Downstream cost in 2026: the technology still exists, the price list never did

What actually happened to Downstream

Downstream Impact was an Amazon advertising platform founded by two former Amazon advertising executives, and by the time it was acquired it was managing, on TechCrunch's reporting, roughly a quarter of a billion dollars a year in Amazon ad spend for large brands. In March 2021 Jungle Scout raised a substantial growth round and bought it.

What followed is the ordinary arc of an adtech acquisition, and it is why your search is confusing. Downstream ran for a period as Downstream by Jungle Scout. Then Cobalt 2.0 merged the market-insight side of Cobalt with the advertising side of Downstream into a single platform. Jungle Scout's site still hosts a Downstream by Jungle Scout guide, and the old standalone domain no longer resolves to a working product site. The technology is alive; the brand is a redirect.

This matters for the cost question in a very direct way. You cannot buy Downstream. You can buy Cobalt, and the advertising capability that was Downstream comes inside it, bundled with market insights and category data you may or may not want. Whatever Downstream cost on its own is now a historical fact, not a quote you can obtain.

It is worth noting this is the second time in this category we have written this paragraph about a different vendor. Adtomic is now Helium 10 Ads and its old add-on price no longer exists either. Ad tooling gets absorbed into suites, and the pricing page is the first casualty.

Why the figures in the roundups are not usable

Search Downstream pricing and you will find a monthly minimum quoted with confidence. Follow the citation and there usually is not one. The figure circulates between listicles, each one sourcing the last, and none of them source Jungle Scout.

There are three separate reasons that number cannot be trusted, and they compound:

• It describes a product that was retired under that name. Even if the figure was accurate when written, it priced a standalone platform that is now a module inside a different one, sold to a different buyer, on a different contract.
• Cobalt is quoted, not listed. Jungle Scout's own pages route brands and enterprises to a demo with a growth expert and describe custom plans. There is no tier card to screenshot.
• Even the self-serve half moves. Jungle Scout's public pricing renders in JavaScript and is tested on visitors. Read it repeatedly on one day and it will not always give you the same set of numbers, which is a normal commercial practice and a total obstacle to quoting it in an article.

So the accurate answer to what Downstream costs is: nobody outside the sales conversation knows, and anybody who prints a figure is repeating one. That is a finding rather than an accusation. Quote-based pricing is a legitimate model for an enterprise platform, where the deployment genuinely varies. It just means your comparison has to be built differently.

What the quote is actually covering

When you take the demo, the number you are given is a bundle. Cobalt is positioned as brand and enterprise software for businesses above roughly a million dollars in annual revenue, and the advertising capability inherited from Downstream sits alongside market share data, category tracking, competitor and pricing analysis, and share-of-voice measurement.

Five things typically move an enterprise marketplace quote, and they are the five you should ask about explicitly:

• Catalogue size. How many ASINs or parent items are in scope, and what happens when that number doubles.
• Retailers and regions. Amazon only, or Amazon plus Walmart, plus international marketplaces.
• Seats. How many people need to log in, and whether read-only viewers cost the same as operators.
• Ad spend. Whether any part of the fee moves with the budget you push through the platform, and if so, at what rate and up to what ceiling.
• Data out. Whether API access or a warehouse export is included or priced separately. This is frequently the line that surprises people.

If the quote contains a percentage of ad spend, that percentage is the real price and the monthly figure is the deposit. On a hundred thousand dollars of monthly ad spend, a single undisclosed percentage point is twelve thousand dollars over a year, and it scales up exactly as the account improves. Get it in writing, get the bands, and ask where it stops.

How to compare a quote against the platforms that publish

You can still run a fair comparison. Build the two published anchors first and put the quote next to them.

Helium 10 prints both halves of its model on its own page, which makes it the cleanest anchor in the category: Diamond at $359 a month, or $279 billed annually, and Diamond customers incur a 2% management fee on PPC spend managed through Helium 10 Ads. That is a complete, checkable cost.

Perpetua publishes Essentials at $695 a month up to $10,000 of monthly ad spend, and Growth at $695 a month plus a percentage of ad spend above that threshold — the percentage is not printed, so you have half the model and must ask for the other half.

Now hold the quote against those. Not on features, on shape. Is the fee flat, banded or variable? Does it move with spend, with catalogue, or with seats? Does it stop? What is the notice period, and what happens to your historical data if you leave?

A useful test: ask the salesperson to price you at twice your current ad spend as well as your current one. The difference between those two numbers is the part of the deal that will actually determine whether you are still a customer in two years, and it takes thirty seconds to ask.

Where Cobalt is genuinely the right buy

We would send some readers there without hesitation. Jungle Scout's underlying strength has always been category and market data, and Cobalt is the enterprise expression of it. If your question is what share of your category you hold, how that has moved quarter on quarter, which competitors gained it and at what price points — that is a real capability, it predates the Downstream acquisition, and the advertising layer sitting on top of it means you can act on the finding in the same tool.

Brands that report to a board on category share, and multi-brand groups tracking a portfolio against competitors, get more from that combination than they would from a focused advertising product. If that is your requirement, take the demo and negotiate well.

If your requirement is that somebody make and defend the daily advertising decisions, market-share data is interesting context rather than an answer.

What we charge, printed, and why

Dr. PPC is $300 a month plus 3% of ad spend, capped, month-to-month, with the first 30 days free and Orbit included rather than sold on. It is autonomous Amazon ad management from Full Circle, a full-service Amazon management company with $500M+ in managed revenue across 100+ brands.

We charge a percentage because most of this category does, including the market leader on its own pricing page, and because it keeps our incentive attached to the account rather than to the contract. The two things we insist on are that the number is printed and that it stops. An uncapped share of spend charges you most at the exact moment you are winning.

What the fee buys is the deciding, not just the executing. Fable 5 reads the whole account, writes a strategy per product against its real economics, and proposes each change with the evidence behind it, a measurement plan and a rollback trigger. You choose whether it proposes only, acts within limits, or acts freely on categories you have approved.

The kind of thing that surfaces: across 47 brands on our own book, 48.5% of search spend — $4,962,963 of $10,243,379 since 1 May 2026 — went to terms that produced no orders at all. Every enterprise platform in this comparison could render that number for your account this afternoon. Someone still has to go and work it, every week, which is the part no platform sells.

One redirect. If the advertising is already efficient and the real ceiling is how many new shoppers ever encounter you, Dr. DSP is the reach conversation rather than the bid one.