Almost every conversation about corporate photo work is a conversation about the shoot. The photographer, the day rate, the lighting, whether the boardroom is big enough. Then the files land, everyone admires them for a week, and the whole thing goes quiet. Eighteen months later the marketing team is quietly pulling images off a stock site again because nobody can find the originals, half the people in them have left, and no one is certain whether the licence covers a paid ad.
That gap between delivery and use is where the money actually goes. This post is about the second half of the job: what a business does with its images once it owns them, and the handful of decisions that determine whether a shoot pays for itself once or fifteen times.
One: it has to be findable by someone who was not at the shoot
The first failure is boring and it is administrative. Files arrive in a transfer link, someone downloads them to a laptop, a subset gets dropped into a shared drive with the folder name the photographer used, and that is the end of the organising.
Two things break. The transfer link expires, usually in thirty days, and the original full-resolution set goes with it. And the person who knows what is in the folders leaves the company.
The fix takes an afternoon. Give every image a name that a stranger can parse, following a pattern you decide once: year, shoot type, subject, variant. Something like 2026-portraits-tan-wl-01 tells you the year, the shoot, the person, that it is a wide frame with layout space, and which take. Keep the untouched originals in one place and never work from them. Keep a working set of web-sized exports somewhere the whole marketing team can reach without asking.
If you run anything more than a hundred or so images, a proper digital asset tool is worth the subscription. If you do not, a disciplined shared drive with a single index spreadsheet does the job. The tool matters much less than the naming rule. When we audit a client's image estate, the ones with a usable library almost never have expensive software. They have one person who insisted on a naming convention and enforced it.
Two: the licence has to cover what you are actually going to do
This is the part that catches Singapore SMEs most often, and it catches them late, usually at the moment someone wants to run a paid campaign.
Photography is licensed, not simply bought, unless the contract says otherwise. What you are paying for is a shoot fee plus a usage grant, and those are separate things even when they arrive as a single number on a quote. The grant can be limited by medium, by territory, by duration, or by all three. A quote that reads "corporate headshots, SGD 3,200" may include web and internal use only, in perpetuity or for two years, with paid media excluded.
None of that is sharp practice. It is the normal structure of the trade. It only becomes a problem when nobody reads it, and then a SGD 6,000 media budget goes live with an image the company does not have the right to run.
Here is how the common arrangements compare in practice.
| Licence type | Typically covers | Usually excluded | Practical risk |
|---|---|---|---|
| Limited web and internal | Website, intranet, decks, email signatures | Paid media, out of home, print advertising | High. Marketing discovers the limit only when a campaign is ready to launch |
| Full commercial, fixed term | All media including paid, for 24 or 36 months | Use after expiry, resale to third parties | Moderate. Easy to forget the expiry date and keep running the asset |
| Full commercial, perpetual | All media, no time limit | Resale, sub-licensing to unrelated brands | Low, but the fee is higher up front |
| Full buyout with copyright assignment | Everything, forever, including resale | Almost nothing | Low. Rarely necessary and often significantly more expensive |
| Editorial or portfolio-only carve-outs | Nothing extra for you | Commercial use entirely | Very high if misread. Uncommon in corporate work but appears in collaborations |
The advice we give clients is unglamorous: buy full commercial perpetual for anything with a person's face in it that you intend to use on the website or in advertising, and accept the higher fee. Re-licensing later almost always costs more than buying the wider grant at the outset, because by then you have no leverage and no alternative image.
Two related points. First, model releases. If a photo of an employee runs in a paid ad, you want a signed release, and you want it to survive their resignation. Get it on the shoot day with a clipboard, not by chasing email six months later. Second, the photographer's own portfolio rights. Most will want to show the work. That is usually fine and occasionally is not, particularly for anything showing unreleased products or a client's premises. Say so in writing before, not after.
Three: it has to hold up across channels that crop differently
A corporate image is rarely used at the shape it was shot. The website hero is wide. The LinkedIn company post is square or four by five. The Instagram story is a tall nine by sixteen. The proposal cover has a text block over the left third. The email header is a letterbox. The staff directory is a tight circle.
If the library was shot only as centred head-and-shoulders portraits, every one of those placements is a compromise, and the compromise shows. The head gets chopped for the letterbox, or it floats absurdly small in the story, or the text overlay sits on someone's face.
Two habits fix this. At the shoot, ask for deliberate negative space on at least one variant of every important subject: a frame where the person sits in the right third with clean, uncluttered area to the left. At the edit, export a set of pre-cropped variants for your actual placements rather than leaving every colleague to crop by hand. Five crops per key image, done once, prevents a hundred bad crops later.
This matters more for social than for anything else, because social is where the volume of use happens. Teams running an active LinkedIn presence burn through images faster than any other channel, and the ones that look coherent are almost always working from a pre-cropped set rather than cropping ad hoc. The same applies to whatever your content team is publishing week to week; a good library is the difference between a post that looks like the brand and a post that looks like a template.
Four: it has to have a refresh rhythm, not a panic cycle
Most companies shoot on panic. Something triggers it, a rebrand, a website project, a big pitch, and suddenly everything must be reshot at once. Then nothing happens for four years while the library decays.
Decay is predictable. People leave. Offices move. Branding shifts. Fashion moves enough that a 2021 set looks dated by 2026 in ways that are hard to name but easy to feel.
A rhythm beats a panic. What works, in our experience, is a tiered cycle: leadership and the small number of genuinely front-facing individuals on a short cycle, the broad staff set on a longer one, environmental and workplace imagery refreshed when the space or the operation changes, and new starters captured in a standing quarterly or half-yearly top-up session.
| Asset type | Practical refresh interval | Trigger to refresh early |
|---|---|---|
| Leadership portraits | 18 to 24 months | Role change, major press cycle, visible change in appearance |
| General staff portraits | 3 years | Brand refresh, background or wardrobe standard changes |
| New starter top-ups | Quarterly or half-yearly | Any client-facing hire |
| Office and environmental | When the space changes | Office move, fit-out, new equipment, headcount jump |
| Product and in-use imagery | With each product cycle | New SKU, packaging change, new use case |
| Event and culture imagery | Rolling, captured as events happen | No refresh needed, just continuous capture |
The top-up session is the one nobody plans and everybody needs. A two hour block, twice a year, with the same lighting setup rebuilt from the original notes, catches every new joiner and costs a fraction of a fresh full-day shoot. Ask your photographer to document the lighting diagram, the lens, the distance and the background at the main shoot specifically so this is possible. Good production teams do this as standard, and if you are commissioning across both stills and motion it is worth checking that the same discipline applies to your video assets too.
Five: it has to actually replace stock, and here is where we disagree with the orthodoxy
The standard line, which we have written versions of ourselves, is that stock photography is the enemy and original imagery always wins. A common mistake we see is companies taking that literally, spending heavily on a full custom library, and then producing work that performs no better than what they had.
In reality, the enemy is not stock. The enemy is imagery that carries no information about your business. A generic original photo of your own staff smiling around a laptop in a neutral meeting room communicates exactly as much as a generic stock photo of strangers smiling around a laptop, which is nothing, and it cost you considerably more.
The images that outperform are the ones that could only be from your company. The specific machine on your floor. The actual queue outside your shop at 12:15pm. The whiteboard with your real workflow on it. The thing your engineer does with their hands that nobody outside the industry has seen. Those are worth commissioning, and they are worth commissioning at the expense of the fifteenth variation of a handshake.
We have seen clients get better commercial results from a small, sharp library of forty specific images than from a sprawling set of eight hundred pleasant generic ones. Most custom photography budgets are spent reproducing stock conventions with more expensive models.
So when you brief a shoot, ask a harder question than "what images do we need". Ask: what does this business look like that no competitor's business looks like, and how do we photograph that. If the honest answer for a given shot is "nothing distinctive", consider whether a well-chosen licensed image plus your own brand treatment is a smarter use of the budget, and spend the savings on the shots that are yours. That reasoning applies just as much to product imagery, where the difference between an in-context shot and a catalogue cut-out is usually the difference between a page that sells and one that does not.
The leaver problem, and how to handle it without a purge
There is one recurring administrative issue that deserves its own treatment, because it catches every company eventually and almost nobody has a policy for it. People leave, and their faces are in your marketing.
There is no single correct answer, but there is a sensible default. Individual portraits of a departed employee should come down from anywhere that implies current employment: the team page, the staff directory, a bio on a services page, an email signature. That is straightforward and mostly automatic if the library is properly indexed, because you can find every use in minutes rather than discovering one on a PDF eighteen months later.
Group and environmental images are different, and companies over-correct here. A wide shot of a busy office in 2024 that happens to include four people who have since left is not a misrepresentation. It is a photograph of an office in 2024. Purging it costs you an asset and gains you nothing. The test we suggest is whether the image makes a claim about a specific named individual's current role. If it does not, keep it.
Where it does matter is advertising. If an image is running in paid media and the person in it has left, check the release. Most releases survive employment, but not all do, and a few carry a term limit. This is another reason to file releases with the images rather than in an HR folder nobody in marketing can reach.
Build the check into offboarding. One line on the leaver checklist, image review, owned by whoever governs the library, takes ten minutes and prevents the awkward discovery of a former employee fronting a campaign six months after a difficult exit.
Six: it has to be governed by someone
The last requirement is the least technical. Somebody has to own the library.
Owning it means deciding what goes in, retiring what goes out, holding the licence records, running the top-up sessions, and telling a colleague no when they want to use the 2022 office shot that still has the old logo on the wall. It is perhaps two hours a month. It is also the single variable that best predicts whether a company's imagery looks coherent, because every other rule in this post survives or dies depending on whether one named person is enforcing it.
When ownership is ambiguous, the library reverts to whatever is fastest, and whatever is fastest is always the first image in the shared drive or the first result on a stock site.
Governance also means writing down the small number of rules that keep the set coherent, and keeping that document to a single page. Which backgrounds are approved. Whether images are cropped square or four by five for social. What the retouching standard is, meaning what gets removed and what does not, because a policy of removing blemishes but not lines is a decision someone has to make once rather than every time. Which images are approved for external use and which are internal only. Where the originals live and who can reach them. Anyone in the company should be able to read that page in three minutes and use the library correctly without asking.
The last piece of governance is a calendar entry. Once a quarter, thirty minutes, open the library and look at it as a stranger would. Are the leadership shots still current. Has anyone in the front-facing set changed roles. Are there images running on the site with a background that no longer exists because the office was refitted. Is the top-up session booked. Almost nothing on that list is urgent in any given quarter, which is precisely why it never happens without a recurring invitation.
One Singapore professional services firm and a very expensive folder
One Singapore professional services firm we worked with had commissioned a substantial shoot roughly three years before we met them. Around eleven hundred images, portraits of the whole firm, partner headshots, office interiors, the lot. By the time we started on their site and content, they were using nine of them.
The reasons were mundane and cumulative. The delivery link had expired and only the web-sized exports survived, which ruled out print. The naming was the photographer's camera output, so finding a specific partner meant opening folders until you recognised a face. The licence was web and internal only, which the marketing lead discovered when legal reviewed a planned campaign. And around a quarter of the people in the portraits had left.
We did not recommend a reshoot first. We recommended an audit. Recovering the originals from the photographer, who still had them and was perfectly willing to help, cost a small archive retrieval fee. Extending the licence to full commercial perpetual cost meaningfully less than a new shoot. Renaming and indexing the surviving set took an assistant a day and a half. That got them from nine usable images to a few hundred.
Only then did we scope a shoot, and it was a small one: partner refreshes, the new joiners, and a half day of specific environmental work in their client meeting space and their research room. Total spend across the recovery and the top-up was well under half what a full reshoot had been quoted at, and the library that came out of it was more usable than the original had ever been.
The lesson we took from it is that the most common photography problem in Singapore SMEs is not a shortage of images. It is a shortage of access to the images already paid for.
Field Notes
When we audit a client's existing image assets before starting brand or content work, the number that keeps repeating is roughly 70 percent. That is the share of commissioned corporate images we typically find sitting unused, unfindable or unusable for the channel the client now needs them for. Expired transfer links and missing originals account for a large chunk of it. Licence limits and staff turnover account for most of the rest.
The corresponding number on the other side is smaller and more encouraging. In the majority of those audits, the original photographer still holds the master files and will release them for a modest retrieval or re-licensing fee. So before anybody signs off a reshoot, send that email. We have recovered entire libraries for a fraction of a shoot day, and it takes one afternoon to find out.
Where to go from here
If you already have a corporate photo set, run the six checks above in order: findable, licensed, croppable, on a refresh rhythm, specific, and owned by a named person. Most companies fail at least three and can fix two of them in a week without spending anything.
If you are commissioning fresh, build the post-shoot requirements into the brief. Ask for full commercial perpetual usage. Ask for the lighting notes. Ask for pre-cropped variants for your real placements. Ask for signed releases on the day. None of these are unusual requests and all of them are cheaper to ask for before the invoice than after.
Our corporate photography work is structured around that second half of the job, because a library nobody uses is not an asset, it is a cost. If you want a view on what you already have, or you want the shoot scoped so it survives contact with your actual marketing calendar, tell us what you are working with and we will tell you honestly whether you need a camera or an afternoon of housekeeping.
If the images are feeding a wider content programme, our guide to choosing a content marketing agency for SEO covers how the two should be planned together.
Natasha Tan is the founder of Digital Marketing Singapore, a full-service SEO and digital marketing agency based in Singapore. With hands-on experience across SEO, paid media, and content strategy, she works directly with Singapore businesses to build organic visibility and generate consistent leads. Natasha specialises in the Singapore market — including local search behaviour, PDPA compliance, and government grant navigation for SMEs.
