Material is rarely the problem. On the foreman's work phone, in a folder on the office computer and across three years of job files, almost every landscaping business holds hundreds of images. What is missing is order. A portfolio gallery made up of a wall of near-identical photographs answers none of the questions people arrive with: do they do paving? Would they travel to me? Is this from this year or from a time when entirely different people worked here? And how large was the area actually shown? This article describes how a gallery has to be built so that it answers those questions — which fields each project needs, how filters by trade, location and year follow from them, how few projects are enough at the start, and why six carefully documented jobs achieve more than thirty unlabelled shots. It is not about photography but about structure: even existing phone images make a usable gallery, provided the details beside them are right.
Key takeaways
- A portfolio gallery is a selection tool, not a photo album: it should answer the three questions visitors arrive with — which trade, which location, which year — before anyone reaches for the phone.
- Three filters are enough: trade, location or region, and year. They only work if these values are maintained as separate fields per project rather than buried in the running text of a caption.
- Every project card carries six details: trade, location, year, the task in two or three sentences, the scope with area, material and working days, plus four images from starting point to result.
- Six fully documented projects across two or three trades are enough to launch; thirty unlabelled photographs perform worse, because no filter can work and nobody can judge the order of magnitude.
- One new project a month from ongoing work keeps the gallery current. The client's approval belongs to the handover appointment, not to a follow-up phone call the following winter.
What the gallery is for — and what it is not
Anyone opening the references on a landscaping company's website usually has a specific intention already. The terrace needs replacing, the driveway has settled, the garden after the house build is still bare soil. That person is not looking for inspiration but for confirmation: can this business do exactly what I need, and have they done it somewhere within reach? A gallery that fails to answer this costs nothing in money, but it also achieves nothing.
From that follows the central sentence on the subject: the portfolio gallery is a selection tool, not a photo album. It is not admired, it is used. The difference shows immediately in the structure. An album presents images in the order they were taken and leaves the interpretation to the viewer. A selection tool offers filters, shows the hard facts for every entry and leads to an action at the end — the phone call or the completed form.
The gallery also has a second and a third audience that is easily overlooked. Applicants look at what they would be working on before they take a job advert seriously; property managers and municipal clients check whether comparable object sizes appear. Both groups read the same fields, but with a different intention. That is precisely why it pays to keep the details per project factual rather than wrapping them in advertising language: scope, material and working days are the interesting figures for all three groups.
Trade
Does this business regularly do what I need? A single image of a pond does not answer that; three pond projects across two years do.
Proximity
Would they travel to me? Locations in the gallery are the most credible evidence of a service area, because they attach to work actually carried out.
Order of magnitude
Does my job fit the range? Area in square metres and working days pre-sort enquiries that are too small or too large without anyone having to phone.
Currency
Is the business still active and in this line-up? A year per project answers that faster than any sentence about long-standing experience.
Way of working
How is the work done? One image of the sub-base says more about execution than a photograph of the finished surface, on which every business looks alike.
Next step
What happens now? Every project needs a route to the enquiry form that carries the context along — otherwise the visit ends at the last image.
Three filters are enough: trade, location, year
Filters multiply quickly. Anyone who starts sorting soon arrives at ten categories: material, style, client type, season, development area, maintenance effort. The result is a filter bar that needs explaining in its own right. In practice three dimensions carry almost the entire selection: the trade, the location and the year. Everything else can appear as a detail on the project card without becoming a filter of its own.
The trade filter matters most, and it is also the one most often cut wrongly. A business such as Viktor Waldbauer Garten- und Landschaftsbau in Nordstemmen works on sand-set slabs, terraces, cisterns and complete garden installations — that is four filter values, not one called "gardening". The rule is simple: a filter value earns its place if you offer it as a service in its own right and can show at least two projects for it. A value with a single entry reads more like an oversight.
With locations, restraint is advisable. Not every client wants their home town printed next to pictures of their garden, and for private gardens a check is due in any case. Where the exact place is not approved, the district or region works just as well — for the question "would they travel to me?" the distance is what counts. What matters is that the spelling stays identical across all projects, otherwise the filter splits into two values for the same place.
| Detail | What the visitor settles with it | Where the value comes from |
|---|---|---|
| Trade (filter) | Whether the business carries out this work regularly | Mandatory field, drawn from your own list of services |
| Location or region (filter) | Whether the journey fits the service area | Town if approved, otherwise district or region |
| Year (filter) | Whether the work is current or from a different era | Year of handover, not the creation date of the image file |
| Area and working days | Whether their own job fits the order of magnitude | Measurements and site diary, taken over rounded |
| Material | Whether look and price bracket match their own wish | Delivery note or quote item, without manufacturer names |
| Client type | Private garden, property manager or public contract | Follows from the job, without naming anyone |
What a project card has to contain
The project card is the smallest unit of the gallery: images plus labelling, in a fixed structure, identical for every project. That uniformity is not a matter of appearance but the precondition for filters working at all. A field left empty on four of six projects is unusable as a filter — the selection would then hide projects that actually fit.
Six details carry the card. Trade, location and year are the filter fields. Added to them are the task in two or three sentences, the scope with area, material and working days, and the images. Nothing more is needed, and in the long run nothing more gets maintained. Anyone adding twelve projects in February notices quickly that every additional mandatory field lowers the chance of the catch-up ever being finished.
What matters is that the values are stored as separate fields and not as one paragraph of running text under the picture. Only then can the same projects appear in several places on the website: filtered in the portfolio gallery, as a selection on the relevant trade page such as the page on paving, and in shortened form on a location page. A project is recorded once and appears in three places — that is where the real time saving lies.
- Trade from a fixed list, spelled exactly as on the service pages
- Location or region, agreed with the client, otherwise the district
- Year of handover, four digits, no month and no full date
- Task: starting point, client's wish, chosen solution — two or three sentences
- Scope: area in square metres, type of material, working days
- Particularity: gradient, drainage, tight access, existing trees — one sentence is enough
- Four images: starting point, build phase with sub-base, detail, result
- The client's approval, filed and findable again
Describe the task, do not praise the result
The most common mistake in reference texts is praise. "A dream garden for happy clients" appears on hundreds of pages and carries no information. The visitor can judge the finished result from the photograph herself; what she cannot see is the route to it. That is exactly what belongs in the text: what did it look like before, what was wanted, what obstacle came up, how was it solved?
An example of the difference. Version A: "Modern terrace with high-quality slabs, a place to feel at home." Version B: "Existing concrete-slab terrace with settlement and waterlogging to the west. New sub-base with drainage, upstand towards the lawn, 42 square metres of natural stone laid in a random bond, six working days." Version B claims less and says more — and it produces enquiries from people with a comparable problem.
The task description does two further things along the way. It supplies the words people actually search with, because nobody searches for "dream garden", whereas settlement, waterlogging and upstand do get typed in. And it lays the ground for the price question: naming area, material and working days makes orders of magnitude visible without quoting a figure. If you also want to work with price ranges, they belong on the pricing page and not in the gallery.
trade: Paving
location: Hildesheim district
year: 2026
title: Driveway with turning area
task: Old concrete surface, cracked, with standing water in
front of the garage. Access had to stay usable for
deliveries, water to soak away at the rear.
scope: 210 m2, grey concrete pavers, central channel, 9 days
particularity: access kept half open throughout the build
images: 01-starting-point.jpg
02-sub-base.jpg
03-detail-channel.jpg
04-result.jpg
approval: client, signed at handover, location approvedHow many projects are enough at the start
The question comes up in every preliminary meeting: how many references do we need? The answer is uncomfortably low. Six projects are enough to start with, provided they are complete (project experience). More important than the number is the spread: those six should cover two or three trades, come from two years and sit in different places. Every filter value then has at least two hits, and no selection leaves the gallery looking empty.
An empty filter result is the reason not to launch with too many categories. A business offering eight trades but able to show projects for only three should offer three filter values to begin with and cover the rest through the service pages. It is less conspicuous not to list a trade in the gallery yet than to list it and then show nothing.
For businesses working mainly in grounds maintenance or winter services the arithmetic differs slightly: there, objects over time are more interesting than individual jobs. A maintained site can be shown across a year in four shots, repeated in the same season the following year. Two or three such sites are enough to convey the way of working, because the evidence there lies not in the appearance but in the reliability over time.
The starting set in one line
Why six good ones beat thirty poor ones
The reflex goes the other way: more must be better, so upload everything. The result is a gallery in which thirty unlabelled shots sit side by side, several of them from the same project and almost the same angle. For the visitor this creates no impression of experience but of effort: she has to decide thirty times whether this picture has anything to do with her plans, and usually gives up before the end.
Six complete projects reverse the ratio. Every entry stands on its own: you see the starting point, understand the task, recognise from the sub-base image how the work is done, and read the key facts. That not only builds confidence faster, it also shortens the conversation afterwards, because the caller already has an order of magnitude in mind. The side effect is better pre-filtering of enquiries — someone who has seen a 210 square metre driveway calls less often about two square metres of doorstep.
The second reason is maintenance. A gallery of thirty unlabelled images can hardly be sorted afterwards, because nobody can reconstruct the details of the older shots: which year, what area, which place, who approved it? A gallery of six complete projects grows on cleanly, because the structure is already there and every new project drops into the same fields.
Removing entries is part of the job
Four images per project are enough
There is a reliable rule of thumb for choosing images: starting point, build phase, detail, result. Four shots in that order tell a sequence. Twenty shots of the finished state tell none. The build-phase image is the underrated one: it shows the base course, drainage, edge restraint or root protection — precisely the work that costs money in the quote and is invisible in the finished picture.
Starting point and result should be taken from the same position. That sounds obvious but is rarely observed, because the before shot is usually taken in passing during measuring and the after shot in good weather from the best angle. Anyone who briefly notes where they stood for the before shot — terrace door, left-hand corner of the plot, front entrance — ends up with a pair that works without explanation. How to manage such shots with a work phone is described under photographs for your website.
Two small technical points belong to each image and tend to get lost in daily work: a descriptive file name instead of the camera's reference, and alternative text describing what can be seen. Together these take less than a minute per image, improve findability and make the gallery usable for people who navigate it with a screen reader.
At the measuring visit
One image of the starting point from a position you can remember. Two seconds of effort, and the project later has a before shot that matches the result.
During the work
One image of the open sub-base before it is covered. This motif exists on a single day only; after that it is lost for this project.
Before handover
One detail: edge restraint, channel, junction with the existing surface, planting. The detail shows care and separates otherwise comparable businesses.
At handover
The result shot from the position of the before image, and the client's approval — short, revocable and signed on the spot.
In the office, same day
Four images into the project folder, fields filled in, task typed up in three sentences. Just after handover this takes minutes; six months later it takes half an hour.
Ordering, home page and cross-links
The gallery's default should show the newest project first. Currency is the signal that lands without anyone touching a filter, and it rewards maintenance. Putting the finest project permanently at the front leaves you, after two years, with a gallery whose first entry comes from a time when half the business looked different.
The home page needs not every project but three — one from each of three trades, with location and year, plus a link to the full gallery. The reverse route matters just as much: below every project there should be a way to the enquiry form that carries the context along, so the form already states what the enquiry refers to. How such a form can be structured is set out under enquiry form for landscapers.
A third route pays off for businesses with a larger service area: location pages showing two or three nearby projects each. The projects are not recorded again but selected through the location field. The same maintenance work then carries three places — gallery, trade page, location page — and the location page consists of documented facts rather than a swapped-out town name.
A reference works not because it is beautiful but because someone recognises their own situation in it. That takes the starting point, not just the result.
Filling the gallery through the year
The starting set is usually assembled in winter, but the continuation has to work during the season, otherwise it does not happen. A realistic rhythm is one new project a month (project experience). Twelve projects a year is more than most landscaping websites show, and the volume builds up in passing, provided the four images arise during the work anyway and the fields are filled in straight after handover.
In practice it helps to tie responsibility to one person and the appointment to a fixed weekday. Half an hour set aside on Fridays for project documentation keeps the backlog small. Doing it "when there is time" leaves forty unassigned images by October. The winter break then means tidying up rather than catching up, and that is by far the more agreeable work.
Upkeep also means letting the gallery follow the season: paving projects to the front in spring, planting and lawns in late summer, tree care and winter services in autumn. This can be steered through the existing fields without rewriting any content. How a website as a whole can be tied to the course of the season is described under seasonal planning.
- Fixed responsibility: one person enters the project, a second checks it before publication
- Fixed appointment: half an hour a week beats an afternoon a quarter
- Obtain approval at the handover appointment, not weeks later by phone
- Clear out once a year: dated motifs, unclear approvals, duplicate views
- Check annually that every filter value on offer still has at least two projects
Common mistakes that cost substance
Three patterns turn up in almost every existing gallery. The first is the wall of pictures without fields: pleasant to look at, not filterable, not extendable, and everything has to be reassigned whenever the website is rebuilt. The second is the catalogue of the finest motifs without any starting point — yet the before shot is exactly what makes someone recognise their own situation. The third is the missing date, left out for fear that older projects might look dated. Without a year they do not look newer, they look unclear.
A fourth point concerns loading time. Images straight from a phone camera are too large for a website, and a gallery is the most image-heavy page there is. Deliver thirty shots at original size and the visitor waits noticeably on a mobile connection — precisely where most evening enquiries begin. Prepared sizes, a sensible image format and loading only on scroll solve this technically; it belongs to building the gallery, not to daily upkeep.
The fifth mistake is the quietest: a gallery without a next step. The visitor works through six projects, finds her own situation — and then the page simply ends. A route to the enquiry form under every project and one at the end of the gallery costs nothing and is the difference between a pleasant visit and an enquiry.