Acorn vs Skai: You Are Comparing Two Different Purchases

One sells labour, the other sells a licence

Comparison articles put these two in the same table because both appear in retail media conversations, but the invoices buy opposite things.

• Acorn-i is an agency. Its own site, read on 20 August 2026, describes "A Technology Led Ecommerce Agency" offering an analytics product called Ignite alongside Expert Services covering ecommerce strategy, content, advertising management and consumer insight. Founded in 2018 by Amazon alumni, acquired by The Brandtech Group in 2022, and announced in November 2023 as being fused with Jellyfish to form Jellyfish Commerce, with its co-founders leading that unit. When you sign, people go to work on your account.
• Skai is a platform. You license the software, connect your accounts and run campaigns through it. The company still files under its original corporate name — the footer of its own site reads "Kenshoo, Ltd" — which is a useful reminder that a rebrand and a legal entity are different things when you are reading a contract.

So the first question is not which is better. It is: who is going to sit in the console on a Tuesday afternoon and move the bids? If the answer is nobody yet, a licence is a cost with no operator attached to it.

One disambiguation, since this corner of the market attracts it: Amazon DSP is the demand-side platform — programmatic display, video and audio. Amazon also uses DSP for Delivery Service Partner, the courier franchise. Everything here is about advertising.

Skai publishes a rate card, and most of this category does not

That deserves saying plainly, because transparency is rarer here than it should be and it is worth crediting where it exists. The tiers below were read on the vendor's own pricing page on 20 August 2026, are quoted in US dollars, and are billed annually — there is no monthly toggle on the page, so there is no second tab to misread.

• Standard — $114k per year, for advertisers spending up to $4M per year
• Advanced — $276k per year, for advertisers spending up to $10M per year
• Enterprise — $504k per year, for advertisers spending up to $20M per year
• Enterprise Premier — $756k per year, for advertisers spending up to $35M per year
• Enterprise Premier + — custom, above that band

Two things follow. First, the tiers are banded by your media spend, which means the licence scales with your budget in much the same direction a percentage fee does — the structure differs less than the vocabulary suggests. Second, prices move. Check the page yourself before you plan a budget around any figure printed anywhere, including here, and check which billing tab you are reading on any vendor including us.

Acorn-i publishes nothing. Stated neutrally: they quote on a call, which is the ordinary practice for agencies in this market.

The comparison nobody runs: licence plus salary versus managed fee

A licence is never the whole cost of the software route, and a managed fee is never only the media buying. Run the arithmetic on your own numbers rather than on anyone's brochure:

• The software route — annual licence, plus the fully loaded cost of at least one competent trader, plus onboarding and integration time, plus the cover when that person is on holiday or leaves. Retail media specialists are not easy to hire quickly.
• The agency route — fee, plus whatever is billed on top. Creative production, clean-room work and audience development sit inside the fee at some firms and outside it at others, and that difference is usually larger than the difference between two headline rates.

The honest tie-breaker is not cost, it is continuity. Software you own keeps working when someone resigns; the knowledge of how to use it does not. An agency keeps the knowledge but rents it to you. Most brands that thrive on the licence route already had a strong operator before they bought the tool, and the tool made that person faster. Buying the console first and hoping it creates the operator is the failure mode worth avoiding.

Who each genuinely suits better than we do

Skai suits a brand running many retail media networks in-house. If you are buying across a long list of publishers and want one console, one taxonomy and one set of reports rather than eight logins, a licensed platform is the correct architecture and no agency should talk you out of it. At the larger spend bands the licence also stops looking expensive relative to percentage-based alternatives, which is a calculation you should run before speaking to anyone.

Acorn-i suits a brand whose problem is content and commerce together. Analytics-led positioning, founders who came out of Amazon, and a group behind them with creative production at scale. If the honest diagnosis is that your listings, assets and data are all mediocre at once, that is a full-service brief and a display specialist is the wrong hire.

We suit a narrower case. reMKTR buys Amazon DSP and nothing else. If your Amazon account is already in reasonable shape and the open question is whether display is creating sales or merely being credited with them, that is our whole job.

Worth stating rather than glossing: on published pricing, Skai is ahead of us. They print tiers; we publish a percentage-of-media model and quote the rate before you sign. Several firms in this market are more transparent than we are, and pretending otherwise would be the sort of claim this page exists to discourage.

What reMKTR brings that neither purchase includes

reMKTR is the Amazon DSP arm of Full Circle, a full-service Amazon management company with $500M+ in managed spend across 100+ brands. We run 109 live Amazon DSP advertiser seats.

Across 30 of those advertisers in July 2026 — a slice of the book, named as such — the portfolio returned 6.04x on ad spend, ran 78.4 million impressions at a $4.00 CPM and a blended $1.42 cost per click, and delivered a $5.49 cost per acquisition across 57,137 attributed purchases, of which 20.1% came from shoppers new to the brand.

The number that matters is not the multiple, though. It is what the multiple is allowed to mean. Last-click attribution cannot prove incrementality and never could; holdouts and matched controls can. Reconciling in Amazon Marketing Cloud is where display and sponsored ads stop claiming the same order, and a console — any console — will happily report a healthy return without ever answering that question. Software reports. Someone still has to design the test.

A five-question script for the build-or-buy decision

Before either proposal lands, answer these internally. They decide the outcome more reliably than any feature comparison, and answering them late is how brands end up owning a console nobody drives.

• Who is the named operator, and are they already employed? If the answer is "we would hire one", add the recruitment timeline and the risk of a bad hire to the software route's cost. Retail media specialists are not quick to find and are expensive to replace.
• What happens when that person is away for three weeks? Cover is the unglamorous problem with in-house operation. An agency has a bench by construction; a one-person team does not.
• How many retail media networks are genuinely in play in the next eighteen months? One channel does not justify a platform licence. Five or six changes the answer completely, because the value of a licence is consolidation and consolidation needs something to consolidate.
• Who owns the measurement design? Software will report whatever you configure it to report. Somebody still has to decide what gets withheld, from whom, and for how long. If nobody in-house will own that, buying the console does not get you an answer — it gets you a dashboard.
• What is the annual total on each route, on your real budget? Licence plus fully loaded salary plus cover, against fee plus whatever is billed on top. Then run it again at twice your current spend, because whichever route you choose you will be living with it through growth.

Two of these questions have nothing to do with either vendor, which is the point. The most common error in this comparison is treating it as a product evaluation when it is a staffing decision wearing a product evaluation's clothes.

Where to go next

Dr. DSP — when you like the idea of buying this capability as a defined product with a scope rather than retaining an agency. It is the closest thing we have to a middle path between a licence and a retainer.

Dr. PPC — when sponsored ads are the real leak. Display amplifies whatever the account already does, so if search spend is going to terms that produce no orders, fix that before adding a display budget on top of it.