Most Amazon teams already talk about Amazon in Slack or Microsoft Teams. Someone pastes a screenshot from Seller Central, someone else asks whether it includes Vendor, a third person opens the ads console to check spend, and by lunchtime the thread has five numbers that do not agree. Viktor is built for exactly that kind of thread: instead of a person fetching the numbers, an AI colleague does it and posts the answer where the question was asked.
This guide covers what Viktor is, what DataDoe gives it, how to connect the two, which requests are worth putting on a schedule, and how to keep control of anything that would change your account.
Viktor is an AI employee made by Zeta AI. A company adds it once to its Slack or Microsoft Teams workspace (Discord and a web app work too), and anyone on the team can mention @Viktor in a channel or a direct message. Behind the chat, Viktor has its own cloud computer where it writes and runs code, so a request can end as a PDF, a spreadsheet, a chart or a small internal app rather than a paragraph of text.
Three things make it different from a chatbot in a sidebar:
DataDoe and Viktor work together, and DataDoe is in Viktor's integration catalogue. Once it is connected, Viktor works from the whole Amazon business rather than one report at a time:
The rest of the picture, with examples, is on our Viktor integration page.
Then go to the channel where Amazon is discussed and ask a first question. Something you already know the answer to is ideal, because it tells you straight away whether the numbers line up with what you see in Seller Central.
Viktor will answer a vague question, but a precise one becomes a routine you can trust. Four details matter more than anything else:
Once the first answer is right, ask Viktor to save it as a skill with those definitions. That is what turns a lucky answer into one that arrives the same way every Monday.
"Each morning at 7:30, post yesterday's orders and sales per marketplace against the same weekday last year. Mark anything more than 20% off in either direction." Short, comparable and quick to read on a phone.
"Every Friday, list Sponsored Products search terms with more than $30 spend and no orders in the last 14 days, grouped by campaign." This is a review list, not an automatic negation; a person decides.
"Twice a day, check our top 30 ASINs for a lost Featured Offer and say how far below us the winning offer is." Pair it with a pricing rule later, once the alerts look trustworthy.
"After each settlement, show what Amazon paid, what it deducted and anything that differs from the previous settlement by more than 10%." Finance teams like this one in their own channel.
"On the first working day of each month, build last month's profit by brand and marketplace as a PDF, with the definitions on the last page." The definitions page saves an argument every month.
"Every weekday, summarise new Vendor Central purchase orders and compare the quantities with the last four weeks." It catches the quiet drop in ordering that otherwise shows up weeks later in revenue.
Viktor and DataDoe each have a lock, and a change on Amazon needs both to be open.
A sensible start is read only for the first week. Then switch on one Action type, usually ad bids, keep Viktor on ask for confirmation, and widen it only when the approvals have been boring for a while.
An agency can add one DataDoe connection per client, name each one after the client, set them to invite-only, and tell Viktor which connection a client channel uses. Clients connect their own Amazon accounts through a DataDoe invitation link, so nobody at the agency handles their logins, and each account can be read only or read and write. The rest of the agency setup is on our agencies page.
If your team prefers Claude inside Slack, the same DataDoe connection works there; see Claude in Slack for Amazon. If you run the business mostly from your phone, Boski does the same by text message. Vendor-specific data is described on our Vendor Central page.
No. You authorize Amazon in DataDoe on Amazon's own pages, and Viktor signs in to DataDoe.
Only if the connection's access allows it. Set it to invite-only to keep it to the people who should see the numbers.
Yes. The connection belongs to Viktor, so it works wherever Viktor does: Slack, Microsoft Teams, Discord and the web app.
DataDoe fetches most tables daily at 2:00 in the seller's local time, with some tables updated during the day. Viktor reads whatever DataDoe holds at the moment of the request.
DataDoe is $97 a month after a 14-day free trial, with every Amazon account included. Viktor is billed separately in credits shared across the workspace and starts free.
Viktor is a product of Zeta AI, Inc., and a DataDoe partner. DataDoe has no affiliation with Amazon and is not endorsed by Amazon. The Viktor steps follow Viktor's documentation as read on 12 November 2026 and may change.
.svg)
