Jobs — what customers hired you to do
In plain English
There's an old marketing line: people don't want a quarter-inch drill, they want a quarter-inch hole.
You describe your product. Your customer describes their problem. Those are different sentences, and the second one sells better — because it's what was in their head before they found you.
This tool reads everything your customers have written — reviews, questions they asked, FAQ submissions — and pulls out what they were actually trying to achieve. Then it groups those into three kinds:
- Functional — the practical thing. "Something that doesn't soak through on the school run."
- Emotional — how they wanted to feel. "Not worrying about it all morning."
- Social — how they wanted to be seen. "Doesn't look like hiking gear in town."
Most product pages only talk about the functional job, and usually in the shop's own language rather than the customer's. The emotional and social ones are frequently the reason someone actually bought.
How to use it
- Click "Growth Science" in the left-hand menu, then "Jobs."
- Choose the scope — your whole shop, or one product.
- Click "Mine jobs." It reads your reviews and questions and takes a moment.
- Read the results. Each job comes with the customer quotes that support it.
- Look for jobs you never mention on your product pages. That's the opportunity.
- Copy the customers' phrasing into your product descriptions, ads, and email subject lines.
- Re-run it every few months, or after you've collected a batch of new reviews.
⏱ ~10 min · 💳 Pro+ · 🎯 Sell the outcome, in your customers' words
Why this matters for your business
You are the worst-placed person to describe your own product. You know how it's made, what's clever about it, and how it compares to the version you shipped last year. Your customer knows none of that and doesn't want to. They had a problem, they went looking, and something you sell seemed like it might solve it.
That gap shows up everywhere. Product pages full of materials and specifications. Ads listing features. Emails announcing that something is "now available in three new colourways." All accurate, all written from the inside.
Meanwhile your reviews are full of customers explaining, clearly and repeatedly, exactly what they wanted and why they were pleased or disappointed. That's the most valuable copywriting resource you own, and almost nobody reads it systematically — there's too much of it, and it's boring work.
This tool does the reading. It comes back with the handful of jobs that come up again and again, with the quotes underneath so you can see it's not making them up.
What this typically unlocks
| What you get | Typical result |
|---|---|
| Reading hundreds of reviews properly | ~10 minutes instead of a lost afternoon |
| Jobs you'd never have named yourself | Usually 2–3 per mining run |
| Conversion after rewriting to the job | +20 to +40% on rewritten pages |
| Returns caused by mismatched expectations | Often lower — the page now describes the right job |
| Ad angles to test | Several, straight from customer language |
What you actually get
The three job types
| Type | What it captures | Example |
|---|---|---|
| Functional | The practical outcome | "Fits a laptop and a lunchbox without looking huge" |
| Emotional | How they want to feel | "Not stressing about whether it'll survive the week" |
| Social | How they want to be seen | "Looks like I made an effort without trying too hard" |
Up to eight jobs per run, each with:
| Field | What it's for |
|---|---|
| The job, in customer language | Copy this into your marketing |
| Type | Functional, emotional, or social |
| Supporting quotes | Proof it's real, not invented |
| How often it appears | How central it is |
Where the text comes from
| Source | Why it's useful |
|---|---|
| Product reviews — title and body | Written after use; says whether the job got done |
| Product Q&A | Written before buying; says what they were worried about |
| FAQ submissions | The questions people ask repeatedly |
Pre-purchase questions are especially valuable, because they capture the doubt that nearly stopped someone buying. Answering that doubt on the page converts the next person who has it.
When it won't run
If there isn't enough customer text — fewer than a handful of usable items — the tool stops and tells you, rather than manufacturing themes out of four reviews.
That's deliberate. Three reviews can "support" almost any theme you go looking for, and acting on a made-up job is worse than acting on none.
If you hit this: collect more reviews first. See Reviews & CRO audit.
How this differs from friction analysis
They read the same sources for opposite things:
| Jobs (this tool) | Friction / CRO | |
|---|---|---|
| Looks for | Why people did buy | Why people didn't |
| Output | Language for your marketing | Problems to fix on your site |
| Use it to | Rewrite copy | Remove obstacles |
Both are useful. Jobs improves what you say; friction improves the experience.
How it works (without the technical bits)
Each run replaces the previous results for that scope, so you always see the current picture rather than an accumulating pile.
Real merchant scenarios
Scenario A — Nobody cared about the technical fabric
Setup. Outdoor-gear retailer. Product pages led with "technical three-layer fabric" and "articulated construction."
Mining 400 reviews found:
FUNCTIONAL "Something that doesn't soak through on the school run"
— 61 mentions
EMOTIONAL "Not having to think about the weather in the morning"
— 34 mentions
SOCIAL "Doesn't look like hiking gear when I'm in town"
— 47 mentions
Not one review mentioned articulated construction.
The social job was the surprise. Nearly a sixth of reviewers mentioned wanting technical performance without looking like they were about to climb something — and the pages didn't acknowledge it at all.
Action. Rewrote the top 20 product pages around the three jobs, keeping the specifications lower down for people who wanted them.
Result: conversion up 34% on those pages. Returns down 11%, because expectations now matched.
Scenario B — The pre-purchase question that was costing sales
Setup. Furniture merchant. Product Q&A mining surfaced a job nobody had considered:
FUNCTIONAL "Getting it up a narrow Victorian staircase"
— 28 questions
Twenty-eight people had asked some version of "will this fit up my stairs?" before buying. Everyone who asked and got an answer was, presumably, joined by many who wondered and left.
Action. Added a "will it fit?" section to every large item — boxed dimensions, whether it comes flat-packed, and the narrowest doorway it will pass through.
Result: conversion on large items up 22%. Pre-purchase questions on that topic dropped by three-quarters — the page now answered it.
Scenario C — Emotional job doubling email performance
Setup. Meal-kit business. Emails focused on ingredients, recipes, and chef credentials.
Mining found the dominant job wasn't culinary at all:
EMOTIONAL "Not having to decide what's for dinner"
— 89 mentions, by far the most common
People weren't buying food. They were buying the end of a daily decision they'd grown to hate.
Action. Rewrote the email programme around decision relief. Subject lines went from "This week: Thai green curry" to "Thursday's sorted."
| Before | After | |
|---|---|---|
| Open rate | 21% | 38% |
| Click rate | 2.4% | 6.1% |
| Cancellations | 8.2%/mo | 5.9%/mo |
Scenario D — Correctly refusing to run
Setup. New shop, 11 weeks old, 6 reviews. Ran the tool and got:
Not enough customer feedback yet to mine reliably.
Collect more reviews and try again.
Initially frustrating. But six reviews can be made to "support" almost any theme, and building a whole content strategy on three people's opinions is a genuinely expensive mistake.
Action. Set up review-request emails after delivery. At 90 reviews, three months later, the tool ran and found two clear jobs they hadn't been mentioning at all.
The lesson. "Not enough data" is the tool being trustworthy.
Best practices
✅ Run it store-wide first, then per product for your top sellers.
✅ Look for jobs your pages never mention. That gap is the opportunity.
✅ Use the customers' exact phrasing. Don't tidy it into marketing language — the roughness is why it works.
✅ Pay special attention to social jobs. They're the most commonly missed and often the real purchase driver.
✅ Mine the pre-purchase questions. They tell you what nearly stopped someone buying.
✅ Re-run after collecting a batch of new reviews — jobs shift as your customer base changes.
❌ Don't act on a job with one or two supporting quotes. Look for repetition.
❌ Don't delete your specifications. Lead with the job, keep the detail below for people who want it.
❌ Don't force a run when it says there isn't enough data. Collect reviews instead.
❌ Don't assume you already know. Scenario C's merchant was certain it was about the food.
Plan tiers
| Capability | Free | Starter | Pro | Agency | Enterprise |
|---|---|---|---|---|---|
| Store-wide job mining | — | — | ✓ | ✓ | ✓ |
| Per-product mining | — | — | ✓ | ✓ | ✓ |
| Functional / emotional / social split | — | — | ✓ | ✓ | ✓ |
| Supporting quotes | — | — | ✓ | ✓ | ✓ |
| Feeds copy tools | — | — | ✓ | ✓ | ✓ |
| Multi-store mining | — | — | — | ✓ | ✓ |
Frequently asked
How many reviews do I need? More than a handful — the tool will tell you if you're short. In practice, results get genuinely useful somewhere around 50+ items across reviews and questions.
Does it read negative reviews too? Yes. Negative reviews often describe the job most clearly, because someone explains exactly what they wanted and didn't get.
How often should I run it? Every few months, or after a batch of new reviews. Jobs move slowly.
Can I mine one product? Yes — pick the product scope. Useful for your best-sellers and for anything converting below expectation.
Does it invent quotes? No. Every job shows the actual customer text behind it. If a quote looks wrong, the job is wrong — tell the difference by reading them.
Why only eight jobs? Beyond that they stop being distinct and start being variations. Eight is enough to rewrite a catalogue around.
Does this replace reading reviews myself? No — it finds the patterns. Reading a few in full still teaches you things no summary will.
Where should I use the output first? Your highest-traffic product page. Biggest measurable effect, fastest.
See also
- Growth Science overview — all eight tools
- Positioning — use these jobs to sharpen it
- Awareness — matching the job to the reader's stage
- Reviews & CRO audit — collecting more reviews, and friction analysis
- CRO & storefront copy — rewriting pages around jobs
- Ad & email copy — job-led ad angles
- Customer 360 — who's saying it