Ask a compliance vendor what their platform costs and you will almost always get the same answer: it depends. That is not evasion — the variables are real — but it leaves you writing a budget line for something you cannot price, and defending it to a CFO who quite reasonably wants a number.
This guide gives you the number, or rather the method to build it. It sets out the pricing models you will actually meet, the costs that live outside the licence and surprise most buyers, and a way to size the whole thing from figures you already hold. Every figure quoted here comes from the benchmarks published elsewhere in this blog; none of them is a quote, and none of them is a substitute for one.
1. Why nobody publishes a price
Three structural reasons, and they matter because each one tells you something about how to negotiate.
Scope varies more than volume. Two institutions with identical client counts can need radically different systems. One onboards natural persons through a single channel; the other onboards corporate entities with multi-level ownership chains, foreign holdings and domiciliary companies. The second buys three times the software. Client count alone is a poor proxy for scope, which is why headline per-client pricing is rare.
The data bill is not the vendor's. Sanctions, PEP and adverse-media screening comes from data providers — World-Check, LexisNexis, KYC Spider, Polixis and others. Some platforms resell it, some let you bring your own contract. A quoted platform price that silently excludes the data is not comparable with one that includes it, and vendors know it.
Implementation dominates year one. For a modern SaaS onboarding platform, expect 4 to 12 weeks from kickoff to production. Enterprise client-lifecycle-management deployments run 6 to 18 months. That range alone can swing the first-year total more than the licence does.
The practical consequence: never compare two proposals on the licence line. Compare them on total first-year cost and total three-year cost, with the data contract and the implementation effort made explicit in both.
2. The four pricing models you will meet
Nearly every proposal in this market reduces to one of four shapes, or a blend of two.
Per named user (seat). You pay per compliance officer, analyst or relationship manager with an account. Simple to forecast and easy to compare. It punishes you for giving read access to the wider business, which is precisely what you want for a defensible audit trail — so check whether view-only or occasional users are billed at the same rate.
Per file or per onboarding. You pay per client onboarded or per periodic review completed. This aligns cost with activity and is attractive when volumes are lumpy. The risk is the opposite of the seat model: success gets expensive. If your growth plan doubles onboarding volume, model the bill at the volume you intend to reach, not the one you have.
Per client under management, or by AUM band. You pay by portfolio size, often in tiers. It is the most predictable of the four and the easiest to budget, but it decouples price from usage entirely — an institution with 500 dormant relationships pays like one with 500 active ones. Ask where the tier boundaries fall relative to your actual growth curve; crossing one can produce a step change you did not plan.
Platform fee plus consumption. A base subscription covering the workflow and the audit trail, plus metered charges for screenings, registry extracts or document processing. The most honest reflection of how these systems consume resources, and the hardest to forecast. Insist on a worked estimate at your real volumes, and on visibility into the meter before the invoice.
None of these is intrinsically better. What matters is whether the model matches the shape of your business: seat-based if your team is stable and your volumes swing; volume-based if the reverse; tiered if you value predictability above all and can live with paying for capacity you do not use.
3. What sits outside the licence
This is where budgets break. The licence is usually the part everyone remembers to include.
Screening data. Depending on provider and coverage, this can rival the platform licence itself. Decide early whether you are bringing your own contract or buying through the vendor, and price both ways. Bringing your own preserves negotiating leverage with the data provider and lets you switch platforms later without renegotiating your data feed — a point worth more than it looks at signature time.
Integration. Connecting to your core banking system, your CRM or your portfolio management system is engineering work, whether you or the vendor performs it. A vendor who cannot describe their integration approach for your specific stack is quoting you an unpriced risk.
Migration. Existing client files have to arrive in the new system in a state that survives an audit. Extraction, mapping, cleaning and reconciliation of legacy records routinely takes longer than the platform configuration.
Training and change management. The platform does not produce savings until the team stops working the old way in parallel. Budget for the overlap period explicitly rather than pretending it will not happen.
Your own people's time. The largest hidden line. Discovery workshops, configuration decisions, test cycles, parallel runs and sign-off all consume your staff, not the vendor's. At a fully loaded Swiss analyst cost of CHF 80,000 to 110,000 a year, a project consuming a third of one person for four months is a real five-figure sum that never appears on any invoice.
4. Build the budget from what you already know
You do not need a vendor quote to size this. You need three numbers you can produce today.
Your current cost per file. Manual onboarding in a private-banking context runs roughly CHF 300 to 800 per file, and can reach CHF 350 to 1,000 where corporate structures are involved. If you have never measured yours, take an analyst's loaded annual cost, divide by the number of files they complete in a year, and add the tooling and archiving overhead. The result is usually uncomfortable, and it is the number that makes the business case.
Your annual volume. Onboardings plus periodic reviews. Reviews are the line most institutions underestimate, because the backlog hides them.
Your analyst capacity. Manual throughput sits around 15 to 25 files per analyst per month. Automated, that figure moves to 80 to 120. The gap is the capacity you are currently buying with headcount.
Multiply cost per file by annual volume and you have your baseline — what compliance costs you today, before any software. Automated processing brings the per-file figure to roughly CHF 50 to 150. The difference between those two totals is the envelope inside which any software investment has to sit to be worth making. If a proposal consumes more than that envelope, the answer is no, regardless of how good the demo was.
5. A worked example
Take the profile used in our ROI analysis: an intermediary with 500 active clients and 2 full-time equivalents on KYC/AML.
The manual baseline comes to roughly CHF 200,000 a year — CHF 180,000 in loaded salaries for two FTEs, CHF 5,000 in tooling and archiving, and around CHF 15,000 absorbed by errors, delays and overtime.
Automation of document flows, client reminders, verification and reporting typically frees 60 to 70 percent of the time spent on repetitive work — about 1.2 FTEs, or CHF 108,000 of capacity returned to the business each year. For an organisation of this size, the all-in annual investment including licences, integration and support lands in the region of CHF 30,000.
That gives a net first-year gain near CHF 78,000, a return around 260 percent, and payback in three to four months. Over three years, with portfolio growth and continued optimisation, cumulative returns of 400 to 500 percent are the norm rather than the exception.
Two cautions about that arithmetic. The freed capacity is only a saving if you actually redeploy it — if the 1.2 FTEs continue doing what they did before, you have bought software and kept the cost. And the figure excludes what is usually the largest avoided cost of all: the regulatory exposure of a file you cannot defend.
6. What too cheap and too expensive both look like
Price is a signal in both directions, and the failure modes are not symmetrical.
Suspiciously cheap usually means the tool automates one link of the chain and leaves the rest to you. Document collection without screening. Screening without disposition. Disposition without an audit trail. Each of these relocates the bottleneck rather than removing it, and you discover the gap during your first inspection rather than during the demo. It can also mean the data contract is excluded, in which case the real total arrives later from a different supplier.
Expensive is not automatically wrong — enterprise CLM platforms genuinely cost more because they genuinely do more. It becomes wrong when the price is driven by modules you will never switch on, by a multi-year commitment that outlives your certainty about your own requirements, or by an implementation whose length means you pay for two years before you save for one. A vendor who cannot commit to a timeline is quoting a cost with no ceiling.
The uncomfortable middle case is the proposal that looks reasonable because the first year is discounted. Model year three. Discounts expire; tier boundaries do not move.
7. Questions to ask before you sign
Commercial due diligence, not technical. These are the questions whose answers change the total.
- What exactly is included in the licence, and what is billed separately?
- Is screening data included, and if I bring my own provider contract, does the price change?
- What is the implementation timeline you will commit to contractually, and what happens if it slips?
- Which parts of implementation consume my staff rather than yours, and for how many days?
- How is migration of existing files priced, and who is accountable if the data arrives incomplete?
- Where do the pricing tiers sit, and what does my bill look like at twice today's volume?
- What does year three cost at list price, with no introductory discount?
- What are the exit terms — can I export my files, in what format, and at what cost?
That last question matters more than its position suggests. The cost of leaving is part of the cost of arriving, and it is far easier to negotiate before signature than after.
8. Frequently asked questions
How much does KYC/AML software cost?
There is no list price in this market, because scope varies more than volume: two institutions with the same client count can need very different systems depending on whether they onboard natural persons or corporate structures. The workable approach is to size the envelope from your own baseline — manual onboarding runs roughly CHF 300 to 800 per file, automated processing brings that to CHF 50 to 150, and the difference across your annual volume is what the investment has to fit inside. For a 500-client intermediary with two FTEs, the published worked example lands near CHF 30,000 a year all in.
Why do vendors refuse to publish prices?
Because three variables dominate and none of them is visible from outside: the scope of what you onboard, whether the screening data contract sits with the vendor or with you, and how long implementation takes. A published price would be wrong for most readers in both directions. It is not, by itself, a red flag — but a vendor who will not give you a worked estimate at your real volumes after a scoping conversation is a different matter.
What is usually missing from a quoted price?
Screening data, integration with your core banking or portfolio system, migration of existing files, training, and your own staff's time during the project. That last item is the largest and the least often budgeted: at a loaded Swiss analyst cost of CHF 80,000 to 110,000 a year, internal project time is a real five-figure sum that appears on no invoice.
Which pricing model is best?
The one that matches the shape of your business rather than the one with the lowest headline. Per-seat pricing suits a stable team with variable volumes; per-file suits the reverse; per-client or AUM tiers suit institutions that value predictability and can accept paying for unused capacity; platform-plus-consumption reflects actual resource use but is the hardest to forecast. Model each at the volume you intend to reach, not the one you have today.
How quickly does the investment pay back?
For a mid-sized intermediary, payback typically falls in the range of three to four months, with a first-year return near 260 percent and cumulative three-year returns of 400 to 500 percent. Those figures assume the freed capacity is genuinely redeployed to higher-value work. If the team continues as before, the software is a cost rather than an investment.
Should we build instead of buying?
Buy, unless your requirements are genuinely unique and you have a standing engineering team to maintain compliance logic as regulation changes. Building typically takes 12 to 24 months to first value and leaves you owning every regulatory update; buying a maintained platform delivers value in 4 to 12 weeks. Most institutions should build only the thin proprietary layer and buy the regulated plumbing.
9. Where Wecan fits
We do not publish a price list, for the reasons set out in section 1 — and we would rather tell you that plainly than print a number that would be wrong for most of the institutions reading this. What we will do in a scoping conversation is give you a worked estimate at your volumes, state explicitly what falls inside the licence and what does not, and commit to an implementation timeline in writing.
Screening runs through the provider of your choice; you keep your own data contract if you have one. And if the envelope you calculated in section 4 does not accommodate what we would need to charge you, we will say so rather than sell you a deployment that cannot pay for itself.
