What it is
Portfolio Insights is an AI capability inside the Strategic Planning Workspace (SPW) that reviews the planning items in a portfolio plan and flags the ones at risk of schedule problems. It was made first available in May 2026 release of Now Assist for SPM and incrementally being enhanced in follow-up releases.
It runs across the planning module and surfaces its findings in the AI Insights panel, a floating side panel that a portfolio manager can open from any tab of the plan. The panel groups its findings into insight cards. Each card answers one question about the portfolio, reports how many items are affected, breaks them down by severity, and explains the likely root causes and recommended next steps. This guide covers four cards that are available today:
- Planning items delayed beyond planned end date - items that should have finished but have not.
- Planning items with delayed starts - items that started later than planned.
- Planned vs. approved date misalignment - items whose planned dates have drifted from their approved baseline.
- Projects at risk - prioritized projects showing early risk signals that could lead to a future delay.
Why it matters
For a portfolio of 50 to 75 planning items, finding schedule problems by hand is slow and easy to get wrong. A portfolio manager has to scan list views, compare planned dates against actuals, open each delayed item, and check resources, dependencies, financials, and scope before reaching a root cause. The investigation is repetitive and the conclusions are hard to defend.
Portfolio Insights does that screening automatically and presents the result in a form the manager can act on. The value for the business is earlier intervention: problems surface while there is still time to correct them, the analysis is consistent from one review to the next, and the reasoning behind each severity rating is explainable rather than a matter of opinion.
A user can generate the insights on demand using the ‘Portfolio Insights’ option in the ‘Planning’ page across all tabs in the Strategic or Portfolio Planning Workspace.
How it works - The four insight cards
Planning items delayed beyond planned end date. Looks at prioritized planning items where the actual end date is empty and the current date is past the planned end date. Completed items are out of scope even if they finished late. Root causes and recommendations are generated from the top 15 delayed items, ranked by severity score.
Planning items with delayed starts. Looks at prioritized items whose actual start date is later than the planned start date within a recent window. Only items whose planned start date falls within the last 30 days are included.
Planned vs. approved date misalignment. Flags prioritized items whose planned start or end date no longer matches the approved dates, which puts the item at risk of starting or finishing off the approved schedule. The card breaks the gap down by start, end, or both, and offers two resolution paths: realign the planned dates to the approved baseline, or update the approved dates to the new reality and seek re-approval.
Projects at risk. Looks at prioritized projects that have active risk signals which could cause a future delay, across task delays, milestone slippage, resource overallocation and conflicts, and financial pressure. An item's overall severity is the highest level triggered across any of these categories.
Each card has a consistent layout. It shows the number of impacted items, a breakdown into Critical, Medium, and Low, an Impact section with monetary benefit and dependencies at risk, the top root causes with the count of items behind each, and recommended actions. Selecting the count of items takes the user to the planning items in the workspace to act on them.
Configuring severity scoring
Severity thresholds and scoring factors come preconfigured. Change them only if your organization's risk criteria differ from the defaults. Configuration does not require a code or prompt change, so it is supported rather than a customization. The required role is sn_align_core.apw_admin.
How an item is scored. Each planning item receives an overall risk score from a set of weighted scoring factors. That score is then compared against the severity thresholds to assign Critical, Medium, or Low. The color coding is fixed: red for critical, yellow for medium, green for low.
Where the settings live
Configuration is held in the Insight topic configuration table, in records whose Type is Portfolio. Navigate to sn_spm_gen_ai_insight_topic.list, filter the Type column to Portfolio, and open the record for the category you want to change. Each category has its own record with its own thresholds and factors, so edit the record for the insight you intend to tune. As of June 2026 store release, the configurable categories are Delayed planning items, Date misalignment, and Delayed start, Risk of Delay planning items. The steps are the same for each.
Severity thresholds
The thresholds and factors shown below are the defaults for one insight, Planning items delayed beyond planned end date. Each insight has its own record with thresholds and scoring factors relevant to that insight, so the factor names and values differ between them. Use these as an illustration of how scoring is structured, not as values to copy across every insight.
In the severityThresholds section of the Default topic config field, set the minimum score for each level. An item is rated at the highest level whose minimum score it meets.
|
Severity level |
Meaning |
Default minimum score |
|
Critical |
Risk score meets or exceeds the critical threshold. |
10 |
|
Medium |
Risk score falls between the medium and critical thresholds. |
5 |
|
Low |
Risk score falls below the medium threshold. |
1 |
Scoring factors
In the scoringFactors section, set the weight of each factor that feeds the overall score. Each factor can be turned off, and most are capped so that a single factor cannot dominate the score. The factors below apply to the delayed-beyond-end-date insight; other insights expose a different set of factors suited to what they detect.
|
Scoring factor |
What it contributes |
Default |
|
Scheduled delay |
Points for each day the item is delayed, up to a maximum. |
0.5/day, max 10 |
|
Priority |
Points based on the priority of the item. |
High 3, Med 1, Low 0 |
|
Dependency impact |
Points for each downstream dependent item, up to a maximum. |
1/dep, max 6 |
|
Financial risk |
Extra points when financial exposure is above a benefit threshold. |
3 if > 100000 |
Worked example. A high-priority item that is 12 days late with 2 downstream dependents and benefit above the threshold scores 6 from delay (12 days at 0.5, capped at 10), 3 from priority, 2 from dependencies, and 3 from financial risk, for a total of 14. That meets the critical minimum of 10, so the item is rated Critical.
Tuning the result. To make a factor matter less without removing it, lower its weight rather than disabling it. An organization that does not want financials to drive criticality can reduce the financial points or set enabled to false. Save the record, and the new thresholds and weights apply the next time insights are generated for the plan.
Verifying the configuration
Open a portfolio plan that contains items with known delays and confirm that the severity ratings match the thresholds you set. If the mix looks wrong, adjust the weights or thresholds and regenerate. The insights are only as good as the dates, dependencies, allocations, and cost plans on the underlying records, so treat data quality as the first thing to check.
Docs page references:
View insights for a portfolio plan
Update severity scoring configuration for portfolio insights
