Enterprise Digital Asset Management

Tag: Procurement

Total cost of ownership across all six lines, commercial models, and the requirements that actually separate vendors.

  • The enterprise DAM requirements checklist, without the vendor bingo

    The enterprise DAM requirements checklist, without the vendor bingo

    Somebody circulates the requirements matrix. It has 240 rows, gathered from every stakeholder who was asked, and every row is genuinely something someone wants. It goes to five vendors. All five score above ninety percent. The decision gets made on which demo felt better.

    Short answer: a requirement only earns its place if a credible vendor could fail it. Most of a standard DAM matrix is table stakes that every serious platform satisfies, which means it contributes nothing except weight. Cut to the thirty or so requirements that actually separate the field, weight them by what you do rather than by who asked, and test the top ten rather than accepting a written answer.

    A requirements scoring chart with five dominant cyan bars at the top, a thin middle group, and seven amber hairlines at the bottom bracketed as noise

    The test for whether a requirement belongs

    One question: could a vendor you would seriously consider fail this?

    “Supports metadata” fails the test. Everyone supports metadata. “Enforces field-level validation on every API write path, not only in the upload interface” passes, because plenty of platforms do not.

    Apply that question to your matrix and it will lose sixty to eighty percent of its rows. What remains is a document that discriminates, which is the entire purpose.

    Two secondary tests worth applying to the survivors:

    Is it observable? If you cannot design a test that demonstrates it, you will accept a written assurance, and written assurances are all identical.

    Does it follow from something you actually do? Every requirement should trace to a workflow, an obligation, or a system in your estate. Requirements with no traceable origin came from a template.

    The requirements that discriminate

    Thirty-two, in six groups. Ordered so the highest-signal group comes first.

    Architecture and API

    • Is the vendor’s own interface built on the public API? Check the browser network tab during the demo. Private endpoints mean a second-class API forever.
    • Can you define fields, types and controlled lists over the API, and are they enforced on every write path?
    • Is search a queryable expression over your own fields, with boolean composition, sorting and pagination past a thousand results?
    • How many SDK languages, and are they first-party and current?
    • Are derivatives produced on request from a URL, or pre-generated and stored?
    • Are delivery URLs versioned and immutable, so a replaced master produces a new address rather than requiring a purge?
    • What events are emitted, and can they be subscribed to?

    The reasoning behind this group is in headless DAM, and it is the group that predicts your integration cost over five years.

    Metadata and taxonomy

    • Are fields typed, with real dates and controlled lists, or is everything a string?
    • Can a field be conditionally required based on the value of another field?
    • Are controlled vocabularies managed as first-class objects, with synonyms, preferred terms and deprecation, in the sense ANSI/NISO Z39.19 means it?
    • Can a term be retired and existing assets remapped without deleting data?
    • Is automated analysis available on ingest, and can its output be directed to specific fields rather than a generic tag bag?
    • Can you export the full schema and all values, in a documented format, without vendor assistance?

    That last one is an exit requirement and it is the one vendors are least comfortable with, which is exactly why it belongs. The field model these support is in the metadata schema an enterprise actually needs.

    Rights and lifecycle

    • Are licence terms structured fields with indexed dates, rather than a document attachment?
    • Does expiry change asset state automatically, or produce a report?
    • Does expiry emit an event that downstream systems can act on?
    • Are lifecycle states configurable, and is state exposed as a search facet?
    • Is version history automatic, with old versions resolvable?
    • Can you identify every asset depicting a given person, for privacy request execution?

    Details in rights and expiry as first-class asset data and the asset lifecycle.

    A checklist sheet with confident cyan ticks at the top, amber handwritten question marks in the middle, and the last third folded out of sight

    Security and identity

    • SSO via SAML or OpenID Connect, with group membership driving roles?
    • Does deprovisioning happen automatically when the directory record is terminated?
    • Can external accounts be time-bound with a required internal sponsor?
    • Does the audit log record reads and downloads, not only writes?
    • Is the log exportable to your SIEM, and immutable to administrators?
    • Can delivery be restricted so an unlisted URL is not an open one?

    The one to test rather than ask about is deprovisioning. Take a real recent leaver and check. Full set in access control, SSO and audit trails.

    Delivery

    • Automatic format selection per browser, from one URL, covering AVIF and WebP?
    • Automatic quality, tuned per image rather than a global setting?
    • Content-aware cropping that keeps the subject?
    • Arbitrary width on request, for responsive images at widths you cannot know in advance?
    • What cache headers are set by default on derivatives?
    • Does the delivery API return descriptive metadata, including alt text, alongside the asset?

    That last one is missing far more often than you would expect and it is why so many properly-described libraries produce pages with empty alt attributes. Covered in delivery is part of your DAM.

    Commercial and operational

    • Is pricing published, and what is the overage rate in writing?
    • Is renewal uplift capped in the contract?
    • What is included in implementation, specifically, and what is time and materials?
    • What is the documented exit path: full asset and metadata export, in what format, at what cost?

    What to leave out

    Four categories that consume evaluation effort and produce no separation.

    Anything every vendor does. Upload, folders, search, preview, download, basic permissions, collections. Table stakes. Listing them makes the document longer and the decision no clearer.

    Interface preferences. “Intuitive user interface” is not a requirement, it is a judgement, and it will be made in the demo regardless of what the matrix says. Handle it with a structured usability test on real tasks, scored separately.

    Requirements derived from your current system’s quirks. “Must support our seven-level folder hierarchy” encodes a problem rather than a need. Ask what the hierarchy is for, and most of the time the answer is faceted search, as in designing a taxonomy people actually use.

    Features nobody has asked for and nobody will configure. Every unused capability is licence cost and configuration surface. If no named person will own a feature, it is not a requirement.

    A balance beam tilted left with three large cyan blocks outweighing twelve tiny amber ones

    Weight before you score

    Unweighted matrices are how a platform wins on features nobody will use.

    Do it in one session with the people who will actually run the system:

    1. Sort the surviving requirements into three tiers. Deal-breaker, important, nice. Cap deal-breakers at eight. If everything is a deal-breaker, nothing is.
    2. Weight by frequency of use. A requirement exercised daily outranks one exercised at audit time, unless the audit one carries legal exposure.
    3. Have each stakeholder group weight independently, then compare. Where marketing and engineering disagree sharply, you have found the real architecture decision, and it is worth an hour of argument now rather than a year of friction later.

    Test the top ten, do not ask about them

    For your eight to ten highest-weighted requirements, design a test and run it in a shared session with each vendor. Half a day per vendor.

    The tests that consistently produce the most information:

    • Upload one asset, request it at five widths in three formats by editing the URL, and time cold and warm responses.
    • Define a new metadata field with a controlled list and a validation rule, then attempt an invalid write over the API and watch it be rejected.
    • Replace a master and confirm existing references pick up the new version with no manual purge.
    • Set a licence expiry in the past and observe what happens without human intervention.
    • Ask for every download of a named asset in the last six months, and time how long it takes to produce.

    Written responses are drafted by people who are good at drafting written responses. A live test is a fact.

    Then take the results into the platform comparison to see which archetype your answers point at, cost it properly with what enterprise DAM costs, and translate it into a business case. If a traditional DAM turns out not to be what you need at all, alternatives to a traditional enterprise DAM is the honest version of that conversation, and the grounding for the whole exercise is in what enterprise DAM actually is.

  • The DAM business case your CFO will actually sign

    The DAM business case your CFO will actually sign

    The business case says the DAM will save 400,000 a year in productivity. The CFO reads it, asks how many people will therefore leave the organisation, and the answer is none. The saving is discounted to zero and the case now rests entirely on the storage line, which is 12,000.

    Short answer: finance discounts benefits it cannot verify, and productivity claims with no headcount consequence are the most discountable claims there are. Build the case out of avoided spend you can point at on an invoice, plus risk you can quantify from incidents that have already happened to you, and treat productivity as a supporting argument rather than the main one. A smaller case built from verifiable numbers gets signed. A larger one built from estimates does not.

    A waterfall chart building from a baseline bar through four rising cyan benefit increments and two falling amber cost decrements to a net bar

    The four benefit categories, ranked by how much finance believes them

    1. Avoided external spend. Highly credible. Money currently leaving the organisation that will stop. Duplicate stock photography purchases, agency charges for re-supplying assets they already delivered, reshoot costs for material that exists but cannot be found, a separate CDN or delivery contract the DAM replaces.

    This is the strongest category because every item is an invoice. You can name the vendor and the amount. Nobody argues with an invoice.

    2. Avoided risk. Credible if you have incidents. Licence overrun settlements, takedown costs, rework after publishing an unapproved asset, audit findings and their remediation cost.

    The trick here is to use your incidents, not industry averages. An industry average is a benchmark; your own settlement from eighteen months ago is evidence. If you genuinely have had no incidents, do not manufacture the category, use it qualitatively instead.

    3. Avoided internal cost. Moderately credible. Contractor days spent on manual asset handling, agency retainer hours logged against asset supply, storage and infrastructure you will decommission.

    Credible to the extent it is contracted spend rather than salaried time. A contractor day is a real number. A salaried hour is not, unless the headcount actually changes.

    4. Productivity. Weakly credible on its own. Time saved searching, time saved on rework, faster campaign turnaround.

    All genuinely real, all correctly discounted by finance, because saved minutes distributed across sixty people do not reduce any budget line. Include it, quantify it honestly, and do not build the case on it.

    How to make the productivity number defensible anyway

    If you want the productivity benefit to survive scrutiny, it has to be measured rather than assumed. Three ways to do that, in ascending order of effort and credibility.

    Instrument the current state. Most DAM cases quote a search-time figure from a vendor’s white paper. Do not. Run a two-week diary study with twenty people recording time spent looking for assets and time spent recreating things that already existed. Twenty people, two weeks, a shared spreadsheet. The number will be lower than the vendor’s and infinitely more defensible.

    Tie it to a throughput commitment. “The team will deliver eighteen campaigns next year instead of fourteen, with the same headcount.” That converts productivity into output, which finance can verify after the fact. It is also a commitment, which is why people avoid it and why it works.

    Convert to avoided hiring. If the marketing operations team was going to grow by one to cope with volume, and now will not, that is a real number with a real budget line. This is by far the strongest form of the productivity argument and it is available more often than people realise.

    A cumulative cash flow curve dipping amber below zero, crossing to cyan above it about a third of the way along, with a marker at the crossing

    The cost side, honestly

    A case that understates costs gets one signature and then loses credibility for the next three years. Use the full stack from what enterprise DAM costs: licence, implementation, migration, storage, delivery, and people.

    Two lines to be specific about, because reviewers look for them and their absence is a tell:

    The ongoing people cost. A named role, or a named fraction of one, funded. If the case does not contain this, an experienced reviewer will assume the programme has no owner and price the risk accordingly. They will be right.

    Year two and beyond. Renewal uplift, growth in volume, and the second wave of integrations that always follows a successful first wave. A five-year model with realistic growth beats a one-year model with a good first-year discount, and total cost of ownership is the framing your finance team already uses.

    Where the real savings hide

    Two lines that are often larger than the headline productivity claim and almost never appear in the case.

    Delivery and bandwidth. If your organisation currently serves original-resolution images to web and mobile, the saving from automatic format and quality selection is a genuine, measurable infrastructure number. Serving modern formats to browsers that support them typically cuts image payload substantially, and the Cloudinary image optimization documentation covers the automatic format and quality mechanism, with WebP and AVIF support now broad enough that this is not a hypothetical.

    Get your current monthly image bytes from your CDN, model the reduction, and price it at your actual rate. It is one afternoon of work and it produces a number from an existing invoice, which puts it in benefit category one rather than category four.

    There is a revenue side to this too, which is worth stating carefully rather than overclaiming: image weight is a primary input to Largest Contentful Paint, and page performance affects conversion. Do not put a conversion uplift number in the case unless you can run the test. Do mention the mechanism.

    Storage multiplication. If every crop and format is a stored file, you are paying for the master times some multiple, forever. Count the distinct masters in your current library, count the total files, and the ratio is your multiplier. In a derived-rendition model it goes to one. That arithmetic is covered from the architecture side in single source of truth is an architecture.

    Extreme macro of a watch escape wheel and pallet fork under a raking cyan light, one tooth discoloured amber

    Benefits to leave out

    Three that weaken a case by being in it.

    “Improved brand consistency.” Real, important, and unquantifiable. It reads as filler in a financial document. Put it in the strategic narrative, not the model.

    “Better collaboration.” Same problem, worse. Finance has seen this phrase in every business case for twenty years and it carries no information.

    Vendor-supplied industry benchmarks. “Organisations typically see a 30% reduction in…” Any experienced reviewer knows where that came from. One statistic sourced from the party selling you the thing damages the credibility of every other number in the document.

    Structure the document like this

    Six sections, ten pages maximum. Longer documents are read less carefully, not more.

    1. The problem, with evidence from your organisation. One incident, one measured number, one quotable line from the diary study.
    2. What is being proposed, in one paragraph, without vendor names.
    3. The cost model, five years, all six lines, with assumptions listed separately so they can be challenged individually.
    4. The benefit model, four categories, each labelled with its confidence, and each traceable to a source.
    5. Payback and sensitivity. When does it turn positive, and what happens if benefits land at half. A case that survives its own downside test is a case that gets signed.
    6. What happens if we do nothing. Usually the strongest section, and usually the shortest. Rights exposure that is already running, migration cost that grows every year you wait, and the specific thing that broke last quarter.

    Then commit to measuring it

    The final paragraph should name the metrics you will report at six and twelve months, with the baseline stated now. Duplicate ratio, search zero-result rate, image bytes served, licensed assets with a valid expiry date on file.

    This is unusual enough that it materially improves your chances of approval, and it also means the benefits get realised rather than assumed. Programmes that do not measure post-implementation are the ones that quietly stop being used, which is the pattern described in why DAM rollouts fail.

    Build the requirement set that generates your quotes from the requirements checklist, and if you are still establishing what the category does at all, start at what enterprise DAM actually is.

  • What enterprise DAM costs, line by line

    What enterprise DAM costs, line by line

    A vendor quotes 180,000 a year. Finance approves it. Two years later the programme has cost roughly double that and nobody can point at the moment it went wrong, because nothing went wrong. The quote was accurate. It was just answering a narrower question than anyone thought.

    Short answer: budget the licence at somewhere between forty and sixty percent of your actual five-year cost. The rest is integration, migration, storage and delivery overage, and the people who run it. The people line is usually the largest single item and it is almost never in the business case, which is why so many DAM programmes are technically successful and commercially disappointing.

    A horizontal stacked bar chart of five cost categories, each split into a cyan quoted portion and an amber not-quoted portion that dominates integration and people

    The full cost stack

    Six lines. Build the model with all of them or you are not comparing vendors, you are comparing quotes.

    1. Platform licence. The number on the proposal. Usually banded by seats, storage, or some composite. Ask specifically what happens on renewal, because year-one discounting is standard and the uplift in year two is where the real price lives.

    2. Implementation and integration. Configuration, schema build, permission model, and connecting the DAM to everything that consumes from it. Vendors will quote the first three and not the fourth, because the fourth depends on your estate. In practice this runs from thirty to a hundred percent of first-year licence, and the variable is how many systems and how good their APIs are. Integration patterns is where that number actually gets set.

    3. Migration. Moving and describing the existing library. Cost scales with how bad your current metadata is, not with how many terabytes you have. Ten terabytes of well-described material is a cheap migration. Two terabytes of IMG_4471_final_v3.jpg is not, and the reasons are in migrating a million assets.

    4. Storage. Usually straightforward and usually small. The trap is not the rate, it is the multiplier. If your platform stores every rendition as a separate file, you are paying for the master plus every crop, every format and every size, forever. A derived-rendition model stores the master only.

    5. Delivery and bandwidth. The line most often underestimated, because nobody models it until they are live. If your DAM also serves your public web traffic, this is a real number that scales with your marketing success. If it does not serve public traffic, you are paying for a separate CDN somewhere else and should count that here too.

    6. People. A DAM manager or information governance lead, plus a share of a developer, plus the taxonomy and cataloguing time that has to come from somewhere. At mid enterprise scale this is typically 0.5 to 1.5 full-time equivalents on an ongoing basis. Fully loaded, that frequently exceeds the licence.

    A five-year cumulative cost chart with a smooth cyan platform line crossed at year two by a steeper amber operating line

    The three commercial models

    Vendors price on one of three bases and they are not equivalent. Which one suits you depends almost entirely on your shape.

    Per seat. Predictable, easy to approve, and it prices the thing you least want to restrict. Every seat you do not buy is a person who emails a colleague for the file instead, which is exactly the behaviour you are paying to eliminate. Watch for the distinction between full users and consumer or read-only users, because the ratio is usually ten to one and the pricing difference is where the negotiation lives.

    Per volume. Storage, bandwidth, transformations, API calls, or a composite unit. Scales with usage rather than headcount, which means adoption success shows up as a cost increase. That is not a reason to avoid it, but it does mean you need a forecast and a monitoring habit, not just a budget line.

    Enterprise agreement. Custom, negotiated, usually unbanded. Fine, and often the right answer at scale, but it removes your ability to benchmark. Get at least one banded quote from a comparable vendor so you know what you are being asked to pay a premium for.

    Cloudinary is a useful reference point on the volume model because its pricing is published rather than gated, which is unusual in this category. The published tiers run from a free plan at 25 credits per month, through Plus at 99 dollars per month for 225 credits, to Advanced at 249 dollars per month for 600 credits, with enterprise agreements custom. The credit is a composite unit, and the definition is worth reading before you model anything: one credit covers 1,000 transformations, or 1 GB of managed storage, or 1 GB of delivered bandwidth, and you spend across those as you use them.

    That composite structure is the part to model carefully. It is genuinely flexible, and it means your bill responds to the mix of what you do rather than to a single dimension. It also means a rough forecast needs three inputs rather than one.

    Two details that materially change the arithmetic: transformations and bandwidth are measured over a rolling thirty-day window while storage reflects your current total, and re-delivery of an already-generated derivative does not count as a new transformation. The second one matters a lot. It means a heavily-cached public site costs far less in transformation terms than a naive calculation suggests.

    Which lines do vendors leave out?

    Reliably four, and none of it is dishonest. They are quoting their product, and these are your costs.

    • Your integration effort. They quote their connector. They do not quote the six weeks your team spends on the other side of it.
    • Delivery at production volume. Pilot traffic is not production traffic. Ask for the overage rate in writing, not just the included allowance.
    • Ongoing cataloguing. Somebody describes the assets. That somebody costs money whether they sit in your team or an agency.
    • The second-year uplift. Ask for a capped renewal in the contract. This single clause is worth more than most of the feature negotiation.
    A fanfold invoice cascading across deep navy under raking light, most lines grey, three circled in amber, one page edge tabbed in cyan

    Where the money actually goes wrong

    Not in the negotiation. In three specific decisions made after signature.

    Storing renditions instead of deriving them. Every stored variant is storage you pay for, a sync problem you own, and a migration item later. If the platform can produce a crop from a URL parameter, use that, and the Cloudinary resizing documentation shows the shape of it. This is the difference between a storage line that grows with your library and one that grows with your library times fifteen.

    Buying seats for people who only download. Most vendors have a cheaper consumer tier or a public collection mechanism. Full seats for the reseller network is the most common single overspend in this category.

    Serving originals. If you deliver the master file to a web page because nobody set up automatic format and quality, you pay for the bandwidth twice: once at the DAM and once in your conversion rate. Automatic format and quality selection is a one-parameter change with a large bandwidth effect, and it is covered from the performance side in delivery is part of your DAM.

    How do you build a number you can defend?

    Do it in this order. It takes about a week and it survives contact with finance.

    1. Count the estate. Masters, renditions, annual growth, and current storage. Not folder size, actual distinct masters. The ratio between the two is itself a finding.
    2. Forecast delivery. Monthly image requests across all channels, plus average delivered size. If you do not have this, your CDN or web analytics does.
    3. List the integrations. Every system that will read from or write to the DAM, with a rough effort estimate each. This is the line that moves most between vendors.
    4. Staff it honestly. Name the roles and the fractions. If nobody is named, the programme has no owner and the adoption risk is your largest unpriced exposure.
    5. Model five years, not one. Total cost of ownership over the realistic life of the decision. Year one flatters the incumbent option and every platform looks cheap in a pilot.

    Then take that model into the business case, where the benefit side gets the same treatment. And when you are comparing the resulting numbers between vendors, the platform comparison is organised by who each one is actually for, which is more useful than a feature grid once the prices are within twenty percent of each other.

    If you are building the requirement list that generates these quotes in the first place, start from the requirements checklist and cut it before you send it. Every requirement you cannot justify is a line item somebody will price. The foundation for all of it is in what enterprise DAM actually is.