What Does a Kofax Capture to TotalAgility Migration Actually Cost?
Short answer: there’s no price list, and anyone who quotes you before counting your batch classes is guessing. The cost of moving from Kofax Capture (and KTM) to TotalAgility is driven by six things: how many batch classes you run, how deep your KTM projects go, how much custom script you’ve accumulated, how many integrations hang off your capture flow, your document volume, and your deployment model. Small, clean estates are a weeks-to-months project. Large, scripted estates run six to twelve months or more. Tungsten’s Move Up program offsets part of the licensing, not the services. Here’s how to size yours before anyone sizes it for you.
Why nobody publishes a price
Search for TotalAgility migration cost and you’ll find technical migration guides and not much else. That’s not vendors hiding a number. It’s that the number genuinely doesn’t exist until someone counts what you’ve built. Two companies running the same Kofax Capture version can be a 10x apart in migration effort, because the product has been quietly customized for fifteen years in one shop and run near-vanilla in the other. So instead of a fake number, here are the six factors that produce the real one.
“Nobody could tell us what fifteen years of batch classes would cost to move. Every answer started with ‘it depends’ and ended without a number.”
“Our validation scripts are the product. The OCR part is easy. The question is what happens to two decades of custom code.”
“The Move Up credits looked generous until we scoped the services. The license was never the expensive part.”
The six factors that drive the cost
the license credit is the visible number. services, retraining and the parallel run are the real budget. size those first
Effort bands by deployment size
These are qualitative bands, not quotes. Your script depth and integration count move you between bands more than your company size does.
| Estate profile | What it looks like | Typical effort band | What dominates the cost |
|---|---|---|---|
| Small / clean | A handful of batch classes, little or no KTM, standard export connectors, light scripting. | Weeks to a couple of months. | Licensing transition, basic rebuild and testing, user training. |
| Mid / typical | Dozens of batch classes, KTM projects with trained extraction, some validation scripts, two or three downstream integrations. | Three to six months. | KTM project conversion and revalidation, script rework, integration re-plumbing. |
| Large / scripted | Hundreds of batch classes, deep KTM tuning, heavy custom code, many integrations, regulated validation requirements. | Six to twelve months, sometimes more. | Custom code rebuild, extraction revalidation, parallel running, formal testing and sign-off. |
What the Move Up program offsets, and what it doesn’t
Tungsten runs a formal “Move Up to TotalAgility” incentive program for Capture and KTM customers, and it’s worth pursuing. Two things to know going in. First, the terms are negotiated per account. There’s no public rate card, so whatever you’re offered, get it in writing, including what happens to your existing maintenance spend. Second, the program lives on the licensing side of the ledger. It helps with the transition to TotalAgility entitlements. It doesn’t pay for the migration services, the script rebuild, the extraction revalidation, the retraining, or the months of parallel running. Those are your budget lines whether the license credit is generous or not. A good business case shows both columns honestly.
The hidden costs that surprise people
- Retraining. TotalAgility is not Capture with a new skin. It’s a process platform with a different designer and a different operating model. Your batch class experts become TotalAgility learners for a while, and that dip in productivity is a real cost.
- Parallel running. You’ll run both environments side by side while volumes shift across. Two sets of infrastructure, two sets of attention, one team. Budget it deliberately or it budgets itself.
- Extraction revalidation. Every trained KTM project needs its accuracy proven again on the new platform before finance or operations will trust it. In regulated environments this is formal testing with sign-off, not a spot check.
- Integration re-testing. Every downstream system that receives your captured data gets a full regression pass. The capture migration is done when the ERP posting is right, not when the batch completes.
How to get a real number: inventory first
The order matters. Estimate before inventory and you get a range wide enough to be useless. Inventory first and the estimate becomes an engineering exercise.
- Export your batch class inventory. Count them, then mark which ones actually ran in the last twelve months. Dead batch classes cost nothing to leave behind.
- Catalog your KTM projects. Which document types, how much training data, which locators are custom-scripted versus configured.
- List every script and customization. Validation scripts, recognition scripts, workflow agents, custom panels. This list is the cost multiplier, so it deserves a day of someone’s time.
- Map the connections. Import sources, email ingestion, export connectors, API consumers. Name the system on the other end of each one.
- Pull twelve months of volume. Pages, documents, and peaks. This sizes licensing and infrastructure.
- Then ask for a fixed-scope estimate. With that inventory in hand, a competent partner can give you a real number and a real timeline, and you can hold them to both.
What the return side looks like
Cost is half the business case. On the delivered TotalAgility IDP projects behind our roughly twenty automation case studies, clients have cut manual document errors by 98%, saved $50-70k a year in processing cost, and freed 100+ hours a month of staff time, with payback typically landing in six to twelve months. Those numbers came from estates that did the inventory first and scoped honestly. The migration cost is real, but so is the number on the other side of it.
Frequently asked questions
Does Tungsten publish TotalAgility migration pricing?
No. What’s publicly available is technical migration documentation, not cost guidance. Licensing is quoted per account, Move Up incentives are negotiated per account, and services depend entirely on your estate. That’s why the inventory-first approach matters: it’s the only path to a number you can defend in a budget meeting.
What happens to our KTM projects in the migration?
TotalAgility carries transformation capability forward, and Tungsten provides converter tooling to bring Capture and KTM assets across. Configuration and structure convert best. Custom scripts, heavily tuned locators, and anything exotic get flagged for rework. Plan for revalidation of extraction accuracy on every project that matters, because “it converted” and “it performs the same” are different claims.
Do the Move Up credits cover implementation services?
Treat the program as a licensing-side incentive. The services work (migration, script rebuild, revalidation, training, parallel running) is a separate budget whoever performs it. Terms differ per account and can change, so get the specific offer, its expiry, and its conditions in writing before you build your business case on it.
Can we run Kofax Capture and TotalAgility in parallel during the migration?
Yes, and you should. The normal pattern is to migrate batch classes in waves, run both platforms side by side, and retire Capture once volumes have shifted and validation has passed. The March 3, 2027 end-of-support date bounds how long that parallel period can responsibly last, which is an argument for starting the inventory now rather than in 2027.
Is there a Kofax Capture 12 we could wait for instead?
No. Capture and KTM are marked Discontinued, and there’s no successor version coming. Version 11.1 is the last stop, with end of support on March 3, 2027 (KTM 7.1 follows on May 9, 2027). Waiting for the next Capture release isn’t a strategy anymore, because there isn’t one.
How long do we have to decide?
Work backwards from March 3, 2027. A mid-sized migration of three to six months, plus a quarter of parallel running and validation, plus procurement and budgeting time in front of it, means a decision made in early-to-mid 2026 lands comfortably and a decision made in 2027 doesn’t. If you’re reading this near the deadline, the effort bands above tell you whether you still fit.
Can we migrate only some batch classes and leave the rest?
Yes. A phased move is the norm, and the inventory usually reveals batch classes that shouldn’t move at all: retired processes, one-off projects, duplicates. Migrating 60 live batch classes instead of 150 total ones is often the single biggest cost saving available, and it costs nothing but an honest triage.
Does our document volume change the cost much?
It changes two things: TotalAgility licensing and infrastructure sizing, and the duration of parallel running and validation. High-volume estates need a throughput design, not just a functional migration, because extraction services behave differently under load. Pull twelve months of real volume data, including peaks, before anyone sizes anything.
Want your number instead of qualitative bands?
Book a Tungsten Migration & AI Readiness Session. We inventory your batch classes, KTM projects, scripts, and integrations, then give you a sized estimate for both paths, Move Up or move out. Backed by roughly twenty automation case studies and delivered TotalAgility IDP projects with 98% fewer manual errors and six-to-twelve-month payback. No pitch, just the number.