My inbox used to be short.
Not always polite. Not always clear. But short. A client would ping: “Checkout’s broken after the plugin update — can you look?” Three lines. I’d reply in four. We’d get on with it.
Lately the replies coming in feel… padded. Warm intros. Soft landings. Three bullet points restating the question I just asked. A closing paragraph thanking me for my patience. Another one offering further clarification should I require it.
Nobody had time to write that. An AI did. And somehow we’ve decided the person on the other end does have time to read the long assed version.
They don’t. I don’t. And if you’re a freelance WooCommerce developer living in support tickets and client email, neither do you.
The new shape of a “quick reply”
Here’s the pattern I keep seeing.
Someone asks a simple question — “Is the staging site still on PHP 8.1?” or “Can we bump free shipping to £50?” — and what comes back is a miniature essay. Context they already know. Options they didn’t ask for. A confident tone that somehow still says nothing.
I’ll paraphrase two versions of the same reply. Same facts. Different length.
The waffle:
Hi Neil, thank you so much for reaching out regarding the shipping threshold question. I’ve carefully reviewed your note and wanted to provide a comprehensive response. After considering the current configuration and our broader promotional strategy, I believe adjusting the free-shipping minimum to £50 could be a valuable next step, though of course we’d want to align with stock levels, margin considerations, and any upcoming campaigns. Happy to jump on a call if helpful. Please don’t hesitate to let me know how you’d like to proceed. Kind regards.
The useful one:
Yes — bump free shipping to £50. I’ll change it on staging first, you check a test order, then we push live. About 20 minutes. Say the word.
Same answer. One of those I can act on before the kettle boils. The other I skim, sigh, and ask again in Slack.
I’m not anti-AI. I use it myself — for drafts, for sorting notes, for turning a messy brain-dump into something publishable. The problem isn’t the tool. It’s the default setting: helpful, thorough, and quietly endless.
Why this lands harder when you work client shops
When you look after live WooCommerce stores, communication isn’t a hobby. It’s how orders stay flowing.
A client emails because checkout failed for three customers. Supporting a theme update. A Stripe webhook that stopped firing. The care-plan inbox isn’t a book club — it’s a triage queue. Every extra paragraph is friction between “something’s wrong” and “someone’s fixing it.”
I’ve had support threads where the first useful sentence was buried under two screens of AI polish. By the time I found “we think it’s the shipping plugin,” I’d already checked half the wrong things.
And clients feel it the other way too. If I send them a long AI reply about a five-minute fix, I’ve trained them that every email from me is homework. That’s a terrible brand for a solo developer who sells clarity.
So this isn’t a LinkedIn rant about the death of writing. It’s an inbox observation from someone whose week is made of short questions that need short answers.
If you didn’t have time to write it…
Here’s the punchline I keep coming back to.
If you didn’t have time to write the reply yourself, why assume I have time to read a long assed email or text that an AI spun up for you?
You outsourced the writing because you were busy. Fair enough. I’m busy too. So is every shop owner watching their till. Length isn’t generosity. It’s often just the model’s idea of being complete.
The fix isn’t “never use AI.” It’s “train the thing before you hit send.”
What I tell the AI before it drafts a reply
When I use AI for client-facing copy now, I give it boring, blunt instructions up front. Something like:
- Answer in under 80 words unless I say otherwise
- Lead with the decision or the next step
- No throat-clearing (“I hope this email finds you well”)
- No restating the whole question
- No fake enthusiasm
- British English, plain, first person as me
- If you’re unsure, ask one short clarifying question — don’t invent confidence
Then I edit. Always. The model drafts; I decide what leaves the building.
That last bit matters. An AI reply that sounds like me but isn’t checked by me is still my reputation. Clients hire Neil Matthews, not a autocomplete with manners.
A before-and-after from support life
Client: “Orders stopped exporting to the warehouse after yesterday’s update. Urgent.”
AI’s first draft (before I trained it): a sympathetic novel about change management, plugin ecosystems, and optional next steps across four paragraphs.
What I actually sent:
Got it. I’ll check the export plugin and the cron on staging mirrors, then production. Expect an update in 30 minutes either way — fixed, or what I need from you.
They didn’t need poetry. They needed a time box and a human on it.
The practical takeaway
Use AI for replies if it helps you keep up. Just don’t let the default “be thorough” setting turn every text into a brochure.
Tell it: short, clear, next step first. Cut the padding before you send. If the reply still feels long, it probably is.
Your clients — and the freelancers in their inbox — will thank you by actually reading what you wrote.
Need a hand?
If you want a WooCommerce developer who answers in plain English and fixes the shop without the waffle, hire me — fixed-price quote, no obligation.
Written by Connie, an AI, from Neil Matthews’ notes, archive, and direction. Neil directed and edited this piece. He did not write these words. Augmented publishing by Neil Matthews.