Competitor Monitoring KPIs: How to Measure Whether Your Programme Is Useful

A competitor-monitoring programme can produce a busy dashboard without producing a better decision.

More websites, products and alerts may look impressive, but volume is not value. A useful programme reliably observes the agreed market, surfaces changes worth reviewing and helps named people make clearer commercial decisions. Only some of those decisions will create a measurable financial result.

The measurement problem is therefore not:

“How much revenue did our alerts generate?”

It is:

“Did the monitoring work as intended, improve the decision process and contribute to an outcome we can defend?”

That distinction prevents a team from inventing ROI while still holding the programme accountable.

Measure a Chain, Not One Headline Number

Competitor monitoring sits at the start of a chain:

1. the agreed competitor and category scope is observed;
2. a relevant change is detected;
3. the change is reviewed and verified;
4. a decision is recorded;
5. an action may be tested;
6. a commercial outcome may follow.

Every link matters. If coverage is unreliable, alert totals are misleading. If alerts are not reviewed, detection has not become intelligence. If a decision leads to an action but no baseline was recorded, the commercial result is difficult to attribute.

HM Treasury's Magenta Book separates three useful evaluation questions:

• process — was the activity implemented as intended?
• impact — what changed, and how far can that change be attributed to the activity?
• economic value — how do attributable benefits compare with costs?

A competitor-monitoring scorecard can use the same logic without becoming a large research project. Start with process measures, add decision evidence and claim financial value only when the attribution is credible.

Begin With the Programme's Intended Decisions

Before selecting KPIs, state what the monitoring is meant to improve. Examples include:

• reviewing the price position of a core sofa range;
• noticing meaningful competitor launches before a category meeting;
• confirming product removals without reacting to a temporary absence;
• reducing repetitive manual website checks;
• giving marketing and merchandising teams one consistent evidence base.

These are different purposes, so they require different measures.

A programme intended to improve price reviews should not be judged mainly by the number of new-product alerts. A programme intended to reduce manual collection should record time spent before and after the change. A programme intended to support range decisions needs a decision log, not merely an inbox count.

If the scope itself is still unclear, use How to Choose Which Competitors and Product Categories to Monitor First before building the scorecard.

Layer 1: Coverage and Reliability KPIs

The first question is whether the agreed monitoring actually happened.

Successful observation rate

successful scheduled observations / planned observations × 100

Define “successful” before calculating the rate. A completed scan should not automatically count if the relevant category was outside its agreed scope or the result was incomplete.

Track the reason for every gap:

• website unavailable;
• monitoring run failed or remained incomplete;
• category not selected;
• website intentionally paused;
• capacity or schedule constraint;
• scope changed during the period.

This makes the KPI actionable. A percentage alone cannot tell you whether to fix a website, narrow the scope or change the schedule.

Scope freshness

Record the share of core competitor-category combinations observed within their agreed review window.

Freshness is more useful than pretending every site needs the same cadence. A commercially important sofa competitor may need a different review rhythm from a watchlist accessories category. Measure against the documented plan, not an arbitrary universal frequency.

Persistent coverage gaps

Count core combinations that missed more than one agreed observation window.

The purpose is not to punish occasional failures. It is to reveal where the programme repeatedly promises evidence it does not supply. Those gaps should lead to a repair, a reduced scope or an explicit expectation change.

Fido presents monitored websites and scan status, while its scheduled monitoring can surface supported product and price changes after successful scans. Those records can support a coverage review, but the customer scorecard should remain focused on useful scope rather than exposing internal scanner diagnostics.

Layer 2: Signal Quality KPIs

Once coverage is dependable, measure whether the changes entering the workflow are worth attention.

Relevance rate

reviewed changes judged relevant to the agreed scope / reviewed changes × 100

A change may be technically correct but commercially irrelevant. Examples include:

• a peripheral category that should not have been monitored;
• a product with no meaningful comparison to your range;
• a repeated observation that adds no new evidence;
• a change too small to affect any agreed decision.

Record the reason when a change is judged irrelevant. A high irrelevant-change count may indicate a scope problem rather than a detection problem.

Verification rate

material changes verified before a decision / material changes reviewed × 100

Verification means confirming that the product identity, price movement, launch or removal is sufficiently clear for the intended decision. It does not mean proving a competitor's private motive.

This distinction is especially important for missing products. A potentially missing item, a confirmed removal and a later reappearance should not be treated as the same level of evidence.

Repeat-noise rate

duplicate or unchanged items closed without new evidence / reviewed items × 100

This is a practical way to identify alert fatigue. Keep the definition narrow: do not label a correct “no action” decision as noise. A relevant alert can still be valuable when the conclusion is to hold price, wait or collect more evidence.

Layer 3: Workflow and Decision KPIs

The next layer measures whether information reaches a person who can use it.

Review completion rate

material alerts reviewed within the agreed window / material alerts received × 100

Choose the window according to the decision. A pricing review may be more time-sensitive than a quarterly range discussion. Avoid describing every alert as urgent.

Ownership completion

Track the proportion of follow-ups with:

• one named owner;
• a due date;
• a recorded decision;
• the evidence used;
• a next review date where uncertainty remains.

This exposes a common failure mode: alerts are read but remain unresolved because responsibility is unclear.

Decision completion rate

reviewed material alerts with a recorded decision / reviewed material alerts × 100

A valid decision can be:

• act now;
• run a limited test;
• wait for another observation;
• escalate to another team;
• take no action and record why.

“No action” is not a failed outcome. Avoiding an unnecessary discount or campaign can be useful. The mistake is assigning a fictional cash value to that avoided action without evidence.

Collection time displaced

Measure:

baseline manual collection time - current collection and review time

Take a baseline before changing the process. Include the human time still required to verify and interpret changes; automation does not remove judgement.

The result can show whether repetitive collection has decreased. It should not be presented as a financial saving unless the business can show that the released time was genuinely redeployed or reduced a real cost.

For the operating rhythm behind these measures, see How to Build a Competitor Monitoring System Without a Big Team.

Layer 4: Commercial Outcome Measures

Commercial KPIs belong at the end of the chain, attached to a specific decision.

Choose one primary outcome before acting:

The metric should match the action. A price decision should not be credited with all category revenue growth. A competitor launch alert should not receive credit for a campaign result when the offer, creative, stock position or season also changed.

Google Analytics' attribution-path reporting illustrates the wider problem: different touchpoints may initiate, assist or close a customer journey, and credit depends on the attribution method. A competitor alert is even further upstream. It may inform a test, but it is not itself a customer touchpoint.

Use a baseline or comparison

Where practical, compare the outcome with:

• the same product or category before the action;
• an unaffected but reasonably comparable product group;
• a holdout audience;
• the expected result already used by the business;
• repeated periods that account for obvious seasonal patterns.

Record other changes that could explain the result, including promotions, availability, media spend, creative, site changes and wider demand.

This does not make every result causal. It makes the evidence more honest.

When an ROI Calculation Is Defensible

A basic programme ROI formula is:

(attributable benefit - total programme cost) / total programme cost × 100

The arithmetic is easy. The words “attributable benefit” are the difficult part.

Include programme costs such as:

• the monitoring service;
• setup and maintenance;
• staff review and verification time;
• implementation cost for actions or tests;
• any additional media, discount or operational cost.

Include benefits only when the evidence supports them. Examples might be incremental contribution from a controlled campaign test or a real labour cost removed from the process.

Do not include:

• the value of every product observed;
• all revenue after an alert;
• an assumed percentage improvement;
• hypothetical losses that “must have” been avoided;
• time saved without a measured baseline;
• commercial gains that could not be separated from other changes.

When attribution is weak, report the operational and decision KPIs instead of forcing an ROI number. “The programme reliably covered the core scope and informed twelve documented decisions” is more credible than an unsupported revenue claim.

A Practical Competitor-Monitoring Scorecard

Use a compact scorecard with an owner and a trend, not a page of disconnected numbers.

Avoid setting universal targets copied from another organisation. Establish a baseline, identify the weakest link and improve it deliberately.

A Clearly Hypothetical Review

Imagine a sofa retailer has completed its first review period.

The team finds that its core competitor-category scope was observed consistently, but several alerts came from a peripheral category. Most material changes were reviewed, although two lacked a named owner. One verified price change led to a limited merchandising test; several others produced documented decisions to wait or take no action.

The honest conclusion is:

• the collection process is working;
• the scope needs one adjustment;
• ownership needs tightening;
• the programme is informing decisions;
• commercial ROI is not yet established.

That is a useful evaluation. It tells the team what to improve without inventing a success story.

Review the Programme in the Right Order

Use this sequence:

1. Confirm the purpose and intended decisions.
2. Measure planned versus successful coverage.
3. Review relevance, verification and repeat noise.
4. Check ownership and decision completion.
5. Measure displaced manual effort against a real baseline.
6. Attach one commercial measure to each deliberate action.
7. Record alternative explanations.
8. Calculate ROI only for benefits that can be credibly attributed.
9. Reduce, repair or stop parts of the programme that remain unused.

The final step matters. Measurement is not only for proving value. It should also reveal which competitors, categories, alerts or meetings no longer deserve time.

The Takeaway

The best competitor monitoring KPI is not the biggest number on the dashboard.

A useful programme covers the scope it promised, produces relevant evidence, reaches an owner, closes decisions and supports measured action. Financial value comes last, after a baseline and a defensible link between the signal, the decision and the outcome.

For marketing-specific tests, use From Data to Action: How to Turn Competitor Alerts Into Smarter Marketing Decisions. That article explains campaign applications; this framework evaluates the monitoring programme itself.

Measure the chain honestly. If the evidence stops at a better decision process, report that. It is stronger than an ROI figure the business cannot defend.

Sources

• HM Treasury: Magenta Book — Central Government guidance on evaluation, updated 15 May 2026.
• Google Analytics Help: Key events attribution paths report, reviewed 17 August 2026.