Sign makers work with their hands. On site, on the ladder, at the vehicle, at the printer. Typing on a smartphone isn't an option in this reality. Dirty gloves, tools in hand, rain, cold — these are the conditions under which information needs to be captured. All features: Feature overview →

Yet the industry types. Or worse: writes on scraps of paper. I'm Björn, founder of PlotonIQ and a sign maker with over 20 years of experience. I wrote on scraps of paper for years. I know how that ends: illegible abbreviations, forgotten dimensions, 47 untagged photos on the phone, and an evening in the office that runs to 10pm.

The audio-first workflow in PlotonIQ flips this. You speak — the AI does the rest. From the car park to the sent quote: under 5 minutes.

The real problem: information loss between job site and office

Every sign maker knows this: you're standing in front of the storefront, the client describes what they want. You write on a scrap of paper. "3x win frost, 1x door hrs wht." Back in the office: what does "frost" mean here? Which brand? And is that a 3 or an 8?

Then there are the photos. 47 photos on the phone, somewhere between lunch and holiday pictures. Which photo belongs to which job? Which side of the building does it show?

The result: you spend 45-60 minutes in the evening reconstructing a 20-minute client conversation into a quote. Most of that time is reconstruction work — trying to remember what you saw and heard on site. That's wasted energy.

How the audio-first workflow works

Step 1: Create a memo

Open PlotonIQ on your phone. Tap + → New Memo. Three snippet types: Text, Audio, Photo. For site work, start with Audio. No login process, no loading screen. You're ready in 3 seconds.

Step 2: Dictate

Press the record button and speak freely. No form, no fields: "Smith's Bakery, 12 High Street. 4 shop windows, roughly 150 by 110 centimetres each. Frosted film, client wants 3M Crystal. Entrance door needs opening hours, white vinyl lettering. Installation Thursday, shop's closed till 11am. Contact is Mr Smith himself, number's in the CRM."

Stop recording. 30 seconds. That's it.

Step 3: Attach photos

Take 3-4 photos directly in the app: full facade view, window detail, entrance door, any issues (a roller shutter that's in the way). Each photo becomes a snippet in the memo — auto-geotagged with timestamp. No more chaos in the camera roll.

Step 4: AI transcribes and structures

Whisper (OpenAI) transcribes the recording. 95%+ accuracy, even with background noise on site, engine running, or wind. The AI automatically recognises: customer name, address, materials, dimensions, and scheduling info.

The memo becomes structured: contact data, positions, materials, dates. From a free-form dictation to a clean job document — without you typing a single character.

Step 5: Memo to quote

One tap: "Generate AI Quote." The AI takes the recognised positions, prices with current market data, generates price justification per position, and creates a print-ready PDF. Review, adjust if needed, send.

Total time from the car park to sent quote: under 5 minutes.

Real-world example: Pylon with header unit and six illuminated signs

Picture a complex job: car dealership, new opening. 1 pylon 3m with illuminated header unit frosted Perspex, 6 illuminated signs 200x80cm on the facade, 4 shop windows frosted film 3M Crystal, 3 vehicles with Oracal 970 wrap.

Old way: 2-3 hours at the survey. Back in the office: 90 minutes writing the quote. Total time: nearly half a working day for one quote.

With audio-first: you take 8 photos on site. You dictate 90 seconds. You use AI photo recognition for the shop windows. In the car you generate the quote. Back in the office you review it once, adjust one price, and send. Total quoting time: 12 minutes.

Facade sketch for complex projects

For facades with many elements, PlotonIQ offers an SVG sketch directly in the memo. You draw a simplified view — rectangles for windows, lines for facade edges — and assign each element the dictated material. This isn't a CAD program and isn't meant to be. It's a quick sketch that tells the office exactly what was meant. Especially useful when you're not the one writing the quote.

Offline capability

On site you don't always have signal. In basements, underground car parks, industrial estates without coverage — you know this. Audio recordings and photos are stored locally and sync when connectivity returns. AI transcription happens server-side — you need signal for that. But capture works offline. You never lose a recording because you're in a basement.

Why not just a notes app?

You could use your iPhone's voice memo app. But then you have an audio file — not structured data. You'd need to listen again, type out the info, build a quote, look up material prices. Same effort as the paper scrap.

The difference with PlotonIQ: transcription is the first step of an automated pipeline. Speech → Text → Positions → Pricing → PDF → Send. Each step flows into the next. It's not a voice memo — it's a voice-to-quote workflow.

Concrete benefits with numbers

  • Information capture on site: 20 minutes typing/writing → 60 seconds dictating
  • Evening office work: 60-90 minutes reconstruction → eliminated entirely
  • Quote creation: 45 minutes → 3-5 minutes including review
  • Forgotten dimensions and return trips: regular (1-2 hours per return trip) → almost never
  • Photo chaos: 47 untagged photos → every photo attached to the right job

How sign shops use the audio-first workflow

A vehicle wrap specialist from the East Midlands — 2 staff, primarily vehicle wraps and window frosting. Previously: scraps of paper on site, office evenings until 9pm. The owner describes it: "I do site visits in the morning and write quotes in the evening. That used to frustrate me. I'm a craftsman, not an office worker. Since using audio-first I dictate in the car and the quote is sent before I get home. It gave me my evenings back."

An exhibition and display company from London (6 staff): "We often have complex projects with many line items. Previously, writing a quote for a medium-sized exhibition build took 3-4 hours. Now we dictate on site and use the facade sketch tool for the layouts. The quotes are more detailed and go out faster than before. And we're winning more jobs because we're with the client when we create the quote — that makes an impression."

What both share: the time saving is real, but the actual gain is the quality of information capture. No information loss between job site and office. What is seen and heard on site goes directly into the system.

Tips for better results

Say material names in full: "3M Crystal frosted film" not "frost." AI recognises brand names better than abbreviations you've invented.

Include units: "150 centimetres by 110 centimetres" not "one-fifty by one-ten." Numbers with units are recognised more reliably than spoken number words alone.

One memo per job: Don't mix multiple clients in one memo. Especially important if you have several sites on the same day.

Photos first: Take photos before dictating — you'll have visual reference while speaking and won't forget any position.

Dictate in the car, not on the ladder: In the car it's quiet. You're sitting. You can see the photos. That's the ideal moment to dictate — not while you're still on site.

Combining audio-first with AI measurement: the complete workflow

The real power emerges when you combine audio-first with AI measurement. On site you take photos with the PlotonIQ app. The photos are automatically linked to the memo. Then you dictate. Back in the office or in the car: you open the measurement tool, tap elements in the photo, enter a reference dimension. The AI recognises polygons, calculates areas, assigns the dictated materials to the measured surfaces.

The result: a complete job document with photos, dimensions, materials, and quantities — from a 5-10 minute workflow on site. What used to mean half an afternoon of office work.

Not every job needs AI measurement. For standard jobs the dictation alone is enough. But for complex facades, large pylons, or jobs where accurate area calculation is critical, the combination is hard to beat.

Frequently asked questions

Does voice recognition work with strong regional accents?

Whisper is trained on accents. Scottish, Welsh, Northern Irish, various regional English accents — recognition rate stays above 90%. Trade terms like frosted film, vehicle wrap, pylon, or channel letters are reliably recognised because the model is calibrated for sign industry vocabulary.

What if I dictated something wrong?

You can edit the transcript directly in the memo before generating the quote. You see the text, correct any errors or number transpositions, and then generate the quote from the corrected text.

How long does AI transcription take?

A 60-second recording is transcribed in 5-10 seconds. Transcription happens server-side — you can continue working or close the memo while it processes, it runs in the background.

Can I dictate in multiple languages in the same memo?

Whisper detects the language automatically. For shops working in border regions or with international clients: English and other European languages in the same memo works reliably.

Conclusion

Audio-first fits the reality. Hands busy, desk far away, time short. Instead of typing quotes in the evening, dictate in the car park and the quote is sent before you start the engine. This works today, on any smartphone, without downloading an app.

14-day free trial. No credit card. Try the audio-first workflow at your next client visit — just once. You won't reach for a piece of paper again. Start for free →