ChannelAdvisor review — half the corpus reviews a company that no longer trades under that name

One definition, then the thing worth knowing

A definition to clear away first, because this site is about a different species of software: DSP here means demand-side platform, the buying system behind programmatic display, video and audio. It is not the Delivery Service Partner courier programme Amazon also runs under those initials, and it is not a marketplace channel manager. We are Dr. DSP, an Amazon demand-side platform product from Full Circle, a full-service Amazon management company with more than $500M in managed revenue across 100+ brands. We do not compete with the platform this page is about, which is part of why we can be straight about it.

Now the fact that reorders your reading. ChannelAdvisor was taken private by CommerceHub in November 2022. On 12 December 2023, CommerceHub, ChannelAdvisor and Dsco were announced as a single brand: Rithum. That announcement is still published on rithum.com, and we re-read it on 20 August 2026. Insight Partners holds the majority position, and the ChannelAdvisor name survives in one visible place — a login link for existing customers.

So the corpus a searcher gets handed spans roughly a decade, two ownership structures, two brand identities and several product generations, mostly without saying which. That is nobody's fault; it is what happens to reviews when a twenty-year-old company changes hands. It does mean the first skill you need is not judgement. It is dating.

Ninety seconds to date any review you find

Four markers place almost any review on the timeline without needing a publication date — useful, because dates on review round-ups are often the date of the last light edit rather than the date of the experience.

• The name used, and how. "ChannelAdvisor" with no mention of Rithum places the experience before late 2023, or the writer downstream of an older source. "Rithum (formerly ChannelAdvisor)" means the writer checked. "Rithum" alone, with ChannelAdvisor as a footnote, is the most recent layer.
• Whether Dsco or dropship appears. The dropship and supplier-network side arrived with the CommerceHub and Dsco lineage. A review treating the product as marketplace listings only is describing something narrower than what exists now.
• The support anecdote. Any story about a named account manager or an onboarding hand-off is a snapshot of one team on one date. Support tracks staffing and process, and both are re-cut in any merger.
• The module vocabulary. Feature names date a review harder than prose does. If the AI layer is unnamed, the review predates it; RithumIQ is the current name on the vendor's own pages.

One more habit: check whether the reviewer names a channel count. Coverage only moves one direction, so a coverage figure is a rough carbon date on its own. Rithum's December 2023 announcement put the network at 420+ marketplaces and retail sites, more than 40,000 companies on the platform and over $50B in annual gross merchandise value. A review citing materially fewer channels is describing an older platform.

Different parts of a platform review have wildly different half-lives

This is the part nobody writing round-ups says out loud, and it beats any star rating. A software review is not one claim. It is a bundle of claims that decay at completely different rates, and an old review is perfectly readable once you know which parts to keep.

The volatile half — months, not years. Support and account management is the first thing restructured after any acquisition and the thing reviewers write about most emotionally in both directions. Commercial terms get rewritten every renewal cycle. Roadmap and AI claims describe a plan rather than a product. Interface screenshots age faster than anything and say almost nothing about whether the system works.

The durable half — years. The data model is architectural: how the platform represents a product, a variant, a channel override and a business rule rarely changes, and it decides whether your catalogue fits or has to be bent. Feed architecture is the same — batch versus near-real-time, how errors surface, how reprocessing works — and it causes most day-two frustration. Channel coverage grows and rarely shrinks, so old reviews understate it. Implementation shape barely moves: a platform that needed a project manager and a quarter in 2021 needs a project manager and a quarter now.

Practical rule: mine old reviews for the durable half, conclude nothing about the volatile half from them, and verify that half by talking to a customer. Which brings us to the questions.

A large share of the reviews you will find answer a different purchase

The second structural problem has nothing to do with dates. Rithum sells marketplace listings, inventory and order management, product feeds, dropship and private marketplaces, commerce reporting, and retail media advertising over the top. It is a channel-management and commerce-operations platform. It is not an Amazon demand-side platform, and it does not claim to be one.

We read the Rithum retail media product page on 20 August 2026. It describes managing campaigns across top retailers such as Amazon, Walmart and Target, connected to product, inventory, pricing and margin data. The words demand-side platform and programmatic do not appear on it. That is a category boundary stated in the vendor's own copy, not a shortcoming.

The consequence is direct. If someone put this platform on a shortlist next to display-buying options, most of the reviews you are about to read are grading feed sync accuracy, variant handling, order routing and marketplace error queues. Those are excellent things to grade, and none of them tell you anything about buying display impressions against Amazon shopping signal — a separate purchase with separate inventory, separate economics and a far harder measurement problem. Two purchases, one shortlist, and the corpus speaks to only one.

What this platform is genuinely strong at, and who should buy it instead of anything we sell

We would not pitch against Rithum for most of what it does, because for most of what it does it is a serious answer and we do not do it at all.

• Breadth of connection. Twenty years of marketplace and retailer integration is not a feature list, it is accumulated edge cases — the tax quirk on one marketplace, the variant rule on another, the retailer whose feed spec moved last spring. That is the hardest thing in the category to replicate quickly, and it is the real asset.
• Dropship and supplier networks. The retailer side — private marketplaces, supplier onboarding, delivery-date and shipping-label work — is a genuinely different capability from marketplace listing, and few vendors carry both.
• Operations at volume. Order, inventory and feed plumbing that holds up across many channels at once is unglamorous and extremely valuable when the catalogue is large and the channel count is high.

Who should buy it rather than us, plainly: any brand or retailer whose bottleneck is listing, catalogue, feed or order operations across many channels; any retailer building a dropship or private-marketplace programme; any business where "we cannot see our inventory in one place" is the sentence that starts the meeting. We solve none of those. If that is your problem, stop comparing us against them and go and read the operations reviews properly, dated using the method above.

Six reference-call questions no review site can answer for you

Each of these is answerable by a real customer in a sentence and unanswerable by a review site, because the answers are specific to a date, an account size and a channel mix. Ask for two references: one over three years in, one live under a year. The gap between them is the roadmap, and it is more honest than any roadmap slide.

• Who is on our account today, and how many other accounts does that person carry? The fastest-ageing variable in any review, answerable in one number.
• How long from signature to your first order flowing correctly — and what slipped? Not the quoted timeline. The actual one, and the reason for the gap.
• What did you end up building yourself around the platform? Every platform has a seam where customers start writing their own scripts. Knowing where it sits beats any feature comparison.
• Which of your channels is the integration weakest on? There is always one. A reference who cannot name it is not using it at your depth.
• What happened at your last renewal, and how much written notice of any change did you get? Ask this of every vendor including us. Written notice of fee changes is a reasonable thing to want in a contract.
• If you left tomorrow, what could you actually take? Catalogue in what format, channel overrides and business rules or not, how much performance history, over what window.

Then trial-import your twenty worst SKUs — the bundles, the odd variants, the regional attributes — rather than a demo catalogue. Demo catalogues are built to import cleanly. Yours is not.

The one thing no review will ever test, and where we come in

Here is the claim no review corpus can evaluate for any advertising product, ours included: whether the spend caused sales that would not otherwise have happened. Reviewers repeat what the dashboard reported. Last-click attribution names the touch nearest the purchase; it cannot establish that the purchase needed the touch. Display is where that gap opens widest, because display reaches people already on their way.

What closes it is a designed test, not a better report: hold display back from a matched set of ASINs or regions for a defined window, run a control, and reconcile in Amazon Marketing Cloud, where DSP and sponsored-ads events sit side by side and stop each claiming the same order. AMC is available at no cost to eligible advertisers. Ask anyone selling you display how they would design that test; the answer separates vendors faster than any star rating.

Our numbers, scoped honestly: a direct Amazon DSP API pull across 30 advertisers in July 2026 — not the whole book — showed 6.04x return on ad spend, 78.4 million impressions at a $4.00 CPM, $1.42 blended cost per click, $5.49 cost per acquisition across 57,137 attributed purchases and 20.1% new-to-brand. One month, Amazon-attributed, promised to nobody. On the buy itself, Amazon's documentation states no minimum for self-service DSP and a typical $50,000 minimum investment for managed service, varying by country.

Dr. DSP runs Amazon display as a managed product, Orbit included at no extra cost. You set how much runs without your click and can move that setting whenever you like; inventory risk, pricing, new products, new creative and any decision to stop display spend always come to a person. No published price — a demo, the first 30 days free, priced on the call, month to month. Two redirects: if the commercial structure is what you came for, our page on how this platform charges reads the historic filings in detail; and if your Amazon problem is sponsored search rather than display, Dr. PPC is the right product and publishes both its price and its cap.