Adviserflow DirectEnterprise Wealth Tech
Compliance

Regulatory Change Management Software: Seven Mistakes That Leave Obligations Unmapped

Most regtech platforms are good at sending alerts and weak at turning them into obligations with owners. Here are seven selection and rollout mistakes, each with a demo test that exposes it.

By the Adviserflow Direct editorial team · November 2026 · 6 min read

Alerts are easy; owned obligations are the hard part

Most firms evaluating regulatory change management software already get more alerts than they can read. What they lack is a dependable chain from a regulator's publication to a specific obligation, then to the control that meets it and the person who signs off on that control. If the chain breaks anywhere, an examiner or internal auditor is likely to find the gap before you do.

The seven mistakes below recur when banks, wealth managers, lenders and payment firms in different countries choose and roll out these platforms. Each comes with a fix and a test you can run live in the demo, so you score what the product does with your material, not what the slide deck claims.

Advertisement

Mistake one: buying a news feed and calling it an obligations library

Horizon scanning and regulatory intelligence feeds tell you something was published. They don't tell you what you now have to do. One rulemaking can produce a consultation, a final text, technical standards and several rounds of Q&A. If the product only delivers those documents, your analysts still break each one into requirements by hand.

What to do instead: Require a structured obligations library. Each requirement should be a single atomic obligation with a citation, version history and a clear record of what changed between versions.

Demo test: Choose a rule your team knows thoroughly, such as the EU's DORA provisions on ICT third-party risk or a Monetary Authority of Singapore AML/CFT notice. Ask the vendor to show the obligations extracted from it, the source paragraph behind each one, and the redline between the last two versions.

Mistake two: accepting thin coverage outside the EU and UK

Coverage maps often claim well over a hundred countries. Depth is a separate question. In some markets a vendor may track primary legislation and central bank headlines but miss circulars, FAQs, securities-regulator guidance or sub-national rules. Examples include Reserve Bank of India master directions, Saudi central bank circulars, Brazilian central bank resolutions and US state money-transmitter rules.

What to do instead: For each jurisdiction, get coverage broken down by issuing body and document type. Also get typical update latency, whether content is in the original language or translated, and whether items are mapped to obligations or delivered only as alerts.

Demo test: Pick three non-EU/UK markets where you hold licenses. Ask for every item the vendor captured there in the last 90 days, and compare the list with what your local compliance staff tracked manually. Any gap in a licensed market should count heavily against the vendor.

Mistake three: leaving obligations unlinked to controls and owners

An obligations register that isn't connected to your control library and accountability map is just a reading list. When a rule changes, nobody is notified and no control gets reviewed.

What to do instead: Insist on many-to-many mapping, since one obligation can need several controls and one control can satisfy many obligations. You also need an impact-assessment workflow with due dates, a named owner on each obligation and evidence capture. If your controls live in a GRC tool such as Archer, ServiceNow IRM or MetricStream, put that integration in scope from the start, not in phase two.

Demo test: Load an obligation already mapped to two controls, then push an amendment to it. Watch whether the system flags both controls, sends tasks to their owners and records the audit trail without anyone intervening.

Mistake four: adopting the vendor's taxonomy without reconciling your own

Vendors tag content by topic, such as AML, conduct or operational resilience. Your firm is organized by legal entity, product and activity. If the two structures aren't reconciled, applicability filtering fails in one of two ways: people are flooded with irrelevant items, or relevant ones are filtered out and nobody notices. The second is the dangerous one.

What to do instead: Before go-live, map the vendor's taxonomy to your entity, product and activity structure. Confirm your custom fields survive the vendor's own taxonomy updates.

Demo test: Ask the vendor to filter for something like "consumer lending entity in Mexico plus e-money entity in the Philippines." Then have them explain why each returned item is included and what was excluded.

Mistake five: trusting automated summaries without source traceability

Many platforms generate summaries and applicability tags automatically. They help with triage, but treating them as your record of interpretation is risky. Supervisors in most jurisdictions expect the firm to show its own reasoning, not a vendor's paraphrase.

What to do instead: Link every obligation to the exact source text, with a version date and a flag showing whether a person reviewed it. You also need a way to record your counsel's interpretation where it differs from the vendor's. Decisions on how a rule applies to you belong with qualified legal or compliance professionals.

Demo test: Click from a summarized obligation through to its source paragraph. Then ask who reviewed it and what the vendor's contractual turnaround is for fixing a reported error.

Mistake six: treating migration of the existing register as an afterthought

Your current obligations register probably sits across spreadsheets, SharePoint lists and an older GRC tool. Moving it is a data-quality project, not an import job.

Say a payments firm operating in 12 markets has 4,000 register rows (a hypothetical). During cleanup it finds that many are duplicates, superseded or too vague to map. If nobody budgeted for that work, go-live slips by months.

What to do instead: Plan a separate cleanup phase covering deduplication, mapping to the vendor's library IDs and retirement of obsolete entries, with business owners signing off.

Demo test: Send a 200-row sample of your register in advance. Ask the vendor to show automated matching against their library, with match confidence scores and a list of everything that didn't match.

Mistake seven: pricing year one and ignoring how the bill grows

License costs usually scale with jurisdictions, content modules, legal entities and seats, and some vendors charge separately for translation or API access. Firms that expand into new markets can find year three costs far more than year one.

Firm profileTypical annual subscriptionTypical implementation (one-off)
One region, 1–5 jurisdictions$40k–$120k$25k–$100k
Multi-region, 6–25 jurisdictions$120k–$400k$100k–$350k
Global group, 25+ jurisdictions, multiple integrations$400k–$1.5M+$300k–$1M+

These are indicative 2026 market ranges in USD. They vary by contract and local pricing differs, so confirm figures with each vendor.

Demo test: Ask for written prices for adding two jurisdictions and 50 read-only users in year two. Ask for a cap on annual uplift too.

Scoring the demo and the decision rule to apply

Use one scorecard for every vendor so the comparison stays honest.

TestPass looks like
Obligation extractionAtomic obligations, cited, with version redlines
Non-EU/UK coverageMatches or beats your manual 90-day log
Control linkageAmendment routes tasks to owners automatically
Taxonomy filterExplainable entity/product results
TraceabilitySource text, reviewer and correction SLA visible
Migration sampleMatch confidence plus a clear unmatched list
Growth pricingWritten year-two pricing with uplift cap

Check where the data is hosted, too. Data-protection regimes such as the GDPR, India's DPDP Act and Saudi Arabia's PDPL may affect where obligation and evidence records can sit. Ask your counsel which apply to you.

The decision rule: any vendor that fails the coverage test in a jurisdiction where you hold a license is out, however good its interface looks. Rank the rest on control linkage first, because that test decides whether obligations end up with owners.

Quick answers

What's the difference between regulatory horizon scanning and regulatory change management?

Horizon scanning spots new and upcoming publications from regulators. Change management takes those publications through to obligations, impact assessments, control updates and sign-off. Many products are strong at one and weak at the other, so test both in the demo.

Can we use our existing GRC platform instead of buying a dedicated tool?

GRC platforms usually handle workflow and control libraries well, but most don't produce regulatory content themselves. A common setup pairs a content and obligations provider with the GRC system through an integration. Before choosing that route, check that the integration carries version changes all the way through to the affected controls.

How long does a typical rollout take?

As a rough range, firms covering one region often go live in three to six months, while global groups can take nine to eighteen. Register migration and taxonomy reconciliation usually drive the timeline more than software setup does. Ask vendors for reference timelines from firms of similar size and footprint.

Written by the Adviserflow Direct editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.