Listings, infographics and handling negativity in one pipeline

The product sits there without selling, and the first decision is to switch on advertising. It multiplies what is already there: if the listing does not convince, every new impression runs into the same bottleneck. Below is the order of work in which advertising comes last.
The product sits there without selling, and the first decision is to switch on advertising. It multiplies what is already there: if the listing does not convince, every new impression runs into the same bottleneck. Below is the order of work in which advertising comes last.
The product is good, the sales are not happening
The seller sees the figures and draws the only conclusion available: not enough impressions, must invest in promotion. He invests. Impressions rise, sales barely. The money has gone.
The cause is usually not impressions. People see the listing and do not buy — and that is a completely different problem, which advertising does not cure but only makes more expensive.
There are not many places where the buyer is lost, and they are predictable: he did not understand what it is; could not make it out on a small screen; did not find the answer to his question; read a bad review and left; did not see how it differs from the neighbouring product at the same price.
Each of those places is fixed separately, and the order matters.
Why advertising does not come first
The root cause: advertising multiplies what is already there.
If a listing converts badly, more traffic increases the spend rather than the revenue. And you will not even see it immediately — rising impressions create a sense of movement.
The misconception: "we need to make it prettier". Beauty does not sell on a marketplace. What sells is being understandable in two seconds: what it is, what it is for, how it differs. A beautiful listing that leaves the product unclear works worse than a plain and clear one.
A second misconception: "price is what matters". Price decides when everything else is equal. Until the buyer has understood the value, any price seems high.
And a systemic layer. The listing, the reviews and the questions work as one whole. You can redo the photographs and get nothing, because the person reaches the reviews and leaves there.
The order of work
Step 1. Where exactly the buyer is lost
Before changing anything — an analysis. You look at which step sags: impressions, clicks into the listing, adds to basket, purchases.
Each failure points at something different. Few clicks with normal impressions means the main image and the title. They click in but do not add means they did not understand the value or did not find an answer. They add and do not buy means price or delivery.
This is the one step that cannot be skipped: without it you will be fixing blind.
Step 2. What buyers say
Reviews and questions are free research that almost nobody reads systematically.
Questions show what is missing from the listing. If the same question is asked twenty times, the answer to it belongs in the description — and every question asked means a dozen people who did not ask and left.
Reviews show real objections. Not invented ones but the ones that have already happened.
Running the mass of reviews and questions through an assistant gives a list of what to fix — and it is always more accurate than hypotheses.
Step 3. The main image
The first thing a person sees and almost the only thing that decides whether he clicks. The requirements are strict: understandable on a small screen, understandable in two seconds, with the product and its size visible.
The most common mistake is an image that looks good at full size and turns into a smudge in the results.
Step 4. Infographics
The task is to answer questions without making people read the description. Size, composition, what is included, how it differs, how to use it.
The rule: one screen, one thought. Trying to fit everything on one image produces something unreadable.
The infographic is built from what was found in the questions — not from what the seller thinks is important.
Step 5. Title and description
The title in the words people search with, not the words the manufacturer uses. The description to a structure: what it is, who it is for, what is included, answers to frequent questions.
Step 6. Working with reviews
Answer all of them, including the bad ones — the bad ones especially.
A bad review with a sensible answer sells better than no review at all: the buyer sees that a problem will be dealt with. A bad review with no answer reads as agreement.
The answer is built on a scheme: acknowledge, explain the cause without excuses, say what has been done, offer a solution. Nobody wins an argument with a buyer in the reviews.
A separate mechanic is reminding satisfied buyers to leave a review. The dissatisfied write by themselves, the satisfied forget, and that skews the picture downwards.
Step 7. And only now, promotion
Once the listing converts, every unit of spend works. Before that, it does not.
How to tell it is the listing rather than the product
An unpleasant but necessary question: sometimes it does not sell because nobody needs it.
Look at the conversion from click to purchase for neighbouring products. If similar items convert noticeably better at a similar price, it is your listing.
Read competitors' reviews. They say what people pay for and what they are unhappy about. If your product closes a named pain and your listing says nothing about it, you have found the cause.
Count returns. A high return rate with normal sales is a sign the listing promises something other than what arrives. That is fixed by the description rather than by advertising.
And the honest option: if conversion is low across the whole category rather than just for you, there may genuinely be no demand, and redoing the listing will not help.
What the seller gets
- A funnel analysis: where exactly the buyer is lost
- A list from reviews and questions — what to fix, in priority order
- A set of images: main, infographics, additional
- A title and description matched to buyers' search phrasing
- A procedure for handling reviews and answer templates for typical situations
- A mechanic for collecting reviews from satisfied buyers
- The order for launching promotion — after, not before
What relaunching product cards costs
The cost is calculated from the volume: the number of listings, whether photography is needed or everything is generated, whether ongoing review handling is required or only setting up the process.
Honestly: the fee here is stated by our methodology from the volume of work rather than taken from market statistics.
What to compare it against. The advertising budget going into a listing with poor conversion. Usually the cost of the rework pays for itself in weeks.
A caveat about priority
I will say this directly, since it bears on the decision: in my audience research the subject of marketplaces came last for interest — noticeably below the other directions.
That does not mean the market is small. It means fewer people want to work in it than is generally assumed. For a seller already there, that is good news: competition among contractors is lower than in other niches.
What relaunching product cards needs
AI assistants — 200+ across 14 categories, including e-commerce: review analysis, listing texts, answers to negativity, analytics.
Image generation — for product shots and infographics with no studio. Paid from the wallet on usage, with the cost visible before you press.
Automation templates — collecting reviews, notifications of new questions. Creator plan and above.
Assistant chats come with Basic. Registration is free and opens three days of Basic access — enough to run one listing's reviews through and get a list of what to fix.
Do one thing today: open the questions on your product and count how many times the most frequent one recurs. The answer to it belongs on the infographic. That is the cheapest fix with the most visible effect.
A conveyor for the whole catalogue
Relaunching one listing is done by hand. Relaunching a catalogue is not.
There is such a pipeline as a ready template: the n8n template “Studio photos and videos for product listings” turns an ordinary product shot into a studio photo and a short clip for marketplaces — no photo studio. A set of listing images (Ozon, Wildberries, Amazon) is built by the skill “Marketplace listing photo generation” through the AI SKILLS connector.
What that changes in the project economics. A set of listings costs $245–730 per product on the market. Preparing one item by hand takes hours. The pipeline processes a whole catalogue, and your fee stops being capped by the number of items.
Plus the templates from the “Earning with ready templates” course — for marketplaces and e-commerce. Descriptions, specifications and listing texts across the whole catalogue at once: drag the formula down a thousand rows and a thousand items are processed.
And the order of work stays the same: first the funnel analysis and fixing the listing, then scaling to the catalogue, and only then advertising. The conveyor speeds up the second step but does not remove the first.
The n8n template is advanced: Full plan, or $9 one-off, or 1000 XP; the marketplace-pack skill file comes from the Basic plan; the “Earning with ready templates” course — Basic.
A set of listings in minutes
You can describe the task in plain words in Cerebrum — the right Gen AI models are picked automatically. The "marketplace listing" task: you attach a link to the product photo and get the full set for Ozon, WB and Amazon: main image, angles, lifestyle, infographics, 3:4 format.
Three adjacent tasks for the same audience: a product showcase — a cinematic 5–15 second clip with no person on camera, a UGC ad with a character — a vertical clip "as shot by a customer", creatives for paid ads — 6–9 images for A/B testing. Single generations are in the Gen AI section.
You pay from the balance on usage, and the price is visible before launch.
Assistants by platform, Basic plan: Wildberries architect, Ozon content master, Yandex Market expert, Avito assistant. Separately, Visual merchandiser — how a listing looks in the results next to competitors, and AI inventory manager for the goods side.
Doing it by hand — the lesson “Working with files. Analysing marketplace product cards” in the “Automator” course, Plus plan.
Skills for relaunching listings
All three skills below are advanced: from the Creator plan, or $9 one-off, or 1000 XP.
- Customer review analysis — where a relaunch begins: sentiment, a priority matrix, a ranked list of complaints. The infographic is made from that list rather than from guesses
- Collecting reviews from shops — the reviews themselves with ratings, dates and verified-purchase markers
- Seller profile analysis — an analysis of competitors: since when, how many reviews, what share is positive
The order that changes the result. The first item first, then everything else. A relaunch started from the pictures is new packaging around the same complaints; the reviews have already written down what needs fixing.
What to take on the platform for this task
- E-commerce marketing — an assistant: marketing on Wildberries, Ozon, Yandex Market, Lamoda; from the Basic plan
- Scheduled review requests with a Postgres log — an n8n template: regular review requests to customers; Creator, Make, N8N, Full
- Ad creative package for targeting — a Claude skill: 6–9 creatives for an A/B test through the AI SKILLS connector (Full plan); the skill file comes from the Basic plan
- UGC ad from a product photo — a Claude skill: a vertical clip "as shot by a customer"; the file comes from the Basic plan
What to read next
AI for marketplace sellers — the listing funnel analysed by stage.
AI for designers — how to assemble a set in one style.
A 3D model from a product photo — what to add when photographs are not enough.
Thirty posts in one launch — how to process a whole catalogue at once.
n8n and Make automation templates — the full guide to automations.