Yes. A shutters company can send a customer a picture of their own window with shutters fitted before the quote goes out, using either a free visualiser the customer runs or an AI image edit you make in about ten minutes. Call it an impression: the written specification in the quote, not the picture, describes what you will fit.
The catch is accuracy. An AI edit redraws the window rather than laying a product over it, so panel counts, frame depth and louvre size can drift from what you would build. Choose the route by how quickly you must reply and what your time costs, then check every image against your survey notes before it is sent.
Four routes to a picture, compared on your minutes and their cost
Two numbers decide the route: the minutes you spend per quote and what the tool costs. Image quality comes third, because a customer who waits three days for a picture has often asked another firm by then. The minutes below are illustrative, so time yourself over ten real quotes and replace them. Prices are the ones on the vendor pages I read on 1 November 2026.
| Route | Who makes the picture | Your minutes per quote | Cost on the vendor page | Where it goes wrong |
|---|---|---|---|---|
| 1. A consumer shutter visualiser the customer uses | The customer, from their own photo | 2 (send the link, read what comes back) | AI Shutters is free to use with no sign-up; its vendor options (leads, a white-label visualiser, quote automation) show no price on the home page | The output is not yours: styles and prices may not match your range, and the page makes no accuracy statement |
| 2. A home-exterior AI web tool | You, from a daytime photo of the house | 6 | Home Exterior AI: first render free with no login (free results watermarked); Pro listed at $19 a month | Built for whole-house changes, so interior louvre detail is outside its job |
| 3. A chat assistant's image editing (ChatGPT, Gemini) | You, from the customer's window photo | 10 | Included in a chat plan you may already have; check the allowance shown in your account, because image limits differ by plan | Redraws the window, so it needs a frozen-window prompt and a check |
| 4. A photo editor or Adobe Firefly, with or without generative fill | You or the office | 30 | Firefly Standard $9.99 a month for 2,000 credits | Slowest, though it gives the most control over panel layout |
Route 1 costs you almost nothing and gives you the least control. Routes 3 and 4 give you the most say over the result, but the picture is only as good as the checks you apply afterwards. A sensible split is route 3 for the everyday enquiry and route 4, or a manual mock-up, for the large, shaped or high-value window.
What the product pages say, and what they leave out
AI Shutters is aimed at homeowners. They upload a room photo, the tool finds the windows, and they pick a style, colour, slat size and mount type before seeing a price estimate with the option to request quotes or book a measure. A separate section for vendors lists leads, a white-label visualiser, quote automation and a lead dashboard, without prices. The page carries no accuracy disclaimer. If you point customers at a tool like this, test what its request-a-quote step does with an enquiry first, and find out where the enquiry goes.
Home Exterior AI makes redesign previews from a house photo. Its pages say renders take roughly 8 to 12 seconds, free downloads carry a watermark, and previews are "concepts, not guarantees". A separate shutter-colour page adds louvred shutters to front windows or repaints existing ones, on the same free-to-try terms. Its pricing page lists Pro at $19 a month for 150 fast generations without a watermark, plus one-off packs such as 10 credits for $2.99. The 20-credit pack appears as $4.99 on the pricing card and $9 in the FAQ, so confirm the price at checkout.
AI Two is a general design platform rather than a shutters tool. Its featured plans bill weekly, at what the page marks as limited-time prices: $6.99 for 100 credits, $9.99 for 300 and $14.99 for 500, renewing each week until you cancel, and purchases are not refundable. Monthly and yearly toggles sit on the same page without prices shown. At the weekly entry price that is about $30 a month, which is easy to forget if you subscribe for a single quote. A fourth name in the same search results, See My Shades, is titled as an AI window treatment visualiser; its page returned no readable detail for me, so I have not assessed it.
None of the pages gives a measured accuracy figure for shutters, and only Home Exterior AI states any limit on its previews. That gap is the reason the checks later in this tutorial exist.
The Complete Excel Course with AI Integration
From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.
By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner
Buy nowCost per quote for a firm sending 40 a month
Take a six-person shutters firm that sends 40 quotes a month, 24 of them from enquiries with a usable window photo. Office time is valued at $30 an hour, and an average job leaves $900 of margin. All of these are illustrative figures, so replace them with yours.
| Route | Minutes per usable photo | Minutes a month (24 photos) | Time cost at $30 an hour | Tool cost a month | Total a month | Per quote sent (40) |
|---|---|---|---|---|---|---|
| 1. Customer-run visualiser | 2 | 48 | $24.00 | $0 | $24.00 | $0.60 |
| 2. Home-exterior web tool, Pro | 6 | 144 | $72.00 | $19 | $91.00 | $2.28 |
| 3. Chat assistant edit (plan already paid) | 10 | 240 | $120.00 | $0 | $120.00 | $3.00 |
| 4. Photo editor and Firefly Standard | 30 | 720 | $360.00 | $9.99 | $369.99 | $9.25 |
Read the last column as the price of a picture on every quote you send. At $900 of margin per job, route 3's $120 is repaid by 0.13 extra jobs a month, which is one extra job every seven to eight months. Route 4's $369.99 needs 0.41 extra jobs, or one every two to three months.
These sums do not show that a picture wins work. They show how low the bar is, and they point to a test that costs nothing: for one month send a picture with every second quote, log the reply rate for each group, and keep the habit only if the picture group replies more often.
The photo brief that stops most bad images before they start
Most poor visualisations start with poor photos. A window shot from a sharp angle hands the AI a trapezium to repair, and it repairs it by inventing. Send the brief below when the enquiry arrives, in the same message that asks for rough measurements.
Hi [first name], thanks for your enquiry. To show you shutters on your own window before we quote, could you send:
1. One photo per window, taken from straight in front, standing as far back as the room allows.
2. The whole frame in view, with a hand's width of wall visible on every side.
3. Curtains and blinds open, so the glass is clear.
4. A second photo with a sheet of paper (A4 or letter size) held flat against the glass, so we can judge scale.
5. Daylight if possible, and no flash.
Rough width and height are welcome too. If your windows are different shapes, please number the photos.Each instruction heads off a specific failure.
- Straight on, whole frame: keeps the window's real proportions, so the louvres are scaled to the glass and not to a distorted shape.
- Wall visible all round: gives the AI wall texture and reveal depth to copy, so it does not crop into the frame.
- Curtains open: otherwise the edit has to guess where the glass ends and may draw shutters over fabric.
- The paper photo: this one is for you, not the AI. A sheet of A4 is 297 mm tall, so you can sanity-check the customer's rough sizes against the survey notes.
- Daylight, no flash: flash bounces off glass and gives the AI a white blob it will turn into something else.
Before you spend ten minutes on an image, a thirty-second photo check saves a bounce-back email. Upload the photo to your assistant with this prompt.
Is this photo usable for a shutter visualisation? Answer PASS, or FAIL with the single most important change the customer should make (angle, framing, glare, curtains, light). Do not describe the room.On a FAIL, forward the reason in plain words. Keep the three reasons you see most often as saved replies, so the customer gets a clear request within minutes of sending a bad photo.
Making the in-house image: a prompt that keeps the window unchanged
The prompt does three jobs. It freezes everything outside the window opening, it names the product in trade terms you can check later, and it caps the output at one image. Fill in the survey facts you already hold, since the assistant will otherwise choose its own panel count and louvre size.
Edit the attached photo. Change only the glazed area of the window: add interior plantation shutters, closed panels with louvres open about 45 degrees.
Keep unchanged: walls, window frame, sill, handles, radiator, curtain rail, lighting, perspective.
Specification: [6] panels in [3] sets, [2] per window, [89 mm] louvres, [white], [hidden tilt], [inside mount].
Do not add or remove window bars. Do not change the number of windows.
Return one image. If the photo does not show enough detail to follow the specification, say what is missing instead of guessing.The last line matters more than it looks. Giving the assistant permission to refuse turns many silent inventions into a visible question you can answer.
Here is what an assistant can return, described the way you would see it. The image looks tidy: white shutters, even louvres, sunlight across the sill. Counting shows eight panels where the specification says six, and the louvres look about half the width you asked for, so the shutters seem lighter than the product you sell. The radiator below the centre window has also disappeared, which makes the sill look deeper than it is. All three are ordinary edit drift, and none is visible at a glance.
The fix is a second prompt that names the faults instead of asking for a better image.
Redo the edit with three corrections.
1. Your image shows 8 panels. The specification is 6: two per window, three windows.
2. The louvres are about half the width I specified. Make them twice as wide, so there are half as many on each panel.
3. Put the radiator back below the centre window, exactly as in the original photo.
Change nothing else.Two rounds is the limit I would set. If the third attempt still drifts, the photo is the problem or this route is wrong for that window, so move it to route 4 and fix it by hand.
Checking the image against the survey notes before it leaves
A picture that looks right is not a picture that is right. Run the checks below on every image, each against the document that settles it, and tick them on the quote record. Once you have done it a few times it takes under two minutes.
| Check | Settled by | What to look for |
|---|---|---|
| Panel count and layout | Survey sheet | Count panels window by window, then compare sets, hinge side and fold direction |
| Louvre size | Range specification | Count louvres up one panel: too many means narrow blades, too few means wide ones |
| Midrail | Survey sheet | Present or absent, and roughly where; AI often adds or drops one |
| Frame style and depth | Survey sheet and reveal photo | Inside or outside mount; the frame should not hide the glass or handles unless the quote says so |
| Tilt rod | Range specification | A visible rod you do not offer is a misdescription, so remove it or say it is hidden tilt |
| Obstructions | The original photo | Radiator, handles, sill, sockets, window furniture: each should still be there |
| Window shape | The original photo | Arched tops, angles and glazing bars should match; the AI tends to flatten or add them |
| Colour | A physical sample | Never promise colour from a screen; name the colour from your range in the quote |
Colour is the one check no image can pass. Screens and AI renderings both shift it, so the quote should name the colour as it appears in your range and the email should offer a sample. If the customer later says the shutters are not the colour in the picture, that sentence is your first line of defence, and a sample in their hand is the second.
The disclaimer line under every image
Put the wording on the image itself, as a caption bar, and not only in the email, because images get forwarded without the message around them. Keep the original photo and the generated image together with the quote record.
Artist's impression created with AI from your photo. Louvre size, frame position, panel layout and colour are indicative. The specification in your written quote is what we will make and fit, confirmed at survey.Two further habits make the line believable. First, only claim what the picture can honestly show: if you have not shown the tilt rod, do not describe it in the caption. Second, send the disclaimer and the specification together, so the customer reads one against the other. I am not a lawyer, and the rules on how a product image may be used in selling differ between markets, so ask a qualified adviser to read your wording once.
One more habit is worth the effort: keep six to ten real photos of finished jobs, labelled by louvre size, colour and window type, and send one beside the AI image. The pairing shows what the product looks like when it is actually made, and it reminds the customer which of the two pictures is the impression.
The ten-minute routine, minute by minute
The "about ten minutes" in route 3 holds only if the steps are in a fixed order. This is the sequence I would give a sales coordinator on their first day.
| Minutes | Step | What you do |
|---|---|---|
| 0 to 1 | Photo check | Upload the customer's photo with the PASS or FAIL prompt; on a FAIL, send the saved reply and stop |
| 1 to 2 | Pull the facts | Open the survey notes and paste the specification line into the image prompt |
| 2 to 6 | Generate | Run the prompt, read the result, and run one correction round if the check table finds drift |
| 6 to 8 | Check | Work down the check table against the survey sheet and the original photo, and tick the quote record |
| 8 to 9 | Caption | Add the disclaimer bar and save the image under the quote reference |
| 9 to 10 | Send | Attach it to the quote email and set the day 2 reminder |
It stays at ten minutes only when the photo passes the first check and the prompt lands within two rounds. When it does not, the time climbs quickly, which is why the rule to move that window to route 4, or to a survey visit, is written down rather than left to mood.
Follow-up that uses the picture: three messages and their timing
The picture gives you a reason to write that is not "just checking in". Send the first message within the hour while the customer is still at their screen, then follow the schedule below.
| When | What it contains | Question it asks |
|---|---|---|
| Within an hour | The image with its caption, the quote, the specification in one paragraph | Does the layout match what you pictured? |
| Day 2 | One line referring to a detail in the picture, such as the bay or the louvre size | Would you like to see this in another colour or louvre size? |
| Day 6 | An offer of one revised image, or a survey slot | Shall I hold a survey slot this week? |
Ask the assistant to draft the second and third messages from the quote data so each one mentions something specific to that window, then read them before sending.
Write a two-sentence follow-up for day 2. Customer: [first name]. Quote sent: [date]. Windows: bay with three sections, white shutters, 89 mm louvres. Mention the bay, ask whether the layout matches what they pictured, offer one revised image. Plain tone, no exclamation marks, no discount offers.A sample result, illustrative: "Hi [first name], I hope the picture of the bay was useful. Does the six-panel layout match what you had in mind? I can send one revised image if you would like a different louvre size." Check it names the right window and promises nothing you have not agreed.
To find out which route earns its minutes, keep a simple log: date, route, minutes spent, time sent, date of reply and outcome. After twenty quotes you will see whether the slow route replies more often than the quick one, and whether either beats a quote with no picture.
When the AI added a panel to a bay window
Here is how it can go wrong, with an invented four-person shutters firm. It sent a picture for a three-section bay. The image showed neat, evenly sized panels with sunlight across the sill. The survey sheet said six panels. The AI had drawn eight, and nobody counted. The customer did, asked why the quote said six, and wondered what else differed. The sale survived, but the survey visit began with an apology instead of a measure.
The firm made two changes. The panel count went into the prompt as a number, and "panels counted against the survey: yes or no" became a tick box on the quote checklist that must be ticked before sending. Neither change takes more than seconds, and both come straight from the check table above.
When to skip the picture
A picture is not free, so spend it where it changes the decision. Use the table as a starting rule, then adjust it from your reply log.
| Situation | Picture? | Why |
|---|---|---|
| Customer has asked about colour or style | Yes | The picture answers the question they are already asking |
| Bay, arched or oversized windows | Yes, route 3 or 4 | These are the windows customers find hardest to imagine, and the value justifies the minutes |
| A competitor has already sent one | Yes, quickly | Speed matters more than polish here, so use route 3 |
| Standard windows, customer already chose the range | No | A sample and the survey do the job |
| Survey is booked within two days | No | You can show samples in the room |
| Quote value too low for ten minutes of office time | Route 1 only | Let the customer play with a free visualiser and send the quote as normal |
For the measuring and quoting side of the same job, see blinds quoting software and the window measurement app tutorial. For the wider pattern of showing a finished room before quoting, read AI remodel renderings from client photos and interior design mock-ups with AI. Once the picture is out, a quote follow-up email sequence keeps the conversation going without nagging.
Shutters visualiser questions from the sales desk
Is an AI shutters picture part of the contract?
Treat it as an illustration only. The written quote and survey specification should say what you will make and fit, and the picture should carry a line saying louvre size, layout and colour are indicative. Ask a qualified adviser to read your wording once, because rules on product images differ between markets.
Can I reuse a customer's window photo in my marketing?
Only with their clear permission, ideally in writing at the quote stage. Photos of lived-in rooms can show family members, post, screens and street details. Crop or blur those, and keep a record of who agreed. If the customer says no, the picture still serves for their own quote.
Which AI tool gives the most accurate shutters image?
There is no safe winner to name, because results change with each model update. Run the same three customer photos, including one bay and one shaped window, through each candidate with the same prompt. Count panels and louvres on every output and keep the tool that needs the fewest corrections.
How many revisions of the picture should I offer?
Two per window is a workable limit: the first image, then one revision after the customer's comments. Beyond that you are designing for free. Say so in the email, and offer the survey visit as the next step for any further changes.
Do free visualisers pass my customer's enquiry to other installers?
Some are built to match homeowners with installers or to sell leads to shutter businesses, so check what the request-a-quote step does before you point customers at one. Read the privacy notice, and test the tool yourself with a dummy enquiry to see who contacts you.
Further reads
- Following Up Silent Quotes With AI: A Decorator's Playbook — A decorator's follow-up playbook that carries over to shutters quotes.
- Glazier Software: Quoting and Scheduling Tools With AI Features — Quoting and scheduling software for window trades, with AI features compared.
- ChatGPT for Contractors: Five Customer Messages to Draft and Check — Five customer messages to draft with ChatGPT and check.
- Getting More Reviews After Every Job With AI Follow-Ups — Ask for a review after the fit, using your job photos.
- Bathroom Fitter Marketing: Before-and-After Posts From Job Photos With AI — Turn finished-job photos into before-and-after posts.
- AI Renders vs Outsourced Visualisation: What Practices Save — When in-house AI images beat paying someone to make them.
- Cabinet Design Software: What AI Adds for a Small Cabinet Shop — Compare cabinet design software by price band, CNC output and cut-list accuracy, and see what AI renders and drafting really add for a small shop.
- Window Company Marketing: A Year of Local Campaigns Drafted With AI — A month-by-month plan for a window installer: fact sheet, autumn and spring campaigns, case studies from job notes and a cost-per-sale table.
- AI for Painting Contractors: Can It Show Colours Before You Quote? — Test room-photo colour previews on your own jobs: what visualisers and AI editors get right, where they mislead, and the approval step that protects you.
- AI Furniture Visualisation for Makers: Show a Piece in the Room — A maker-side workflow for a room-photo preview that supports decisions without pretending to be a measured installation drawing.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: product pages for AI Shutters, Home Exterior AI (home, pricing and shutter-colour pages) and AI Two (pricing), viewed 1 November 2026; Adobe Firefly plan price from Adobe's Firefly plans page, read in dollars.