Platform · Compare
Compare like with like — or not at all
A comparison is only fair when both sides were measured the same way. Spy Monk compares an entity with its own past, or with up to three others, on the same model version and the same sources — and tells you plainly when that is not possible.
Updated By Drive Digital Editorial
Your first snapshot is the baseline
Same model version, same sources · Estimates marked as estimates · No benchmarks we cannot source
Direct answer
How does Spy Monk compare reputation over time or with competitors?
Spy Monk compares an entity with itself over time, or up to four entities over the same window, using one model version and only the sources every side has in common. When source panels differ, the differences are marked as estimates; when versions differ or fewer than two sources are shared, the comparison is blocked.
- Three verdicts: comparable, compare with care (estimates), not comparable (blocked)
- The harmonised source panel and each side's missing sources are always shown
- Scenario planner: broad ranges labelled “Scenario estimate — not a forecast or guarantee”
Principle
Why most reputation comparisons mislead
Two scores can differ for reasons that have nothing to do with reputation. One audit reached YouTube and the other did not. One month covered thirty days and the other eight. One number came from an older formula. Put side by side without saying so, those differences look like a story — we improved, they are worse — that the data does not support.
DRS treats a comparison as a claim that has to earn its place on the screen. Before it shows two numbers together, it checks that they come from the same model version, that the time windows line up, and which sources every side actually has in common. That common set is the harmonised source panel, and it is the only basis for the comparison. Sources that only one side has are listed as that side's missing sources, not quietly mixed in.
What Spy Monk will not do is invent context. There is no "average score for your sector", no percentile, no league table. We have no benchmark data we could source, and a made-up benchmark is worse than none.
Verdicts
Every comparison gets one of three verdicts
Comparable
Same model version, aligned windows and the same sources on every side. Differences are shown as differences.
Available nowCompare with care
The model version matches but the source panels differ. The comparison uses the shared sources, and every delta is marked as an estimate with each side's missing sources listed.
Available nowNot comparable
The model versions differ, or fewer than two sources are shared. Spy Monk shows the reason and no numbers side by side, because any number it showed would mislead.
Available now
Step by step
How a comparison is built
Choose what to compare
One entity across two time windows — this quarter against the last, say — or two to four entities over the same window.
Check the versions
Every side must have a score on the same model version. A v1 Health Index from an older audit is never compared with a v2 score.
Build the shared panel
Spy Monk lists the sources each side returned data from and keeps only those common to all. With fewer than two in common, the comparison stops here as not comparable.
Show what is missing
For each side: the sources it lacks against the panel, and its sample size. A side with a thin sample says so.
Calculate the differences
Deltas between sides, or between windows, are calculated on the shared panel and flagged as estimates whenever the panels were not identical to begin with.
Attach the caveats
Every result carries its caveats in words: same version and aligned windows only; a competitor's score describes the public sources about them, not their business; calculated figures such as share of voice are estimates.
Competitors
Comparing with other businesses — what is fair
You can add a competitor or a peer as an entity and compare on the public sources Spy Monk can read about them. That is ordinary market awareness, and it is all Spy Monk does: it does not connect to their accounts, look behind logins, or try to find out who their customers or reviewers are.
A competitor's score describes the public picture of them in the sources checked — not the quality of their business, and not anything you could fairly publish as a finding about them. Use it to understand where you stand and what customers talk about, not as material for a campaign. Monitoring or profiling private individuals is not allowed under the Acceptable use policy, and comparisons are for entities you are entitled to look at.
Reach, influence and share of voice are calculated estimates wherever they appear, shown with their source panel and their gaps. Where they depend on connected accounts Spy Monk does not have, they are not shown at all.
Scenarios
Improvement scenarios, not predictions
Alongside a comparison, the scenario planner answers a narrower question: if one thing about our service improved, roughly where might the score land? You choose a lever — for example better response and resolution, or fewer recurring issues — and a hypothetical change. Spy Monk shows a broad range around a midpoint, labelled Scenario estimate — not a forecast or guarantee.
The planner needs a scored baseline, so it is not available while your score is in the insufficient-data state. It assumes the same sources and model version as your current snapshot. Its levers describe real improvements in service and response; it never models synthetic reviews, rating manipulation or taking criticism down. Platforms, publishers and customers decide outcomes, and nobody can promise you a number.
Availability
Comparison features by status
- One entity over two time windowsAvailable now
- Two to four entities over the same windowAvailable nowEach entity is audited and monitored like any other, within your plan's limits.
- Verdicts, harmonised source panel and per-side missing sourcesAvailable now
- Deltas marked as estimates when panels differAvailable now
- Improvement scenarios with labelled rangesAvailable now
- Share of voice and reach from connected accountsPending feasibilityNeeds account connectors and platform permissions; until then, not shown.
- Sector benchmarksPending feasibilityOnly if a sourced, comparable data set ever exists — none is used today.
- Available now
- Available now5 items
- Pending feasibility
- Pending feasibility2 items
Questions
Comparison questions
Why does my comparison say “not comparable”?
Either the two sides were scored with different model versions, or fewer than two sources returned data for every side. Spy Monk shows the reason. Running a fresh audit for each side on the current model usually fixes the first; adding sources or widening the window can fix the second.
Why are my differences marked as estimates?
Because the sides did not have identical source panels. The comparison uses only the shared sources, but each side's score was built from a slightly different picture, so the deltas are approximate and labelled that way.
Can I compare myself with a competitor?
Yes, on public sources only, for an entity you are entitled to look at. A competitor's score describes the public sources about them in that window — not their business — and Spy Monk never tries to identify their customers or reviewers.
Is there an industry average I can compare against?
No. Spy Monk has no benchmark data it could source, so it shows none. Comparisons are always between entities or periods you chose, on a shared panel.
Will the scenario planner tell me what my score will be?
No. It shows a broad range for a hypothetical improvement, labelled as a scenario estimate and not a forecast. It assumes nothing else changes, which in practice never holds exactly.
Can I compare an old audit with a new one?
Only if both were scored on the same model version. Audits from before the current model keep the v1 Health Index, which uses a different formula and is never placed beside a v2 score.
Start with a baseline you can trust
Your first audit gives you a snapshot with its sources and confidence — the only fair starting point for any comparison.
No card needed