Smart Cycle Discounts Campaign Intelligence: What the Numbers Actually Tell You
Plugin Guide · Smart Cycle Discounts
Smart Cycle Discounts Campaign Intelligence: What the Numbers Actually Tell You
Your dashboard shows a verdict, a confidence band, a reason, and a next action. Here is what each of those actually means — and how to use them to run better promotions.
What Campaign Intelligence is (and what it is not)
Campaign Intelligence is the analysis layer built into Smart Cycle Discounts that evaluates each of your discount campaigns and produces a single, actionable verdict: is this campaign safe to run right now, or does something need your attention?
It is not a reporting dashboard in the traditional sense. It does not give you a table of revenue and impressions to scroll through. What it gives you is a decision — a clear answer to the question “what should I do with this campaign right now?” — backed by the data that drove that answer.
The system runs continuously in the background. Each time you load the dashboard, a campaign’s overview panel, the campaign list, or the campaign wizard, Campaign Intelligence re-evaluates the campaign against your live store data: active conflicts, current stock levels, schedule integrity, discount configuration quality, and (in Pro) actual order economics. The result is cached for a short window — between 90 seconds for the wizard and 3 minutes for the overview panel — so you are always looking at a reasonably fresh picture without every page load triggering a full recalculation.
Campaign Intelligence vs. the analytics dashboard
Campaign Intelligence evaluates campaign quality — conflicts, stock, configuration, schedule, economics — and tells you what to do. The analytics dashboard (available with full depth in Pro) shows campaign performance data: revenue, orders, conversions. They serve different purposes. This post covers Campaign Intelligence specifically.
The five verdict states explained
Campaign Intelligence collapses its entire analysis into one of five verdict states. Each state has a specific meaning and a specific appropriate response.
Blocked
Something fundamental is broken and the campaign cannot work until it is fixed. Examples: no products are selected, no discount rule is configured, or all products in the campaign are out of stock. The system will not let this pass silently — you will see exactly what is blocking it and a direct link to fix it.
At Risk
The campaign is running or can run, but active damage is happening. This state requires prompt attention. Typical triggers: the discount is consuming your entire margin on the affected products, a high-priority conflict is overriding your campaign on a significant share of products, or stock is depleting faster than expected. “At Risk” uses stronger impact language than “Needs Attention” because the consequences are already happening.
Needs Attention
A genuine campaign-quality issue exists, but it is not yet causing active harm. Readiness is in the fair range (70–89), a conflict exists at low severity, or a degraded warning is active. This is the most common non-green state and often indicates that a campaign is technically functional but not as effective as it could be.
On Track
The campaign is healthy. All checks have passed, scores are within safe ranges, and there are no active issues. No action is needed. If you are in this state, Campaign Intelligence’s job is essentially done — it is confirming that nothing is wrong.
Ready
All pre-launch checks pass. The campaign is configured correctly, there are no detected conflicts, stock is adequate, and the schedule is sound. This state only appears before activation. Once a campaign is live with low confidence data, the system may show “On Track” rather than “Ready” to avoid overclaiming certainty.
The verdict is not the score
The verdict state (Blocked / At Risk / Needs Attention / On Track / Ready) is the primary output — not the numeric scores you may also see. The verdict is what tells you what to do. The scores explain why the system reached that verdict. Read the verdict first; use the scores to understand it.
How the scores are calculated
Campaign Intelligence computes four composite scores, each on a 0–100 scale. Understanding what goes into each score helps you interpret why a campaign ends up in a particular verdict state.
The six health dimensions
All four scores ultimately derive from the same six underlying health dimensions. Each dimension maps to a specific aspect of your campaign’s quality, and each is assigned a weight in the final health score calculation:
| Dimension | Weight | What it evaluates |
|---|---|---|
| Conflicts | 30% | Overlap with other active or scheduled campaigns on the same products. The highest-weighted dimension because conflicts directly undermine which discount a customer actually receives. |
| Schedule | 20% | Schedule completeness, timing risks, recurring configuration correctness, and whether the campaign ends before it has meaningfully run. |
| Discount | 15% | Discount configuration integrity — whether the discount type is correctly configured, whether the depth is extreme or too weak to convert, and type-specific issues (BOGO target missing, bundle minimum not met, tiered steps illogical). |
| Stock | 15% | Inventory exposure on discounted products. Running a promotion while products are low-stock or out-of-stock wastes the campaign and frustrates customers who find nothing to buy. |
| Coverage | 10% | Product selection quality — whether the product set is valid, non-empty, not accidentally too broad or too narrow, and whether deleted products have been removed from the scope. |
| Products | 10% | General product readiness: whether selected products are published, purchasable, and not filtered out by other active rules. |
The four scores
From those six dimensions, the system derives four distinct scores:
- Health score — the weighted average of the six dimensions above. A score of 100 means no issues on any dimension. The health score directly drives the “campaign health” indicator you see in the operational signals.
- Risk score — mathematically the inverse of the health score (100 minus health score). It represents how much downside is present. A risk score of 30 means the campaign has a health score of 70.
- Readiness score — derived from the health dimensions but adjusted to reflect how prepared the campaign is to activate, not just how few issues it has. A campaign can have a good health score but a lower readiness score if, for example, its schedule is incomplete or it has not been reviewed.
- Composite score — a blend of opportunity (45%), readiness (35%), and inverted risk (20%), adjusted by store-level signals like competition pressure and whether the campaign is already active. This is the single “intelligence score” number you may see in the UI on some surfaces.
What the dashboard gauge actually shows
The gauge on the main dashboard does not display the composite intelligence score for any single campaign. It shows the store-wide health score — a blended aggregate across all active and scheduled campaigns, where active campaigns are weighted twice as heavily as scheduled ones. A single unhealthy active campaign pulls the gauge down noticeably. A blocking issue hard-caps the store-wide score at 40 regardless of what other campaigns are doing.
Confidence: why it matters more than the score
The confidence band (Low / Medium / High) is arguably the most underappreciated piece of output Campaign Intelligence produces. A verdict at High confidence means the system had good data and its recommendation is solid. A verdict at Low confidence means the system reached a conclusion, but with limited evidence — and that is important context for how much weight you should give it.
Confidence is computed from four weighted factors:
- Signal completeness (24%): Does the system have all the data it needs? Campaign setup, health data, store context, and diagnostics all contribute. A newly created campaign with sparse configuration will score lower here.
- Data freshness (22%): How current is the analysis? This depends on the “analysis mode” — exact analysis (where the product set is fully known) scores 1.0; post-activation exact analysis scores 0.74; estimated analysis (where the product set is not yet compiled) scores 0.58. Stale data applies an additional penalty.
- Cross-signal consistency (18%): Do the readiness, risk, and health scores tell a consistent story? A campaign where readiness is high but risk is also high contains a contradiction that lowers confidence.
- Evidence quality (28%): The largest single factor. Diagnostic evidence contributes 40% of this component; actual economic data from attributed orders contributes the remaining 60% when available. No attributed orders means confidence is capped regardless of other factors.
The final confidence value is clamped between 0.10 and 0.98 — the system never claims perfect certainty, and it never falls below a minimum floor. Confidence bands are: High (0.80 and above), Medium (0.55–0.79), Low (below 0.55).
Several caps apply automatically. A campaign in “estimated” analysis mode with no attributed orders is capped at the low band (0.54) even if all other signals look healthy. This prevents a brand-new campaign — where the system is still guessing about product scope — from appearing confidently “Ready” when it does not yet have real data behind that verdict.
What “Low confidence” actually means in practice
Imagine you have just activated a new campaign with a randomly selected product pool. Campaign Intelligence will classify this as “estimated” analysis mode — it cannot know the exact product set until after the first activation compiles it. The verdict might be “On Track” at Low confidence. That is not a contradiction: the system is telling you “nothing obviously wrong, but we do not have enough data yet to be certain.” The right response is to let the campaign run a few days, come back, and check whether confidence has moved to Medium once attributed orders accumulate. If the verdict stays Low confidence and the readiness is also low, that is the real signal to investigate.
Reason codes and what triggers them
Every verdict includes a reason code — a specific, named cause that explains why the system reached its decision. The reason code is not the verdict itself; it is the dominant signal that pushed the verdict in that direction. There are currently over 80 reason codes across six families.
The six reason families
| Family | What it covers | Common triggers |
|---|---|---|
| Conflicts | Campaign overlap on shared products | Two campaigns at the same priority targeting the same products; a higher-priority campaign overriding yours; a scheduled campaign that will conflict when it goes live |
| Stock | Inventory problems on discounted products | Low-stock products in the campaign, products already out of stock, high depletion risk from the discount driving demand faster than stock replenishes |
| Coverage | Product selection issues | Empty product set, deleted products still referenced, a category filter that now returns zero products, a product selection so broad it applies to the entire catalog accidentally |
| Discount | Discount configuration problems | No discount rule configured, extreme discount depth, BOGO missing its “get” product, bundle with fewer products than the minimum, tiered steps that are illogical, discount rate so weak it is unlikely to affect purchase behavior |
| Schedule | Timing issues | No schedule set, schedule already expired, recurring campaign with no valid next occurrence, large idle gap between recurring instances, timing overlap with another campaign’s schedule window |
| Reliability | Data quality and evidence concerns | Stale data (analysis not refreshed), estimated analysis mode, insufficient order data for the campaign to have meaningful economic evidence, no active campaigns in the store at all |
The “review” family is a seventh category that handles positive and operational states — things like “launch ready,” “monitor campaign,” “campaign performing well,” and usage-limit notifications. When you see a reason code from this family, Campaign Intelligence is not flagging a problem; it is giving you operational context about where the campaign stands in its lifecycle.
Each reason code has up to seven vocabulary dimensions: what happened (the label), why it matters (the impact — phrased more urgently when the verdict is At Risk, more gently when it is Needs Attention), what to do (the recommended action), why that action helps, and the button text for the CTA. This is why clicking the CTA in Campaign Intelligence takes you directly to the relevant wizard step — the system knows which step addresses each specific reason code.
The operational signals (free tier)
Beyond the verdict, reason, and confidence band, Campaign Intelligence surfaces a set of operational signals that give you more granular visibility into what is happening. In the free version of Smart Cycle Discounts, four signal categories are available:
- Campaign health — the readiness score expressed as a percentage. This is the same readiness number that feeds into verdict determination, made visible as a standalone indicator. A readiness of 90+ is the threshold for a “Ready” verdict; 70–89 triggers “Needs Attention.”
- Conflict pressure — how much overlap exists with other campaigns targeting the same products. This is only surfaced when there are actual conflicts; if there are no conflicts, the indicator is hidden rather than showing a redundant “None.” Conflict pressure directly maps to the conflicts dimension, which carries the highest weight (30%) in the health calculation.
- Inventory risk — stock exposure on discounted products, shown when count is above zero or severity is medium or high. Like conflict pressure, it is hidden when there is no active signal — a clean store does not need noise on this indicator.
- Schedule pressure — timing concerns on the current campaign. This consolidates ending-soon flags, recurring configuration issues, and timing overlap with other campaigns’ schedules into one signal.
A fifth signal, override intensity, shows the percentage of the campaign’s product scope that is being overridden by a higher-priority campaign. It only appears when override intensity is non-zero.
Hidden signals are good news, not bugs
If you open a campaign’s overview panel and see fewer operational signals than you expected, that is intentional. The system only surfaces signals that have an active value — if conflict pressure is zero, it is not shown. A clean panel with only the verdict and confidence visible means there is nothing unusual to flag, which is a healthy state.
The economic metrics (Pro)
With Smart Cycle Discounts Pro, Campaign Intelligence gains access to four additional economic metrics. These require actual attributed order data — the system needs at least five attributed orders and ten line items from the campaign before it will compute them. Until that threshold is met, the economic metrics are unavailable (not wrong — just absent).
- Revenue per discount dollar — how much revenue the campaign generates for every dollar of discount it gives away. A value below 1.0 means the campaign is costing more in discounts than it generates in additional revenue, which is a strong signal to revisit the discount depth or targeting. This is the most direct efficiency indicator available.
- Break-even uplift required — the percentage increase in order volume the campaign needs to generate to cover its discount cost, expressed as a percentage. A campaign offering a 20% discount does not automatically require a 20% volume increase to break even — the actual number depends on your base conversion rate and margin. Campaign Intelligence calculates this from your campaign’s attributed order data rather than from a generic formula.
- Margin risk — a qualitative profitability risk level (low / medium / high) derived from the relationship between your discount rate and the estimated margin on affected products. This is directional rather than precise — the system does not have access to your cost-of-goods figures unless you have configured them.
- Promotion lift — the estimated revenue increase attributable to the campaign relative to a pre-campaign baseline, expressed as a percentage. Campaign Intelligence uses store-level order data to estimate what the baseline would look like without the promotion running. This is a baseline-comparative estimate, not an exact attribution model — it works best for campaigns with clear start dates and meaningful historical data prior to launch.
Economic metrics require real attributed order data
The economic metrics compute from actual orders attributed to the campaign — at least five orders and ten line items. If you run a short flash sale that generates only two or three orders, the economic metrics will not appear. This is by design: generating a “revenue per discount dollar” figure from two orders would be statistically meaningless. If you need economic visibility on low-volume campaigns, a longer test window or higher-traffic products will get you there faster. For more on how to design meaningful discount tests, see testing percentage versus fixed-amount discounts in WooCommerce.
How to read the dashboard in practice
Most store owners open the dashboard, see a green verdict, and move on. That is a reasonable habit when things are healthy. But the dashboard is worth a slower read in two situations: when a non-green verdict appears, and when you are deciding whether to extend or expand a campaign that is currently running.
Starting with the verdict row
The main dashboard shows a focused verdict on whichever campaign the system considers most important — typically the highest-urgency active campaign, or the one with the most severe issues if multiple campaigns have problems. This is the “focus campaign.” The verdict row displays the state, the primary reason, the recommended action, and the CTA that takes you directly to the relevant fix.
Below the verdict row, the campaigns strip shows per-campaign verdict badges for all active and scheduled campaigns. Campaigns needing attention appear first. If all campaigns are healthy, the strip collapses. This is the fastest way to check that nothing is wrong across your whole campaign portfolio at once.
Reading confidence before acting
Before you take any action based on a Campaign Intelligence verdict, check the confidence band. A High-confidence “At Risk” verdict is different from a Low-confidence “At Risk” verdict. The former means the system has substantial evidence that something is wrong. The latter means it has detected a pattern that resembles a problem, but does not yet have enough data to be certain. In a Low-confidence scenario, investigate before making significant changes — especially before pausing or ending a campaign that may actually be performing adequately.
The analysis mode indicator (shown near the confidence band on some surfaces) tells you whether the analysis is exact, estimated, or exact-after-activation. Estimated mode means the product set was not fully compiled at the time of analysis — this is normal for campaigns that use random product selection or complex category filters before they have been activated. The system will upgrade to exact mode after the first activation and compilation cycle.
Following the reason code to the right fix
The primary reason code is the single most actionable piece of information in the entire Campaign Intelligence output. It tells you not just that something is wrong, but specifically what is wrong — and the CTA button that accompanies it takes you directly to the wizard step where you can fix it.
If the reason code is in the “conflicts” family, the CTA takes you to the Basic information step where you can adjust campaign priority. If it is in the “discount” family, it goes to the Discount configuration step. If it is in the “schedule” family, it goes to the Schedule step. This step routing is baked into the system — you never need to figure out where to start.
When to scale, pause, or redesign
Campaign Intelligence gives you data. What you do with it still requires judgment. Here is a practical framework for the three most common decisions:
When to scale a campaign
A campaign is a good candidate for scaling — broader product scope, more aggressive discount, or extended duration — when Campaign Intelligence shows all of the following:
- Verdict is On Track or Ready at Medium or High confidence
- Conflict pressure is zero or low
- Stock risk is low across the affected products
- If you have Pro: revenue per discount dollar is above 1.0 and margin risk is low or medium
Do not scale a campaign that is showing Low confidence even if the verdict is green. The system is telling you it does not have enough data yet. Let it accumulate a few more days of orders before making the campaign bigger. For recurring promotions in particular, the second and third cycles often have meaningfully higher confidence than the first — the system learns from each cycle. See recurring WooCommerce discount campaigns for how to structure cycles that improve over time.
When to pause a campaign
Pausing makes sense when the verdict is At Risk or Blocked and the reason is something that cannot be fixed immediately — a supplier stock issue, a conflict that requires another campaign to be rescheduled, or a pricing error that affects margin on the affected products. Pausing a campaign that Campaign Intelligence has flagged as “At Risk” because of margin exposure stops the bleeding while you work out the fix.
Do not pause based on a Low-confidence warning alone. Investigate first. Low confidence often means “not enough data” rather than “definitely a problem.” The post on when to actually pause a WooCommerce campaign covers this decision in more depth — including what other signals (refund rate, stock velocity) should be in the picture alongside the intelligence verdict.
When to redesign a campaign
A campaign needs redesigning — not just a tweak — when:
- The same reason code keeps appearing across multiple Campaign Intelligence refreshes, even after you have followed the recommended action
- Pro metric “break-even uplift required” is above 30% and actual lift is well below that threshold — the math does not work for this discount structure on this product set
- The conflict family keeps triggering because the campaign’s priority position is structurally wrong relative to your other campaigns, rather than because of a specific fixable overlap
The margin impact calculator for WooCommerce is a useful companion here — before you redesign a campaign’s discount structure, work out the break-even point outside of the plugin to validate that the new structure is viable, then use Campaign Intelligence to confirm that the configured version matches your intent. And if you run bundle campaigns in particular, the bundle discount setup guide covers the three pricing modes and their margin implications in detail.
Common questions
Does Campaign Intelligence auto-pause or auto-stop campaigns?
No. Campaign Intelligence in Smart Cycle Discounts is an advisory layer, not an automation layer. It evaluates campaigns and tells you what to do, but it does not take action on your campaigns automatically. The decisions remain yours. You can follow the recommended action by clicking the CTA, which takes you to the relevant configuration step — but nothing changes until you make the change yourself. The system tracks when you override a guardrail warning, but does not penalize you or revert the campaign.
Why does a campaign show “At Risk” immediately after I activate it?
Most commonly, this happens because a discount is configured at a depth that consumes your estimated margin on the affected products — even before any orders come in, Campaign Intelligence can flag this because the margin calculation only requires the discount rate and the store’s order history (not campaign-specific attributed orders). Check the reason code — if it is in the “discount” family and references margin or extreme discount, the fix is to reduce the discount percentage or exclude low-margin products from the campaign’s scope.
How many orders does a campaign need before the Pro economic metrics appear?
At least five attributed orders and ten attributed line items. “Attributed” means orders where a customer added products that are part of the campaign’s product scope while the campaign was active. On a high-traffic store, this threshold is typically reached within the first day or two of a campaign. On a lower-traffic store, it may take several days or may not be reached at all for a short flash sale.
Is Campaign Intelligence available in the free version of Smart Cycle Discounts?
Yes — the core Campaign Intelligence verdict (state, reason code, recommended action, confidence band) is available in the free version. The operational signals (campaign health, conflict pressure, inventory risk, schedule pressure) are also available for free. The four economic metrics — revenue per discount dollar, break-even uplift required, margin risk, and promotion lift — require a Pro license and attributed order data. The full analytics dashboard with advanced exports is also Pro-only.
Why does confidence stay Low even after several days of running?
Three common causes: (1) The campaign is in “estimated” analysis mode because it uses random product selection or a complex category filter that was never compiled — check whether the product scope was properly resolved after activation. (2) The campaign has been active but generated too few orders to build meaningful attributed economic evidence. (3) The health scores and risk scores are inconsistent — high readiness alongside high risk — which the cross-signal consistency factor penalizes. In case 1, a manual recompile or editing and re-saving the campaign usually resolves it. In case 2, more traffic is the only fix. In case 3, look at the reason code — a conflicting signal is usually the root cause.
Key takeaways
What to take away from this post
- The verdict is the primary output. Blocked / At Risk / Needs Attention / On Track / Ready tells you what to do. The scores explain why. Read the verdict first.
- Conflicts carry the highest weight (30%) in the health score. If two campaigns share products and neither has clear priority, that is the single biggest drag on Campaign Intelligence scores across both campaigns.
- Confidence matters as much as the verdict. A Low-confidence warning is an invitation to investigate, not a mandate to act. A High-confidence warning is a mandate to act.
- The reason code tells you exactly what is wrong. The CTA button routes you to the specific wizard step that fixes it — you do not need to figure out where to look.
- Free users get verdict, reason, confidence, and operational signals. Economic metrics (revenue per discount dollar, break-even uplift, margin risk, promotion lift) require Pro and at least five attributed orders.
- The dashboard gauge shows store-wide health, not a single campaign’s score. A blocking issue anywhere in your portfolio hard-caps the gauge at 40.
- Campaign Intelligence is advisory, not automated. It recommends; you decide. Nothing is paused or changed without your action.
Try Campaign Intelligence in your store
Smart Cycle Discounts includes Campaign Intelligence in the free version — verdict state, reason codes, confidence band, and operational signals for every campaign. Pro adds the economic metrics (revenue per discount dollar, break-even uplift, margin risk, and promotion lift) once attributed order data accumulates. No guessing about whether a campaign is working.
