Does the internet understand your company?
The Digital Reality Score measures how accurately, completely and visibly a company is represented across the sources assessed. This page publishes how: the definition, the rules, the thresholds and the limits. Every Score page links back to the evidence behind its number.
What the Score combines
The Digital Reality Score combines accuracy, completeness and presence across the sources assessed. It is an index, never the probability that information about a company is correct. Its three components are shown beside it wherever it appears.
| Component | Question | What is counted | Out of |
|---|---|---|---|
| Accuracy | Are the assessed material representations correct and current? | Each place a fact is shown, weighted by the fact’s importance and how much the place is seen (below) | The representations of facts whose truth sources establish or dispute |
| Completeness | Are the facts that apply stated across the sources assessed? | Each fact in four groups of sources: the company’s own site and profiles, independent sources, AI assistants asked by name, and the records on listings where the company is present. Stated counts 1, partly stated ½; stated is not the same as right | The fact-and-group checks that apply and could be verified, weighted by importance |
| Presence | Is the company on the relevant, applicable surfaces for its category and market? | The category’s listing and review sources and the main company databases, Google page 1 for the name, the category questions put to AI assistants, and partners’ own sites. Present 1; not found 0. A listing under a wrong name or website is present: its error counts once, under Accuracy | The checks that apply and could be verified |
The Score is the weighted geometric mean of the three, with equal weights (1 / 1 / 1). Equal weights are a starting point, not a calibrated choice. A geometric mean lets a weak component pull the Score down: strength in one cannot hide a gap in another. Near zero it is so steep that one small listing could move the Score by twenty points, so in the Score each component counts as at least 10. The component itself is always shown as measured.
What applies. A source a company is not eligible for is left out, with the reason: an agency is not expected on software review sites, a software product not on agency directories, a company assessed in Portugal not on Danish registers, and a platform’s app marketplace only applies when an integration with it is established. Facts follow the company type: a services company is not asked for a named product unless it adds one; positioning is checked for every company and the plans offered for software. Absence is scored only after this.
What “not found” means. Each check records what it looked at. A search with no matching result is “not found in this search”, not proof of absence. A failed fetch, a blocked page or a failed search is “could not verify”: it is left out and never counted as missing.
When there is no Score. Each component needs enough evidence: Accuracy at least 3 facts established, one of them high importance, and 30% Evidence Coverage; Completeness 60% of its checks verified; Presence 60% of its weight verified and at least 5 checks. If one falls short, the profile shows the components it has and no Score, rather than a number built on a small readable subset.
Weights inside the components. Completeness weighs its four source groups equally (1 / 1 / 1 / 1), and Presence its channels equally (1 / 1 / 1 / 1). Inside the listing channel each source weighs by how often AI assistants cite it for the category, combined with the share of the category’s other assessed companies listed there, where at least 2 were checked on it. Neither is a measure of how many buyers see a source. AI category answers count with at least 4 usable answers, relative to the companies named most; those companies are a comparison, never labelled competitors. Each category’s reference (its questions, answers, sources and the companies already assessed in it) is frozen for the calendar month, so a change in the benchmark is never read as a change in the company.
Outside the Score. Evidence Coverage, how much of each component could be verified, and confidence are shown beside it, never in it. How many independent sources back each fact is shown per fact.
Accuracy: what it measures
Accuracy is the expected accuracy of what people and machines encounter about a company, weighted by how much each fact matters and how often each surface is seen.
Take every material claim about the company whose Business Reality could be established. Take every place that claim is represented: the company’s own website, registries, reference works, databases, press, search results and AI answers. For each representation, estimate the probability that it matches what is actually true. Weight it by the claim’s importance and by how much that surface is encountered. Accuracy is the weighted average, on a scale of 0 to 100.
Because the weights work as sampling weights, the number keeps a plain reading: a representation of the company, drawn in proportion to importance and exposure, is accurate with probability Accuracy / 100, across the material claims and surfaces assessed.
Earning the percentage sign
A probability is only meaningful if it is calibrated: of all the representations the model rates 80% likely to be accurate, about 80% should turn out accurate when checked independently. Until that has been shown, the Score is published as an index, 80 / 100, with this reliability interpretation. It will be published as a percentage only when calibration evidence supports it.
Calibration is measured against a growing set of claims whose truth has been verified independently. That set deliberately does not rely on claimed companies alone. Companies that claim their profile are self-selected and interested parties, so a random sample of unclaimed companies is verified by analysts every quarter. Calibration results, and the size of the validation set, will be published here.
Bands and confidence
Higher means better represented. The Score always travels with a band, a confidence level, Evidence Coverage, the date and the methodology version.
| Band | Range |
|---|---|
| Aligned | 85–100 |
| Minor gap | 70–84 |
| Moderate gap | 50–69 |
| Significant gap | 0–49 |
On a report, each of the three components carries a word beside its number: Strong from 85, Fair from 70 and Weak below 70. The word is a reading aid; the number is the measure.
The sites that matter
Each category has its own list of the sites that matter: the review, comparison and listing sites that AI assistants cite when asked for a shortlist in the category (at least 2 answers in the last 180 days) or that an analyst adds, together with the main company databases and open records (wikidata.org, crunchbase.com, linkedin.com, brandfetch.com). The list follows the category’s answers, so it moves as they do, and an analyst reviews it every quarter. A report checks the company on the sites that apply to it and counts each site once, with the reading that needs most work.
Sites are ranked by how much fixing them improves the Score, then by how often AI assistants cite them for the category, then by how many competitors are listed there.
Coverage. A report also counts how many of the category’s listing sources the company and the competitors shown are on, from the same checks. Present means we found the company’s own record, on the page or in a search of the site; a link to its website alone is not a record. Sources that could not be verified are shown as such and left out of the count. Competitors are checked for presence, not accuracy. Coverage is a plain count: sources are not of equal value, it is not added to the Score, and it is not the chance that an AI assistant recommends a company. The Presence component reads these sources with its own weights, so the two do not match.
Statuses
| Status | What it means |
|---|---|
| Right | The site shows the same facts as your website. |
| Right, could be stronger | Right, but a fact that applies is missing from it. |
| Outdated | The site shows an earlier value: your old positioning, a previous CEO, a former address. |
| Conflicting | The site shows a different value, and nothing proves which is right. Counted with the sites to fix. |
| Wrong | Only when the company has corrected the fact, or a company register states the company’s value. Disagreement alone is Conflicting. |
| Missing | Not listed on a site where competitors in your category are. |
The breakdown on a report always adds up: the sites checked are the sites right (including could be stronger) plus those outdated, conflicting, wrong and missing. The sites to fix are the outdated, conflicting, wrong and missing ones.
How to fix, and when it’s usually live
Each site shows how it is fixed:
| How | What it means |
|---|---|
| Profile | You create or edit your own profile on the site. |
| The site changes its record when you write to it. | |
| Form | The site has a correction or submission form. |
| Edit | An open record anyone can edit, with sources (Wikidata) or through its editors. |
| Your website | The fact is stated on your own website, where sources and AI read it. |
Usually live in. The time from submitting a fix to seeing it live on the site, as the median of every tracked fix on that site, shown as “(usually live in 3 days)” or, for a fix across several sites, “(usually live in 3 days – 2 weeks)”. It appears only once a site has at least 10 tracked fixes. Before that the report shows no time. We never estimate a time we have not measured.
Marking a fix as submitted records the date on the company’s profile. PageRadar SIGNAL can then check each site and record when the change is live; every tracked fix adds to the site’s timing. A company’s first tracked fix is free.
Places to get listed
The sites that matter are where a company is checked and scored. Places to get listed are a second list: further directories and marketplaces where a company could add itself. They are not part of the Score, a report does not count them, and they live in PageRadar SIGNAL.
What we don’t do
- We don’t edit, submit or write to sites for a company. Fixes are made by the company, on its own accounts.
- We don’t buy, sell or arrange listings or reviews.
- Paying never changes a Score: a Snapshot or a tracked fix adds detail, never points. A Snapshot run never becomes the public Score.
- We don’t call a source wrong because it disagrees: it is Conflicting until the company corrects the fact or a register settles it.
- We don’t pass bot challenges or pose as another crawler.
Claims are the unit
A claim is an important fact about the company, written as subject, predicate and value. Which claims are worth checking, and what can be true of a claim, are defined once, by the Digital Reality Gap framework. This page adds only what scoring requires.
Acme → primary_category → Enterprise operations platform
Acme Flow → product_available → Generally available
Claims are weighted by importance (high 3, medium 2, low 1) and come in five kinds, because not every claim can be known the same way:
- Fact
- Independently verifiable: leadership, founders, headquarters, ownership, size, certifications. Needs independent evidence.
- Self-defined
- The company is the main authority: category, positioning, products, pricing. The company’s own statement counts for more.
- Subjective
- Opinion or marketing, such as “leading provider”. Recorded, never scored.
- Tracked
- Read like a fact and shown as findings, never part of the Score: leadership beyond the CEO. Each source names a different part of a team, so it rarely settles, but a former executive still listed in a role is worth seeing.
- Diagnostic
- About being found rather than about the company, such as whether a search for the name leads to it. Shown as diagnostics, never part of the Score.
Two questions, kept apart
The Score answers two different questions, and never uses the answer to one as evidence for the other.
1. What is true?
For each claim the methodology estimates the probability that the value the company states is Business Reality.
- Start from a prior. A publicly stated fact starts at 60%; a self-defined claim at 78%. When the company states nothing, a value proposed by sources starts at 50%.
- Only independent, citable evidence moves it: registries, reference works, press, databases, directories. Each one shifts the log-odds by its authority × reading confidence × freshness, times 2 when it supports the stated value and -2 when it contradicts it.
- Copies collapse to their origin. A directory that repeats Crunchbase is Crunchbase again. One hundred copies are one vote, cast by the strongest of them.
- Search results and AI answers never count as proof. They are how a company is represented, not evidence of what is true. Five AI answers agreeing leave a claim exactly as unresolved as none.
- The company’s own properties never prove themselves. The website sets the stated value; it cannot corroborate it.
- Evidence ages. It loses half its weight every 365 days.
Business Reality counts as established when the probability rises above 80% or falls below 20%. Below 20%, the company’s own statement appears to be wrong, which is itself a finding: Stated Reality is often where an error starts. Between the two, the claim is unresolved.
2. What is shown?
Given that probability, each representation of the claim is scored on how likely it is to be accurate: a representation that matches the stated value is accurate with the same probability; one that is outdated or contradicts it is accurate only if the stated value is wrong. Each representation is weighted by exposure, how much that surface is encountered. The same wrong fact costs more in a knowledge panel than in an obscure directory.
Exposure in drs-0.12 (preliminary) uses published proxies by kind of surface. They will be replaced by measured exposure: search demand × rank position for search surfaces, answer frequency for AI.
| Surface | Exposure proxy |
|---|---|
| Search results, knowledge panels | 1.00 |
| Company website | 0.80 |
| AI answers | 0.70 |
| Company profiles (LinkedIn, maps) | 0.50 |
| Wikipedia, Wikidata | 0.50 |
| Press | 0.35 |
| Company databases | 0.30 |
| Directories and review sites | 0.20 |
| Registries and filings | 0.05 |
Three claim states
- Aligned
- Important sources represent the claim as the company states it.
- Gap
- Sources represent the claim differently from the company, with enough evidence to say so.
- Unresolved
- Not enough independent evidence to decide. Unknown is not treated as wrong.
A claim is aligned when Business Reality is established and its representations are at least 75% likely to be accurate. Below that it is a gap, and a gap always names its kind, using the Digital Reality Gap framework’s claim states:
- Incorrect
- Wrong, at the source or somewhere downstream.
- Stale
- True once, and still circulating after it changed.
- Conflicting
- Different sources give different answers.
- Unsupported
- Stated, with no independent evidence behind it.
- Missing
- True, and absent where people look for it.
- Misunderstood
- Present, and interpreted as something else.
- Divergent
- Right in one market or language, wrong in another.
“Incorrect” is reserved for a stated value that the evidence says is wrong, a company correction, or a registry-grade source (authority 0.9+) contradicting it. Disagreement alone is “Conflicting”.
Evidence Coverage
The Score measures how accurately, completely and visibly the company is represented. Evidence Coverage says how much of the company’s material reality could be assessed confidently.
Coverage is the importance-weighted share of fact and self-defined claims whose Business Reality could be established. There is one Score, not a public one and a verified one. Unresolved claims are left out of the Score and lower Coverage instead. Unknown is never treated as wrong, and a Score of 95 on 30% coverage says something very different from 95 on 97%.
Diagnostics
Six views explain where a Score comes from. From drs-0.2 they are not averaged into it.
| Diagnostic | Question |
|---|---|
| Accuracy Probability that a source, any source, represents each claim correctly, with every source counted equally. The Score weights the same thing by exposure. | Are important company facts represented correctly? |
| Consistency Whether sources tell the same story as each other. Diagnostic only: sources can agree on something wrong, so agreement never raises the Score. | Do important sources agree? |
| Evidence How many independent origins corroborate each claim. Copies of the same source count once. Two independent origins count as sufficient. | Are important company claims sufficiently supported? |
| Authority The standing of the strongest independent source supporting each claim. Registries and reference works rank above directories. | Are credible and authoritative sources reinforcing those claims? |
| Discoverability The share of surfaces checked that pass: crawler access, each AI assistant asked, partners’ own sites, the knowledge graph, the company’s own site and search. Partial counts half; a surface not checked is left out. Preliminary: every surface counts equally until calibration. Never part of the Score. | Can the places people and machines look reach the company and know it? |
| Representation Probability that search results, knowledge panels and AI answers represent each claim correctly, weighted by exposure. | Do search and AI systems describe the company correctly? |
Discoverability: where buyers and machines look
The Score asks whether what is shown about a company is right. Discoverability asks whether the places buyers and machines look can reach the company and know it. It explains the Score and is never part of it.
Each run checks a set of surfaces. Each surface is a pass (1), partial (0.5) or fail (0). Discoverability is the share of the checked surfaces that pass, with a partial counting half. A surface that was not checked is listed on the profile and left out. It is never counted as a failure. Each surface is also shown as a finding, with what was observed and when.
| Surface | Pass | Partial | Fail |
|---|---|---|---|
| Crawler access | Our honestly identified crawler can read the site, and robots.txt lets AI answer crawlers in | Readable, but robots.txt keeps some AI answer crawlers out | The site cannot be read, or robots.txt keeps every AI answer crawler out |
| Each AI assistant asked (named with its model, access route and whether it searched) | Answers at least 60% of the facts asked, none of them wrong or out of date | Answers some, or an answer differs from what the company states | Answers none |
| Partners’ own sites (a brand’s distributor list, a vendor’s partner directory) | Every partner checked lists the company | Some do | None of those found lists it |
| Knowledge graph (Wikidata) | An entity with the company’s website | — | No entity |
| The company’s own site | States at least 75% of the core facts | States at least 40% | States fewer |
| Search (findability claims, where a run has them) | Searching the company leads to it | Partly | Leads elsewhere |
| Search results: the company’s own pages our searches return | Every one still exists | One or two answer “not found” (404/410) | Three or more, or half of them, answer “not found” |
| Structured data (schema.org Organization) | Present with name, url, address, legalName and sameAs, and nothing that contradicts the site | Present, but one or more of those fields missing | None, or it contradicts the site (name, address or founding year) |
| Contact details: the site, its structured data, partners’ own pages | Address and phone agree wherever they are given | A partner’s page gives a different address or phone | The site and its own structured data disagree |
| Category questions: AI assistants asked for a shortlist of companies like this one, without its name (each question asked twice of every assistant, web search on) | Named in at least 75% of the answers | Named in some answers | Named in none (not checked below 4 usable answers) |
| Where the category is read: the listing and review sources of the company's category (up to 20 from the category map, plus the main company databases) | Listed on sources carrying at least 70% of the weight | At least 35%; a listing under a wrong name or website counts half | Less than 35% |
| Google results: page 1 for the company’s name (up to 10 organic results, through a search API; Google’s edition for the domain’s country, else the United States, named on the profile) | The company’s own site ranks first | Its own site is on page 1 below other results, or first beside a knowledge panel for another name | Its own site is not on page 1 |
| Independent record: the facts we check that at least one independent source states (a register, a database, the press, a directory, a reference work), weighted by importance | At least 90% | At least 50% | Less than 50% |
Category questions are asked once per category and market, in the same words for every company in it, and never with a company’s name, products or features in them. A global category is asked “Can you recommend some good …?” and “Who are the leading …?” with the category in plain words; a country’s category is asked the same in the market’s language and in English. Their answers are read for every company scored in that category within 7 days of the day they were asked, so the companies in one category are compared on the same questions and the same answers. Each profile shows the date the questions were asked.
Not sampled yet, so not counted: Google AI Overviews, Perplexity, and the consumer apps of ChatGPT, Claude and Gemini. AI assistants are asked through their makers' APIs, and each profile names the model, the access route and whether it searched the web. That is not the same as the consumer app, which can answer differently.
Where the category is read
Buyers and AI assistants compare companies on listing and review sources: review platforms, alternatives directories, marketplaces, company databases. Each category has a map of them. A source enters the map when AI assistants, asked for a shortlist in the category, cite it in at least 2 answers in the last 180 days, or when an analyst adds it. A source is weighted by its share of those citations; a source not cited yet weighs by its type, below every cited source. Companies’ own sites are cited too, and are never counted as sources.
Each run checks the company on the map’s 20 heaviest sources plus wikidata.org, crunchbase.com, linkedin.com, brandfetch.com. Per source it records whether the company is listed, under which name and linking to which website. A listing counts 1, a listing under another name or website 0.5, no listing 0. A source that could not be checked is left out. Many of these sources block automated readers; their records are then read from search results, at 60% confidence, and from a link index that records where their pages link, with its date. We never pass a bot challenge or pose as another crawler.
A listing’s name and website are also representations for the Score, weighted by how exposed the source is. Limits: assistants are asked through their APIs, not the consumer apps; Perplexity is not sampled; weights firm up as runs accumulate; a category without a map yet is not checked.
Preliminary. Every surface counts equally because the weights have not been calibrated yet. Calibration against a larger reference set will weight them, and that will be a new methodology version.
Sources and provenance
Every observation keeps its provenance, using the framework’s source states:
- Declared
- Supplied and confirmed by a verified person at the company.
- Observed
- Found on properties the company controls.
- Evidenced
- Found in an independent, citable source.
- Perceived
- How search results and AI systems represent it.
- Unknown
- Not enough evidence either way.
Authority and exposure are different properties. Authority is how much a source counts as evidence of truth; exposure is how often it is encountered. A registry has high authority and little exposure; an AI answer has exposure and, for truth, none.
| Source | Counts as evidence | Authority | Exposure |
|---|---|---|---|
| Companies register, regulatory filing | Yes | 0.95 | 0.05 |
| Wikipedia | Yes | 0.85 | 0.55 |
| Business press, analyst reports | Yes | 0.80 | 0.40 |
| Industry publications | Yes | 0.75 | 0.30 |
| Company databases | Yes | 0.70 | 0.30 |
| Review sites | Yes | 0.55 | 0.35 |
| Directories | Yes | 0.50 | 0.15 |
| Company website | No | — | 0.80 |
| Search results | No | — | 1.00 |
| Search knowledge panel | No | — | 0.90 |
| AI assistant answers | No | — | 0.70 |
When the company answers
A verified person at a claimed company can confirm, correct or ask to clarify a claim. Every answer is evidence, labelled Declared. None of them sets the Score. An answer shifts the log-odds by 2 while pending and 3.5 once an analyst has reviewed it, and it can never make a claim certain.
Answers mostly raise Evidence Coverage. They can lower the Score: confirming a fact the wider web gets wrong turns an unresolved claim into a gap. A company cannot choose or buy a better Score. It can only help establish what is true. Company-supplied answers are private unless the company publishes them.
Badges
A claimed company can display its Score when Evidence Coverage is at least 80%, the Score is no older than 120 days, confidence is not low and no high-importance claim is unresolved. A badge means the company’s assessed digital representation is highly likely to be accurate and current. It does not say the company is trustworthy, ethical or good at what it does. The Score measures information, not character.
Limitations
- An automated Score cannot know every fact about a business. It knows what can be observed and established. That is why Coverage is published and unknown stays unresolved.
- Priors, evidence strengths, authority and exposure are published starting estimates. Calibration will adjust them, and each change is a new methodology version.
- Dependency between sources is recorded where it is known. Undetected copying still overstates corroboration.
- AI answers vary by model, date, language and phrasing. Each run records which assistants were tested and when.
- drs-0.12 scores in English. A company can be represented accurately in one language and badly in another.
- A Score is a snapshot at the date shown on it.
Versions
| Version | Change |
|---|---|
| drs-0.12 (preliminary) | drs-0.11 with two more facts, positioning (every company) and the plans offered (software), in Accuracy and Completeness; leadership beyond the CEO tracked and shown as findings, never scored; and in the Score each component counts as at least 10, so a component near zero no longer swings it. |
| drs-0.11 (preliminary) | The Score combines three components: Accuracy (drs-0.10’s Score), Completeness and Presence, as their weighted geometric mean with equal weights. Sources a company is not eligible for, and checks that could not be verified, are left out with the reason. Each component needs enough evidence, or no Score is shown. A source repeated with the same value is one representation, and a company’s answer counts only once reviewed. Demo companies stay on drs-0.10: they have no listings to measure completeness and presence from. |
| drs-0.10 (preliminary) | Disputed facts count in the Score. A fact whose truth is not established because sources show different values now counts at its expected accuracy: each representation weighs the probability that it is right, given the evidence. Before, such a fact left the Score, so one more source getting it wrong could raise the Score. A fact nothing disagrees with but that could not be established still lowers Coverage, never the Score. Coverage and Discoverability are unchanged. |
| drs-0.9 (preliminary) | One more surface for Discoverability: the independent record. The share of the facts we check, weighted by importance, that at least one source the company does not control states: a register, a database, the press, a directory or a reference work. It passes at 90% and is partial from 50%, and the facts no independent source states are named. Not having findings is a finding: where nothing independent speaks for a fact, people and AI assistants have only the company’s own word, or nothing. A fact sources disagree on is not silent. The Score’s rules are unchanged. |
| drs-0.8 (preliminary) | One more surface for Discoverability: Google results. Page 1 of Google for the company’s name (up to 10 results, Google’s edition for the domain’s country, else the United States) shows whether the company’s own site comes first. It is part of Discoverability, never of the Score. The research now reads Google’s results for its fixed searches through a search API instead of searching itself: the same searches, a fixed set of results per run, and no knowledge panel or search page counted as evidence. The Score’s rules are unchanged. |
| drs-0.7 (preliminary) | The category layer. Each category has a map of the listing and review sources it is read from, taken from the pages AI assistants cite and from analysts’ lists. The company is checked on the map’s heaviest sources and the main company databases: listed, under which name, linking to which website. A listing under an old name or website is a representation for the Score, weighted by the source’s exposure, and a finding. Discoverability gains one surface, “Where the category is read”. The counting rule is unchanged. |
| drs-0.6 (preliminary) | Four changes from a 20-company reference set. A source that shows an earlier or a different value is a finding of its own, even when most sources are right or the truth is not settled; the Score is unchanged by it. A self-defined value (name, category, product, pricing, markets, integrations, website) that an independent source shows is taken as the company’s, so another value elsewhere lowers how accurately it is shown instead of removing the claim from the Score. The company register is looked up by the company’s country, and confirms a stated CEO and the registered name (a different registered director is noted, never counted: registers can lag an announced change). The official website is a claim. |
| drs-0.5 (preliminary) | One more surface for Discoverability: category questions. AI assistants, with web search on, are asked for a shortlist of companies like this one, without its name, two questions twice each; the surface passes when the company is named in most answers. The Score and the counting rule are unchanged. |
| drs-0.4 (preliminary) | Three more surfaces for Discoverability: the company’s own pages that search results still send people to but that no longer exist; structured data that is complete enough to identify the company and says what the site says; and contact details that agree across the site, its structured data and its partners’ own pages. The Score and the counting rule are unchanged. |
| drs-0.3 (preliminary) | The Score, Coverage and the other diagnostics are unchanged from drs-0.2. Discoverability is now counted from surface checks: crawler access, each AI assistant asked, partners’ own sites, the knowledge graph, the company’s own site and search. Every surface counts equally until calibration, surfaces not checked are left out, and each surface is shown as a finding. |
| drs-0.2 | The Score becomes an exposure-weighted expected accuracy of representations. Truth is estimated from independent evidence only, with copies collapsed and freshness decay. Company answers become weighted evidence. Consistency and the other five dimensions become diagnostics, outside the Score. Findability claims move to diagnostics. Badges gain eligibility rules. |
| drs-0.1 | Weighted mean of six dimensions, including source agreement. |
Published Scores keep the version that produced them. Scores from different versions are not directly comparable.
How often Scores are updated
Published Scores are re-run periodically, and every run is kept in the company’s history, with the methodology version that produced it: a Score from drs-0.11 on combines three components, earlier Scores measured accuracy alone, and the two are not directly comparable; drs-0.12 adds facts, so its Scores are not directly comparable with drs-0.11’s either. The research reads Google’s results for a fixed set of searches. A company assessed again within 7 days of its last search reads the same results again, so a re-run is scored on the same evidence (Google returns different pages for the same searches from day to day). After 7 days, or when the searches change, the results are collected fresh. Continuous monitoring of search, mentions and AI representation is not part of the Score; that is what PageRadar SIGNAL does.
Independence, and the framework
The Score is maintained by PageRadar, which also sells services that help companies improve how they are represented. How that conflict is handled is set out in the independence policy. The Score is built on the Digital Reality Gap framework: the framework explains why the realities diverge, the Score measures it.
Preview status
This is methodology drs-0.12 (preliminary). Scores of real companies are automated and unlisted: not in search engines or our sitemap, and opened only by people with the link. A sample report on a fictional company is run through this exact methodology from invented evidence.