Яндекс Маркет

Work Market sales, prices and boost bids right in chat

Connect Yandex Market to the AI agent: campaigns, orders, products, prices and stock. The agent summarizes sales by campaign and breaks down returns and refusals.

How the agent works with Яндекс Маркет

The agent works with the Market cabinet over a Yandex OAuth token and sees both planes at once — the cabinet level and the individual shop level: orders, stock, prices, sales boost, reviews and financial reports. It handles Market's first trap itself: some Partner API methods expect the business identifier, others the campaign identifier, and the wrong level returns "not found" on live data. The second trap is date formats: different methods expect different ones, and the agent supplies the right one for each.

Orders come either for a single shop or for every shop in the cabinet as one feed — no manual walk over campaigns. The "impressions → cart → order → delivered" funnel and the buyout rate per SKU come from a single asynchronous Sales Analytics report instead of hundreds of per-campaign calls: where naive automation explodes on dozens of shops, the agent has one report. Market rate-limits report generation hard, so the agent schedules the runs and waits for readiness instead of hammering the endpoint.

Financials come from the official reports rather than from re-adding orders: monthly turnover and every Market service from united-orders, weekly settlements from united-netting. Market's Excel exports with their summary sheets and Russian headers are parsed by sheet content, not by sheet order — so the next reshuffle of sheets in the export does not turn into scrambled numbers. Market hands those reports over asynchronously: the agent queues the job, waits for the finished file and parses it itself, with no manual export from the cabinet.

Sales boost is the full cycle: current bids per SKU, recommendations for a target share of impressions, and applying them. Market takes bids in kopecks — the agent will not be off by a factor of a hundred. Bulk operations (prices, stock, bids, review replies) run in chunks with Market's quirk in mind: one bad item rolls back the whole batch, so the agent checks the result of every chunk and retries exactly the failing ones. It can reply to reviews in bulk from a template generator, but never auto-replies to negatives — complaints need a human.

Some events arrive without polling: the agent listens to Market notifications about order status and substatus changes, delivery date and order composition changes, cancellation requests, expired stock, price quarantine and offer card status changes. It diagnoses key permissions honestly: "no data" from Market often means missing access rights rather than an empty result, and in that case the agent asks you to reconnect the integration with the right permissions instead of reporting zero sales. Any change — prices, bids, replies — is shown as a plan first and applied only after your confirmation.

Integration scenarios

Sales summary across Market campaigns

The agent collects orders and revenue across every campaign and fulfillment model (FBY/FBS/DBS) and shows the trend week by week.

Price and stock control

The agent reconciles Market prices and stock against your accounting system and finds the discrepancies and the listings hidden from sale.

Unredeemed orders and returns

The agent analyzes unredeemed orders and returns for the period, groups them by reason and counts what the logistics cost you.

Sales boost bids

The agent shows the current bids per SKU, works out recommendations for a target share of impressions and applies them only after your confirmation — in kopecks, never off by a hundred.

Reviews with no reply

The agent collects fresh reviews, answers routine questions and thank-yous from a template and hands the negative ones to you — complaints never go into an auto-reply.

Alerts from the cabinet

The agent posts to chat the moment Market sends a cancellation request, expired stock, a price quarantine or an offer card status change — with no polling on a schedule.

Яндекс Маркет examples

Яндекс Маркет
Frozen chats — click a question
Разбери невыкупы за июнь.

Невыкуп 11,3% против 8,1% в мае, потери на обратной логистике 187 тыс ₽.

  • Основной рост дала кампания FBS — там невыкуп 15,7%
  • Половина причин — «передумал» по товарам дороже 6 тыс ₽
  • 42 возврата пришли с пометкой о повреждении упаковки

How to connect

Before you connect

  • A Yandex account with access to the Market partner cabinet
  • At least one campaign in the cabinet — FBY, FBS or DBS
  • A role above content manager — it cannot open the API section
  1. 1

    Pick a long-lived Yandex account

    Access is issued to a specific Yandex login, not to the account. Use an account the business will keep: a password change, a 2FA change or recovery revokes it.

  2. 2

    Invite the login at partner.market.yandex.ru

    The owner invites the login into the partner account at partner.market.yandex.ru and sets a role per shop. The agent sees only campaigns it can reach.

  3. 3

    Check the role opens «API и модули»

    The section is open to the account admin, manager, shop admin, developer and operator roles. A «Контент-менеджер» (content manager) cannot see it.

FAQ

Through Yandex OAuth: you sign in with an account that has access to the partner cabinet, so there are no API keys to copy. There is also the option to connect with a static API key from the cabinet.

Campaigns, orders, products, prices and stock — the Partner API sections your account has access to. Permissions follow your account's role in the cabinet.

Yes. The agent sees every campaign in the connected cabinet and can roll them into a single report — comparing FBY and FBS by margin, for example.

Skills for Яндекс Маркет

Бренд-аналитика на маркетплейсахАнализ бренда на маркетплейсе: доля рынка, динамика выручки vs категории, brand health, защита от контрафакта, конкурентное позиционирование.
Аудит комиссий и удержаний маркетплейсовСправочник всех типов удержаний Wildberries, Ozon и Яндекс Маркет. Маппинг названий из отчётов, нормальные диапазоны, алгоритм выявления аномалий, шаблон претензии при расхождении. Используй при анализе финансовых отчётов маркетплейсов, проверке корректности списаний и подготовке претензий.
Мониторинг конкурентов на маркетплейсахСистематический мониторинг и анализ конкурентов на Wildberries, Ozon и Яндекс Маркет: идентификация конкурентов, ценовой мониторинг, SEO-позиции, ассортимент, отзывы, рекламная активность. Матрица реакций и чеклист регулярных проверок. Аналог конкурентных модулей MPStats, Moneyplace, SellerStats.
Кросс-площадочное сравнение маркетплейсовСравнение маркетплейсов для конкретного товара: комиссии + логистика + трафик + конкуренция + возвраты + реклама. Скоринг площадок, юнит-экономика на каждой, рейтинг и мультиплатформенная стратегия.
Аналитика рекламы на маркетплейсахКомплексный анализ рекламных кампаний на маркетплейсах (Wildberries, Ozon): ключевые слова, ставки, бюджеты, юнит-экономика, ценообразование, минусация, оптимизация ДРР.
Стратегия рекламы на маркетплейсахМетодология управления рекламой на Wildberries, Ozon и Яндекс Маркет: выбор типа кампании, распределение бюджета по ABC-анализу, стратегии ставок (авто vs ручная), расчёт предельного ДРР/ROAS, масштабирование кампаний, привязка рекламы к жизненному циклу товара, сезонная адаптация, чеклист готовности карточки к рекламе.

Related integrations

Category
Marketplaces
Connection
Device code

Connect Яндекс Маркет

Connecting takes a couple of minutes, and your data stays in your own systems.

Connect