Builder.
Sets or transforms per-recommendation values. Useful for computing dynamic acceptance labels or priority scores.
Calls a CRM Analytics or Einstein Discovery model and appends the score. Used for inline predictive branching.
Caps the output at the top N. Three is the practical maximum for a single component.
- Strategies do not log execution by default. Use the Test tab on the canvas; do not expect debug logs in production.
- Ties on the Prioritize node fall back to entry order, which is often non-obvious. Make priority expressions tie-breaking when it matters.
- Inline Predict nodes add latency at page load. Pre-compute the score on the record when the score is stable; reserve inline predicts for context-sensitive scoring.
- Strategies need version bumps for non-trivial changes. Editing the active version directly is technically possible and operationally risky.
- Stale branches surface stale recommendations. Schedule a quarterly walk-through of every active strategy.
Trust & references
Cross-checked against the following references.
- Strategy BuilderSalesforce Help
- Einstein Next Best ActionSalesforce Help
Straight from the source - Salesforce's reference material on Next Best Action Strategy.
- Strategy Builder referenceSalesforce Help
- Einstein Next Best Action overviewSalesforce Help
About the Author
Dipojjal Chakrabarti is a B2C Solution Architect with 29 Salesforce certifications and over 13 years in the Salesforce ecosystem. He runs salesforcedictionary.com to help admins, developers, architects, and cert/interview candidates sharpen their fundamentals. More about Dipojjal.
Test your knowledge
Q1. What is a Next Best Action Strategy?
Q2. What does Strategy Builder combine?
Q3. What's a common mistake?
Discussion
Loading discussion…