Every screening programme has to answer the same policy question. Screening every client constantly is unaffordable. Screening each client once, at onboarding, is indefensible the first time a regulator asks when a high-risk relationship was last reviewed. The real design problem with adverse media screening is not whether to run the adverse media check but when. This article works through the three timing modes, onboarding, risk-tiered refresh and event-driven review, and how a firm sets a cadence it can afford, defend and prove.
Onboarding: The Baseline Check
The onboarding check is the one every later review will reference, so it is the one that justifies depth. It runs after identity and ownership have been resolved under the KYC process, across the customer and the natural persons behind it, against licensed news sources rather than an open web search. Licensed coverage matters because paywalled, regional and non-English reporting is where material findings tend to sit, and because a licensed archive can be searched the same way twice. The article on global news coverage for adverse news monitoring sets out why source breadth changes what a check can find.
Each result is dispositioned: relevant to this party or a namesake, material or immaterial, and if material, what it says about conduct, litigation, regulatory action or association. Material findings feed the initial risk rating alongside the sanctions and PEP checks, and the rating in turn decides which refresh tier the client enters.
The record is the point. A first check that captures the sources searched, the date, the results reviewed and the disposition of each gives every later check something to compare against. A refresh then asks a narrow question, what has appeared since this date, rather than re-reading the customer's entire history. An event-driven review asks whether the new story is already in the file. Depth at onboarding is paid for once and drawn on across the whole relationship, which is the proportionality argument for spending the time here rather than spreading a thinner check across every review.
Risk-Tiered Refresh: Setting a Defensible Cadence
Refresh is the scheduled mode, and the risk-based approach is what makes it proportionate. Rather than one cycle for the whole book, cadence is tiered by the risk rating assigned at onboarding and updated since. Firms commonly review high-risk relationships at least annually, and more often where the rating or the sector warrants it, with medium-risk relationships on a longer cycle and low-risk relationships on the longest, subject to a maximum interval the firm sets in policy.
Each tier needs a proportionality argument attached. For the high-risk tier the argument is exposure: a relationship rated high because of jurisdiction, ownership or sector is the one most likely to generate new coverage, and the one where a late finding costs most. For the standard tiers the argument runs the other way: material new coverage is less likely, screening the whole tier is costly, and the periodic review of customer due diligence that already occurs on that cycle is the natural point to attach a KYC refresh of the adverse media check.
The trap is a cadence that exists on paper but not in the data. A policy that says annual review for high-risk clients is only as good as the completed-review records that show it happened, on time, for every client in the tier. Where the review backlog is invisible, the policy is a statement of intent rather than a control. The documentation for each refresh should therefore reference the onboarding check or the previous refresh it was compared against, so that a reviewer can see both the cadence and the drift.
Event-Driven Reviews: Catching What the Cycle Misses
The scheduled cycle has a structural blind spot. A client reviewed in March and due again next March can be named in an investigation in June or change ownership in August, and the cycle will not notice until spring. Event-driven review is the mode that closes the gap: a check run between cycles because something has happened.
Trigger events fall into a few recognisable types. New adverse coverage naming the client or a connected party. An ownership or control change filed at the registry. A sanctions designation of the client, an owner or a counterparty. A transaction pattern that the monitoring system has flagged. Each is a documented reason to open the file, run the adverse media check again against the last one, and decide whether the risk rating still holds.
What makes this mode work is ongoing monitoring rather than manual re-checks. An analyst cannot search every client's name every week, and an open web search produces results that cannot be compared with the last one. Adverse media monitoring built on licensed sources and alerting does the watching instead: the client's resolved identifiers are held as a standing query, new coverage matching them generates an alert, and the alert becomes the trigger event that opens a review. The article on how technology has transformed adverse media screening covers how this shift from search to monitoring happened.
The record for an event-driven review notes the trigger, the date, what was found that the earlier checks did not hold and whether the rating or the refresh tier changed as a result.
Making the Three Modes One Policy
The three modes are not three policies. They are one screening policy with four moving parts: the baseline check, the tiered refresh cycle, the trigger list and the escalation path that applies to all of them. Written down together, they answer what a regulator or an auditor will ask: when every client is screened, how often each tier is re-screened, what prompts a review between cycles, and what happens when a review finds something.
The documentation thread is what holds the policy together. Each check references the one before it: the first refresh compares against the onboarding baseline, the next refresh compares against the last, and an event-driven review compares against whichever was most recent. A file built this way shows drift, both in the client's risk and in the firm's own adherence to its cadence. A file built as a series of unconnected searches shows neither. The same negative news finding might sit in both files, but only one of them can show when it first appeared and what was done about it.
Running Every Mode in Nexis Diligence+
Nexis Diligence+
does not treat the three modes as three products. It runs one documented check and the policy decides when. The onboarding search covers licensed news in multiple languages, global sanctions and PEP data and company records, and its saved report is the reference point in the customer file. A refresh re-runs that search on the same entity and compares it with the saved report, so the reviewer sees what is new rather than re-reading what was already dispositioned.
For event-driven review, the check is run on demand when a trigger arrives, against the same sources and with the same output, so the third mode produces a record that looks like the first two. The August article on licensed international adverse media behind screening describes the content layer that Nexis® Data+ supplies to automated pipelines; Nexis Diligence+ is where an analyst applies that content inside a documented check, whichever mode called for it. Because every mode exports to the customer file in the same format, the thread from onboarding to refresh to trigger stays legible across the life of the relationship.
Set Your Screening Cadence with Nexis Diligence+
Final Thoughts
A screening policy is defensible when every client's last adverse media check is as recent as their risk demands, and provable when every check references the one before it. Onboarding sets the reference point, risk-tiered refresh keeps it current in proportion to exposure, and event-driven review catches what the cycle misses.
The three modes make sense as one workflow rather than three tools, and Diligence+ runs all three the same way, with the same output, into the same file. Cadence is then a policy choice with evidence behind it rather than a default inherited from the last audit, and a firm that can show its timing was proportionate for every tier is in a far stronger position than one that simply screened often.