Document Management Software Cost

Most business software gets more expensive when you hire someone. Document management gets more expensive when you do nothing. The seat price behaves normally, the storage meter underneath it climbs every month against a volume that only goes one way, and the capture meter climbs with how busy the business is. Three years in, plenty of buyers are paying two or three times the storage line they signed for, on the same plan, with the same headcount, having never once approved an increase.

This verdict prices document management the way a small business actually gets billed for it, which means treating the per user number as the smallest interesting part of the question. It maps the three meters, walks the storage arithmetic across three years rather than one, prices scanning and OCR where volume pricing genuinely hurts, treats migration as the decision problem it really is, and separates the governance features that justify the category from the cheap storage that keeps getting mistaken for it. It sits alongside our verdict on the true cost of business software, which takes per seat billing apart in general, and our manual on migrating to new software, which is the single most relevant piece on this site to what you are about to spend. Put your own volumes through the true-cost calculator and the companion on this page before you shortlist anything.

What document management software actually costs

Ask what a document management system costs and the useful answer has four parts, and only one of them behaves like a normal software line.

There is the per user subscription, which is what the pricing page advertises and what everybody compares. There is metered storage, charged against the volume you hold once you pass an included allowance, which rises continuously and quietly. There is metered capture, charged per page or per document for scanning, OCR and automated extraction, which rises with business activity. And there is the one off project cost of getting your existing content into the system, which in this category is routinely the largest number in year one and is almost never on any pricing page.

Those four parts scale on four different things. Licences scale with headcount. Storage scales with retained volume, which is a function of time and of whether anyone ever deletes anything. Capture scales with throughput. Migration scales with how disorganised your current shared drive is, which is a cost you have already incurred and are about to be billed for.

The fifth part, which appears on no invoice, is the tier you were pushed into by a single retention or audit requirement. Everything below takes these apart with illustrative planning figures, and the companion on this page reprices the whole stack against your own volumes as you read.

The headline: illustrative per user pricing by tier

Start with the numbers people search for, framed as planning bands rather than quotes, because pricing in this category moves constantly and varies by vendor, region, contract length, storage allowance and how much of the records machinery you switch on.

File sync and share seats commonly land around $8 per user per month. This is not document management, and the reason it appears in the ladder at all is that a large share of buyers compare against it and conclude the category is overpriced. It gives you space, sharing, a recycle bin and a filename search.

Entry document management plans commonly land around $12 per user per month. Expect real version history, folder and document level permissions, metadata fields you define, full text search across text based files, and a modest storage allowance.

Standard plans commonly land around $22 per user per month. This is the rung where retention rules become schedules rather than reminders, where approval and review workflow appears, where an audit trail starts recording views as well as edits, and where OCR is included at some allowance.

Advanced plans commonly land around $38 per user per month and carry the records management machinery: disposition schedules, legal hold, granular field level permissions, richer capture and classification, and the administrative reporting a compliance function asks for.

Enterprise tiers commonly start around $55 per user per month and are bought for single sign on, directory provisioning, data residency choices, sandbox environments and contractual commitments rather than for anything about documents.

Read that ladder against your own obligation rather than against a competitor’s stack. A business that needs shared files with history is shopping the entry rung. A business that has to apply a retention schedule and show who saw what is shopping the standard rung at minimum. A business under a records regime is shopping advanced and should not pretend otherwise.

The three ways this category gets metered

Almost every pricing surprise in document management traces back to the fact that three meters run simultaneously and only one of them is on the page you were shown.

The seat meter. Per user per month, billed monthly or annually, exactly as you expect. It changes when headcount changes and is the only line most buyers model.

The storage meter. An allowance is included, commonly expressed per user, and everything beyond it is billed per gigabyte per month. This verdict uses an illustrative allowance of 10 GB per user and an illustrative overage of 25 cents per gigabyte per month. The important property is not the rate, it is that the quantity ratchets upward.

The capture meter. Pages or documents processed through scanning, OCR, classification or data extraction, billed against an allowance. This verdict uses an illustrative allowance of 500 pages per user per month and an illustrative overage of 4 cents a page.

The seat meter is the one you negotiate. The storage meter is the one that quietly compounds. The capture meter is the one that spikes when the business has a good quarter, which is a genuinely uncomfortable pricing shape: the tool costs more precisely when you are busiest.

Storage is the meter almost nobody models

Here is the mechanism that makes this category different from every other one this site prices. Storage volume in a document system is monotonic. It only goes up, unless a human being makes a decision to delete, and human beings almost never make that decision voluntarily.

Nothing about ordinary operation reduces it. Superseded versions are retained on purpose, because retaining them is the point of version control. Scanned originals are kept alongside their extracted text. Email attachments filed into matters stay filed. Departed employees’ folders are kept because nobody wants to be the person who deleted them. Duplicates accumulate because two people filed the same document into two places and neither knows.

So the storage line behaves like a subscription that raises its own price. If you hold 120 GB against an 80 GB allowance today, you are paying 40 GB of overage. Add 45 GB a year and by the start of year three you are paying 130 GB of overage on exactly the same plan, for exactly the same people, having approved nothing.

The practical consequence for budgeting is that a one year total is close to meaningless here. Model three, and model the growth explicitly rather than assuming today’s number holds. The companion on this page does the three year arithmetic against your own figures.

Three years of storage growth, priced

Take the illustrative eight person professional practice used later in the worked totals. Standard plan at $22, so licences are $176 a month and $2,112 a year. The included allowance at 10 GB per user is 80 GB. The practice holds 120 GB today and adds about 45 GB a year.

Year one is charged against 120 GB, so 40 GB of overage at 25 cents is $10 a month and $120 for the year. Year two is charged against 165 GB, so 85 GB of overage is $21.25 a month and $255. Year three is charged against 210 GB, so 130 GB of overage is $32.50 a month and $390. Three year storage total: $765.

The interesting figure is not $765. It is that the annual storage line more than triples, 120 to 390, on a business that did not grow and did not upgrade. Run the same arithmetic on the three person consultancy and it goes from $30 to $138, a multiple of over four and a half, because a small allowance is a smaller cushion against the same absolute growth. Run it on the thirty person firm and it goes from $900 to $2,460.

That is the honest headline of this category. The seat price is stable and the bill is not.

Pricing by the page and by the document

Some systems, particularly those built around capture rather than collaboration, price all or part of the service by volume of documents processed instead of by user. You will see per page rates for scanning and OCR, per document rates for classification and data extraction, and occasionally a per transaction rate on a workflow that touches a document.

The logic is defensible. A three person insurance brokerage processing forty thousand pages a month is a heavier customer than a thirty person consultancy processing two thousand, and a per seat model prices those identically. Volume pricing puts the cost where the load is.

The problem is that it converts a fixed cost into a variable one that tracks your revenue activity, and small businesses budget badly for that. A per seat bill is predictable and boring. A per page bill is a line that rises when the phone rings more, which is exactly when nobody is looking at the invoice.

If you are quoted per page or per document, ask for three things in writing: the exact definition of a countable page or document, whether reprocessing a failed document counts again, and whether there is a ceiling. Vendors that will commit to a monthly cap are pricing honestly. Ones that will not are selling you an open meter.

OCR and where volume pricing actually hurts

Optical character recognition is the feature that turns a scanned image into something search can read, and it is the single most common place a document management bill goes somewhere the buyer did not model.

The mechanism is simple. An allowance is included with the plan, expressed per user or per tenant, and it is generous enough that most collaboration led businesses never touch it. Capture led businesses blow through it in the first month and then pay per page indefinitely.

Take the thirty person firm from the worked totals. The included allowance at 500 pages per user per month is 15,000 pages. The firm actually runs about 26,000 pages a month, so 11,000 pages are billable at 4 cents, which is $440 a month and $5,280 a year. Across three years that is $15,840, which is larger than the entire three year cost of the eight person practice.

Note what that means for tier choice. The advanced capture module at an illustrative $6 per user per month would cost this firm $2,160 a year, and if it carries a materially larger allowance it pays for itself several times over. That is the rare case in this category where buying up is straightforwardly cheaper, and it is invisible unless you have counted your pages first. Count them before you shop.

Scanning a paper backlog and what changes the quote

Software OCR is not the same purchase as sending boxes of paper to a scanning bureau, and both may be on your project.

A bureau quote for routine loose paper, covering preparation, scanning and basic indexing, commonly lands around an illustrative 8 cents a page. That number is a starting point and moves a long way. Staples, bindings, folded plans, fragile originals, double sided pages, poor contrast, mixed sizes and anything requiring handling rather than a feeder all push it up, sometimes several times over. Indexing depth matters most of all: one field per document is cheap, six fields keyed by a human is a different service.

The thirty person firm in the worked totals has a 240,000 page backlog, which at 8 cents is $19,200. That single line is larger than a full year of its licences and is a bigger share of the three year total than storage and migration combined.

Which raises the question nobody enjoys. Does the backlog need scanning at all? A great deal of archived paper is retained for a retention period and then destroyed, never once retrieved. Scanning it converts a cheap storage problem into an expensive digitisation project plus an ongoing storage line. Scanning what is actively referenced and leaving the rest in boxes until its retention period expires is frequently the right commercial answer, and it is a question for whoever advises you on your retention obligations rather than for a vendor.

Cheap cloud storage is not this product

This is the comparison that decides most purchases, usually badly. General file sync and share at an illustrative $8 a seat looks like the same thing for a third of the price. It is not, and the difference is worth naming precisely rather than hand waving at “features”.

Version control in a document system means a defensible chain: every version retained, attributable, restorable, with the ability to say what the document contained on a given date. Consumer grade version history is a convenience feature with a retention window, which is a different promise.

Retention policy means a schedule that acts on its own. Files are classified, a rule attaches, and disposal happens on time whether or not anybody remembers. A shared drive has no opinion about how long anything should live.

Audit trail means a record of who opened, viewed, printed, downloaded and altered a document, kept in a form somebody outside your business could be shown. Basic activity logs record edits and shares, not reads.

Permissions granularity means control at the document, and sometimes the field, rather than at the folder. Folder level permissions are why the wrong person can see the whole directory because they needed one file in it.

Search means reading the content of scanned images and structured metadata, not matching filenames.

If you need none of those five, the cheaper product is genuinely the correct purchase and this verdict is not going to pretend otherwise. If you need any of them, you are not comparing prices for the same thing.

Version control and what it is actually worth

Version control is the cheapest feature in the category to describe and the hardest to value until the day it matters.

The value is not that you can undo a mistake. Every product on the ladder can undo a mistake. The value is that the history is complete, attributable and cannot be quietly edited, which means it functions as evidence rather than as a convenience. A contract, a specification, a policy or a set of accounts that changed over time is far more useful when you can show exactly which text was in force when, and who put it there.

The practical cost implication is storage, because keeping every version of everything is precisely why your volume climbs faster than your document count. Some systems let you set version limits or age out old versions automatically, and that setting is one of the few genuine levers on the storage line. Ask about it before you sign, because a product that keeps unlimited versions forever with no pruning policy has an unbounded storage bill built into its design.

The other implication is discipline. Version control only produces a clean history if people work on the document inside the system instead of downloading it, editing on a desktop and uploading a copy called final version two. That is an adoption problem, not a software one, and it is covered further down.

Retention policy is the feature you are really buying

If you strip this category back to the one capability that justifies its price over cheap storage, it is retention: the ability to say that a class of document lives for a defined period, then is disposed of, automatically and provably.

That matters in two directions. Keeping something too long is a liability in some contexts, because a document you still hold is a document that can be requested. Deleting something too early can be worse. The system’s job is to hold whichever rule applies to you and act on it without depending on anyone’s memory.

What this verdict cannot tell you is what your rule is. Retention periods vary by industry, by jurisdiction, by document class, and often by the specific agreements you have signed with clients, insurers or funders. They also change. Anyone who tells you a single number for how long to keep business records is not describing your business.

The correct sequence is to establish your obligations with a qualified professional, write them down as a list of document classes and periods, and then evaluate software against that list. Buying software first and hoping its defaults match your obligations is how businesses pay for records machinery they never configure. Our manual on running a software trial is worth reading in that order, because a retention requirement is exactly the kind of thing a trial should test rather than assume.

Legal hold and why it sits near the top rung

Legal hold is the mechanism that suspends normal retention for a defined set of documents, so that nothing in scope is deleted while a matter is live, and it records that the suspension happened.

It sits at the advanced rung, or in a records module at an illustrative $9 per user per month, for a reason that is more commercial than technical. It is a low frequency capability with a high consequence, and the businesses that need it will pay for it, which is a textbook description of what gets placed behind an upper tier.

Two practical questions decide whether you need it now. First, does your industry or your contracts create a realistic prospect of a matter in which document preservation is required? Second, if that happened tomorrow, could you actually stop automated deletion across your entire content set, and prove you did? If the answer to the first is yes and to the second is no, the module is not an upsell.

If neither applies, buying legal hold as insurance is buying capability at full price for a scenario you have not assessed. That assessment belongs with a qualified professional who knows your sector, not with a feature comparison, and this verdict makes no claim about what any regime requires.

The audit trail and the question of who saw what

Ordinary file storage records changes. Document management records reads, and that difference is doing most of the work when the category is bought for governance reasons.

A trail that answers “who opened this, from where, and when” is what lets a business respond to an access question, investigate an incident, or demonstrate that confidential material was confined to the people it should have been. A trail that only records edits cannot answer any of those, because the risk being managed is usually a person seeing something rather than changing it.

Three things are worth checking before you assume you have this. How long is the trail retained, and is that period configurable or fixed? Can it be exported in a form somebody outside the business would accept, or does it only exist inside an admin screen? And is it tamper evident, meaning an administrator cannot quietly remove entries?

Those three questions separate a genuine audit capability from a feature bullet, and they are frequently the reason a business ends up on the standard rung rather than entry. It is also a cost you are unlikely to negotiate away, because it tends to be the tier gate itself rather than a line item.

Permissions granularity and the folder that leaks

The most common real world failure in document handling is not a breach. It is that somebody needed one file, was given access to the folder containing it, and now has standing access to two hundred documents nobody thought about.

Folder inheritance is the cause, and it is the default model in ordinary file storage because it is simple. Document management systems generally support permissioning at the document level and increasingly at the metadata or field level, so a case file can be visible while a specific attachment inside it is not, or a record can be visible while a salary field on it is not.

The cost angle is that granularity is a tier gate. Field level control in particular tends to sit at the advanced rung. Whether you need it is a question about your document set rather than your headcount: businesses that hold documents mixing routine and sensitive content in the same object need it, and businesses whose sensitive material lives in separate places usually do not.

The other cost is administrative. Granular permissions require somebody to maintain them, and a permission model nobody maintains degrades into everyone having access within about a year. Budget the ongoing attention, not just the licence.

Search is a cost line, not a feature

Every product in this category has a search box, which is why search rarely gets evaluated properly. What differs is what the box can actually reach.

Filename search reaches names. Full text search reaches the text inside text based files. OCR backed search reaches the text inside scanned images, which is the difference between a hundred thousand scanned pages being an archive and being a filing cabinet in the corner. Metadata search reaches the fields you defined, which is what makes structured retrieval possible at all.

Search quality is a cost line because retrieval time is the actual daily expense of document handling. If four people each spend fifteen minutes a day looking for things they should have found in thirty seconds, that is a real number, and it usually dwarfs the storage overage this verdict spends so long on. Businesses that already log where their hours go can price that directly, and our verdict on time tracking software cost covers what it takes to know. This is also the one benefit in the category that a trial can measure honestly, which is covered further down.

The practical test is not whether search exists. It is whether search finds a document when you remember only one detail about it, which is how people actually search, and whether it finds the scanned version as readily as the typed one.

Migration is the largest single cost in year one

Every worked total in this verdict has the same shape in year one: the one off project is bigger than the software. That is not an accident of the numbers chosen, it is the normal condition of this category.

Migration is priced here at an illustrative $15 per gigabyte of existing content. That rate is deliberately expressed against volume rather than against seats, and it is a proxy for human sorting effort rather than a data transfer charge. Moving bytes is nearly free and largely automated. Deciding what the bytes are is not.

On the eight person practice holding 120 GB, that is $1,800. On the thirty person firm holding 600 GB, it is $9,000. On the three person consultancy holding 40 GB, it is $600. In each case it lands in year one alongside the first year of licences, which is why a twelve month budget built from the pricing page is always wrong.

What actually consumes the money varies. Some of it is vendor or partner services: mapping folder structures to metadata schemas, configuring retention classes, building templates and workflows, testing permissions. Some of it is scripted movement and validation. Most of it, on most projects, is people inside your business answering questions only they can answer.

Deciding what to keep costs more than the software

This deserves its own section because it is the observation that most changes how buyers plan.

A shared drive that grew for eight years without a naming convention contains an unknown number of documents in an unknown number of near duplicate versions, owned by people some of whom have left, in folders whose logic died with the person who created them. Nobody knows what is in it. There is no tool that can tell you, because the question is not technical: it is whether this particular 2019 spreadsheet is the one that matters.

So the migration turns into a series of decisions. Which folders come across as they are? Which get restructured, and into what schema? Which documents get metadata applied, by whom, and how many fields? Which are archived rather than migrated? Which are deleted, and who is willing to sign that off?

Each of those questions costs staff hours from people who have other jobs, and the answer to the last one is usually “nobody”, which is why so many migrations become lift and shift and the new expensive system inherits the old mess along with a bigger storage bill.

The honest budgeting position is this: if your current content is well organised, migration is a project. If it is not, migration is an organisational exercise that happens to involve software, it will cost more than the first year of licences, and pretending otherwise at the planning stage is the single most reliable way to blow the budget. Our verdict on the true cost of business software makes the general version of this point across categories.

Integrations decide whether anyone opens it

A document system that people have to remember to visit is a document system people stop visiting. The integrations that matter are the ones that put it where work already happens.

Email is first. If filing a message and its attachment into the right place takes six clicks and a separate window, it will not happen consistently, and inconsistent filing destroys the retrieval benefit that justified the purchase. Look for a mail integration that files with metadata in one step.

Accounting and finance is second for most small businesses, because supplier invoices, statements and receipts are the highest volume document class in the building. Whether your document system pushes to, pulls from, or duplicates your accounting tool is a real question, and our verdict on accounting software cost is worth reading against it, because some of that storage may already be paid for.

Signature workflow is third, and it is where hidden cost hides. Some systems include it, some resell a partner product per envelope, and some hand you an export. The per envelope model can be significant at volume.

Case, matter, project or ticket systems are fourth. If your business organises work into units, documents want to hang off those units. Our verdict on help desk software cost covers the adjacent case where attachments live inside tickets instead.

Compliance requirements vary and cannot be shopped from a table

Vendors list regime names on comparison pages because buyers search for them, and that listing does very little for you.

Three things are true at once and are frequently conflated. A product can be built so that a compliant configuration is possible. A specific tenant can be configured compliantly. And an organisation can operate compliantly, which requires people following the configuration. Only the first of those is anything a pricing page can speak to, and it is the least important of the three.

Requirements themselves differ by industry, by jurisdiction, sometimes by sub sector, and they change over time. Data residency, permitted processors, breach notification, access logging, retention minimums and disposal evidence are all live questions with different answers depending on who you are and where you operate. Nothing in this verdict states what any of them require.

What this verdict can say about cost is the shape. Compliance requirements usually push you up a rung rather than adding a line item, because the capabilities they need are tier gates. Budget for the rung, not for a module. And establish the requirement with a qualified professional before you evaluate software, because the alternative is discovering after purchase that you bought an advanced tier for a regime that did not apply to you, while missing the one that did.

Users who need access but not a full seat

One of the few genuine pricing levers in this category is that not everyone who touches a document needs to be a licensed user, and most buyers over-provision because it is easier.

Systems typically distinguish several classes. Full users create, edit, approve and manage metadata. Light or view only users read and download, and are commonly cheaper, sometimes at a fraction of a full seat and sometimes bundled at a ratio. External or portal users submit and retrieve documents through a limited interface, and are frequently not billed per head at all. Administrators may or may not consume a licence.

The savings are real. On a thirty person firm where twelve people only ever read, moving those twelve from an advanced seat at $38 to a view only seat at an illustrative third of that saves in the region of $300 a month, which is more than the entire storage overage line in year one.

Ask three questions in writing. What exactly makes a user billable? Is view only access charged, and at what rate? Do external parties, clients or contractors count as users? Vendors answer these clearly when asked directly and vaguely when not asked at all.

Contract length, prepay and the renewal you should expect

Annual prepay discounts in this category behave like the rest of business software, commonly in the region of ten to twenty percent against monthly billing, and the calculation is the usual one: the discount is worth taking if you are confident about the next twelve months and worth declining if you are still validating the product.

What is different here is the storage interaction. If you prepay a year at today’s volume and your volume grows, the overage is usually billed as it accrues rather than being covered by the prepayment, so a prepaid contract does not lock your total. Ask explicitly whether the prepay covers a storage allowance for the year or only the seats. That single answer changes what a multi year commitment is actually worth.

The other question is what a multi year deal fixes. A three year term that fixes the seat rate while leaving the storage and capture rates open has fixed the smallest of the three meters. If you are being asked for a multi year commitment, ask for the overage rates to be fixed too, or at least capped. Our manual on negotiating SaaS pricing covers how to approach that conversation without giving away your position.

Then ask what happens at renewal to a business that has become substantially harder to move because all of its documents are now inside the product. Switching cost is real leverage for the vendor, and the honest response is to establish your export rights in writing at the start, not to hope.

A worked three year total at three business sizes

Three illustrative scenarios, using the bands above throughout: seats as listed, 10 GB per user included storage with 25 cents per gigabyte per month overage, 500 pages per user per month included capture with 4 cents per page overage, migration at $15 per gigabyte of existing content, and bureau scanning at 8 cents a page. Storage overage in each year is charged against the volume held at the start of that year.

A three person consultancy, entry plan, no paper backlog. Licences at $12 are $36 a month and $432 a year, so $1,296 across three years. It holds 40 GB against a 30 GB allowance and adds 18 GB a year, so overage runs $30, then $84, then $138, totalling $252. Capture stays inside the allowance at 700 pages a month, so nothing there. Migration on 40 GB is $600. Year one is $1,062, year two $516, year three $570. Three year total: $2,148, or about $19.89 per user per month.

An eight person professional practice, standard plan, modest backlog. Licences at $22 are $176 a month and $2,112 a year, so $6,336 across three years. It holds 120 GB against an 80 GB allowance and adds 45 GB a year, so overage runs $120, then $255, then $390, totalling $765. Capture at 2,600 pages a month sits inside the 4,000 page allowance. Migration on 120 GB is $1,800, and scanning an 18,000 page backlog at 8 cents is $1,440. Year one is $5,472, year two $2,367, year three $2,502. Three year total: $10,341, or about $35.91 per user per month.

A thirty person capture heavy firm, advanced plan, large backlog. Licences at $38 are $1,140 a month and $13,680 a year, so $41,040 across three years. It holds 600 GB against a 300 GB allowance and adds 260 GB a year, so overage runs $900, then $1,680, then $2,460, totalling $5,040. It processes 26,000 pages a month against a 15,000 page allowance, so 11,000 billable pages at 4 cents is $440 a month, $5,280 a year and $15,840 over three years. Migration on 600 GB is $9,000, and a 240,000 page backlog at 8 cents is $19,200. Year one is $48,060, year two $20,640, year three $21,420. Three year total: $90,120, or about $83.44 per user per month.

Notice the spread. Per user per month across three years, the thirty person firm costs more than twice what the eight person practice costs and over four times what the three person consultancy costs, and almost none of that difference is the seat price. It is paper and history. Load your own volumes into the companion on this page to see where your version lands, and note that the companion prices the three meters and migration but leaves bureau scanning out, because that quote has to come from a bureau.

Signs your document management bill has drifted

Because two of the three meters move on their own, drift here is the norm rather than the exception. Five signatures recur.

Storage overage is now a material line. Compare this month’s storage charge with the one from twenty four months ago. If it has doubled and headcount has not, that is the mechanism in this verdict working exactly as described, and it is a prompt to look at version pruning and disposal rather than to accept it.

Nothing has ever been deleted. If your retention schedule exists on paper but no disposal has actually run, you are paying to store material you decided years ago you would not keep.

Licensed seats exceed people who log in. On an illustrative $38 advanced plan, four dormant seats are about $1,824 a year.

View only people hold full seats. The most common over-provision in the category, and the easiest to fix at renewal.

You are on a rung for a capability you never configured. Records management modules and legal hold are frequently bought during a compliance scare and never switched on. If it is not configured, it is not protecting you and it is not worth the rung.

How to trial a system on retrieval rather than upload

Every product in this category demos beautifully, because uploading a file and seeing it appear is a demo that cannot fail. Structure your trial around the opposite motion.

Load real documents, including ugly ones. A scanned fax, a photographed receipt, a fifty page contract with handwritten amendments and a spreadsheet with six tabs. Clean sample data proves nothing.

Test retrieval from partial memory. Have somebody who did not file the documents find four of them knowing only one detail each: a client name, an approximate date, an amount, a phrase from the body. Time it. This is the daily cost of the system and the only benefit you can measure in a fortnight.

Configure one real retention rule and let it run. Set a short period on a test class and confirm that disposal actually happens and is recorded. A retention feature you have not watched execute is a claim, not a capability.

Break a permission on purpose. Give someone access to a document inside a folder they should not otherwise see, then check what else they can now reach. Folder inheritance surprises are best discovered in a trial.

Run the exit. Export a hundred documents with their metadata and version history and open the result. What comes out tells you what you own, and it is the question nobody asks until they want to leave. Our manual on migrating to new software covers doing that properly in both directions.

What to ask a document management vendor before you sign

Nine questions, in writing, before any signature.

What storage is included, per user or per tenant, and what is the overage rate? Then ask whether the rate is fixed for the term.

Do retained versions count against storage, and can old versions be pruned automatically? This is the main lever on the meter that grows.

What capture allowance is included, how is a page counted, and what happens to failed or reprocessed documents?

What exactly makes a user billable, and are view only and external users charged?

How long is the audit trail retained, is that configurable, can it be exported, and can an administrator alter it?

Can retention rules be applied automatically by document class, and can I see a disposal actually execute in a trial tenant?

What does migration include, what is expected from us, and what is the change order rate if scope grows? Scope always grows here.

What is the annual prepay discount, the renewal uplift, the notice period, and does prepay cover storage or only seats?

What is the export if we leave, does it include metadata and version history, and in what format? A system holding your only copy of your records should never be the only place they exist.

The bottom line

Document management is the one software category where the pricing page tells you the least. The per user bands are real and useful as a floor: around $12 at entry, $22 at standard, $38 at advanced, $55 and up at enterprise, against roughly $8 for the file sync product that keeps getting mistaken for it. But seats are the only meter that stays where you put it.

The two that move are storage, which ratchets upward because nothing in normal operation deletes anything, and capture, which rises with how busy you are. Model them across three years rather than one, because a twelve month budget in this category is systematically too low. In the eight person example, the storage line more than triples on unchanged headcount.

And budget the migration honestly. At an illustrative $15 a gigabyte it is the largest single line in year one for every scenario in this verdict, and the reason is not technical. It is that somebody has to decide what to keep, and that somebody works for you.

Buy the governance if you need the governance. Buy storage if you need storage. The expensive mistake is paying for one while needing the other. Run your own volumes through the companion above and the true-cost calculator before you shortlist, and count your pages before you talk to anybody.

VetLoft is funded by its readers rather than by the vendors it tests, and this verdict is published on that footing: educational material only, and not procurement, records management, data protection or legal advice for your organisation. Every plan rate, storage allowance, overage rate, per page figure, migration estimate and multi year total on this page is an illustrative planning number chosen to show how the three meters in this category behave, not a quotation from any vendor or bureau, and pricing here moves often enough that a figure typical when this was written may read differently by the time you shop. Feature placement across rungs varies by product and shifts between releases, so treat the ladder above as a shape to verify rather than a specification. Record retention periods, disposal obligations, legal hold duties, data residency rules and access logging requirements differ by industry, by jurisdiction and sometimes by the contracts you have signed, they change over time, and nothing on this page states what any of them require of you; establish your own position with a qualified professional before configuring retention, deleting anything, or relying on a vendor’s compliance claim. Confirm allowances, overage rates, user definitions, migration scope, audit trail behaviour, export rights and renewal pricing directly with each vendor in writing before you commit.