New analysis

Outlook 2027: six shifts in the open-source threat picture, and what each one changes for investigators.

Read the outlook

Results

What happened after we delivered.

Alerts sent, pages collected and hours billed are easy to count and prove nothing. We count what clients did with our findings and whether those findings turned out to be right. This page sets out the numbers, the rules behind them and the cases where we were wrong.

  • Outcomes confirmed by clients or the public record
  • Counted the same way since 2014
  • Misses published beside the hits
2,600+ Investigations closed since 2014 Across 30+ countries
91% Of attributions later confirmed by a court, regulator or law enforcement action 412 attributions with a known outcome
$310M+ In fraud losses frozen, recovered or avoided by clients Client-reported, 2021 to 2026
11 days Median time from tasking to finished report Urgent tasking: first assessment in 24 hours

How we count

Five rules for calling something a result.

A number without a counting rule is marketing. These rules were written in 2014 and have not been loosened since.

They make our figures smaller than they might be. Most of our work is confidential, many matters never reach a public outcome and clients do not always tell us what happened next. We count only what we can stand behind.

Read our methodology

An outcome counts only when

  • The client has confirmed it in writing, or it appears in a public record
  • Our finding was used in the action taken, not merely delivered before it
  • It has actually happened. A referral is not an arrest, and a freeze is not a recovery
  • Any monetary value was reported by the client, not estimated by us
  • It is counted once, however many engagements touched it

We do not count

  • Alerts sent, items collected or sources monitored
  • Leads that nobody acted on
  • Outcomes we believe happened but cannot confirm
  • Savings calculated from what might have gone wrong

By discipline

One measure for each kind of work.

Each discipline is held to the measure that best shows whether it did its job. Where a figure shows how often we cleared someone or found nothing, that is deliberate. It is part of the job.

Outcome measures by discipline
Discipline What we measure Result Period
Investigations & Attribution Attributions later confirmed by a court, regulator or law enforcement action 91% of 412 2014 to 2026
Due Diligence Engagements in which our report changed the terms of a transaction or stopped it 1 in 6 2021 to 2026
Threat Monitoring Critical threats escalated to the client within 60 minutes 98.6%, median 22 minutes 2026 to date
Protective Intelligence Anonymous threat authors identified, where identification was the question 64% of cases 2021 to 2026
Influence Operations Inauthentic accounts removed by platforms following our referrals 27,400 across 38 networks 2020 to 2026
Dark Web Intelligence Leak and breach claims resolved to a definite conclusion within 48 hours 50 of 58 2026 to date
Screening & Vetting Subjects screened, and the share with a material discrepancy 14,200+ screened, 3.1% 2024 to 2026
Insider Risk Matters in which the evidence pointed to an external cause and cleared the insider 34% of matters 2024 to 2026
Supply Chain Intelligence Supplier entities mapped, and projects whose key finding sat at tier three or below 9,400+ mapped, 7 in 10 projects 2022 to 2026
Geopolitical Risk Events we reported as verified that were borne out by later reporting 99.2% 2026 to date
OSINT Training Practitioners trained, and the share passing the practical assessment 3,100+ trained, 88% 2021 to 2026

The other nine percent

Where we were not confirmed.

Since 2014 we have made 1,037 attributions. For 412 of them we know what happened next. The remainder are open, confidential or were never reported back to us, and we do not extrapolate from them.

Of the 412, 375 were later confirmed and 37 were not. Twenty-nine of those 37 had been reported at moderate confidence, which is what that label exists to warn about. Eight had been reported at high confidence. Each of the eight triggered a formal review of the method, and three of those reviews changed our standard.

  • 2017: a distinctive username is no longer accepted as a link without a second, unrelated identifier
  • 2019: lines of evidence must come from distinct source types to count as independent
  • 2022: every unexplained contradiction caps confidence at moderate, however strong the rest

Read: Attribution is a standard of proof

375 Attributions later confirmed by a court, regulator or law enforcement action
29 Not confirmed, reported at moderate confidence
8 Not confirmed, reported at high confidence Each one reviewed

Client perspectives

What clients measured.

We had seventy-two hours and no way of proving a negative. Aldervane proved a positive instead: where the files really came from.

General CounselInternational freight forwarder

The seven they cleared mattered as much as the forty-six they linked. We paid those customers the same week, and we could defend every one of the declinations.

Head of Special InvestigationsMotor insurer

Our audits stopped at the factory gate. Aldervane kept going for three more tiers and showed us the evidence for every step, including the one step they could not prove.

Vice President, Responsible SourcingInternational apparel and footwear brand

When the answer is no

Three results that never appear in a headline.

Some of the most valuable outcomes are the ones where nothing dramatic happened, because the evidence did not support it.

Declined

One request in seven

Turned away at screening because the purpose, the scope or the client did not meet our standards. Work we should not do is not counted as lost revenue.

Cleared

The person it was not

Findings that pointed away from the suspect a client arrived with, reported as plainly as an attribution would have been.

Unanswered

The question we handed back

Matters where open sources could not lawfully establish the answer, closed with a statement of what would, and where it could be found.

Start an engagement

Bring us the question nobody else could answer.

Tell us what you need to know and why. A senior investigator will scope it with you, tell you candidly what open sources can and cannot establish, and propose a plan with a fixed fee and a firm timeline.