Tinuiti Pricing: How an Enterprise Agency Quote Gets Built

What is on their site, and what is not

We read Tinuiti's own Amazon services page rather than a roundup of it. There is no fee, no retainer range, no minimum spend and no rate card anywhere on it. The commercial content is a contact form, which is entirely normal at this end of the market and worth stating without any edge to it.

What the page does tell you is the shape of the offer, and that shapes the price. The Amazon practice covers sponsored ads, Amazon DSP, Amazon Marketing Cloud, Sponsored Brands and video, Stores, A+ content and Amazon Performance Plus. That is a full-stack retail media practice rather than a bid-management shop, and a proposal will reflect the breadth of the team behind it whether or not you use every part.

Plenty of specific retainer figures circulate online for this agency. Nearly all of them sit on competitor blogs whose business model is ranking for this exact keyword. We are not reprinting numbers we cannot verify against Tinuiti's own material, and neither should the spreadsheet you take to your finance director.

One term worth disambiguating for anyone who arrived here from a broader search: Amazon DSP in this context is Amazon's demand-side platform for programmatic display and video advertising. It is unrelated to an Amazon Delivery Service Partner, which is a logistics franchise. This page is about the advertising one.

The trap in agency pricing articles: market ranges are not rate cards

Before you compare any figures at all, learn to spot the single most common error in this genre — one we have made ourselves and had to correct.

Agencies publish industry pricing guides. Those pages look exactly like pricing pages: same layout, same confident numbers, same URL structure. But the numbers on them describe what the market charges, not what the publisher charges. A figure lifted from one of those guides and reprinted as a named agency's rate card is a false claim about that agency, and it is a claim that appears in a great many articles about agency pricing.

The test takes five seconds. Is this the price they charge, or the price they say the industry charges? If the surrounding sentence frames it as a range for the category, it is the second, and nothing about it belongs in a column headed with a company name.

The related discipline: check which billing view you are reading before quoting anything, on any vendor including us. Monthly and annual tabs on the same page produce different numbers, and reading the wrong one is where most confident pricing errors begin.

Where a firm genuinely does publish, say so. Darkroom lists a starting monthly figure for every service line on its own pricing page — Amazon management from $6,000/mo, in US dollars, read 20 August 2026. On pricing transparency that is ahead of most of this category and ahead of us: reMKTR quotes a percentage per account rather than publishing a rate card, and only Dr. PPC carries a public price across our group.

Three ways agencies price, and what each does to behaviour

Whatever number arrives, it will be built on one of three structures. The incentive each creates is more useful to understand than the digit.

• Flat retainer. Predictable, easy to budget, indifferent to your spend. The risk runs both ways: a fixed fee against a fast-growing account gets quietly under-serviced, and against a shrinking one it starts to feel punitive at exactly the moment cash is tight.
• Percentage of media spend. The fee moves with the account, which ties the agency's revenue to growth. reMKTR charges this way, so we are not going to pretend it is the wrong model — but it needs two safeguards written in: a rate that steps down as spend rises, and a review trigger if spend grows faster than results.
• Hybrid. A base plus a percentage, the most common enterprise structure and the one where cost is easiest to lose track of, because the two components usually get negotiated in separate conversations weeks apart.

The honest argument against our own model, which we would rather raise than have raised at us: a flat fee banded by revenue can genuinely beat a percentage when your ad spend is large relative to your revenue. Some vendors price that way on purpose. If that is your shape, say so and make us argue for the cap rather than for the structure.

Then add the fees that sit outside all three: onboarding and setup, creative production, marketplace expansion, and anything described as a "project". Ask whether each is one-time or repeats on scope change.

Five inputs that decide where the number lands

You can shape all five before a proposal is written, and doing so is worth more than negotiating afterwards.

• Media under management. The first question any agency asks and the biggest single lever. Give both your current figure and your twelve-month projection, and ask for both prices.
• Channel count. Amazon alone is one scope. Amazon plus Google, Meta, CTV and programmatic is a different relationship with different staffing and a different number.
• Creative. Display and video production is real work and is frequently a separate line. Ask whether it is in or out, and what "in" covers per quarter.
• Catalogue complexity. Hundreds of ASINs across variations and marketplaces consumes more hours than a tight range does, and hours are what you are buying.
• Measurement scope. Amazon Marketing Cloud work, incrementality testing and custom reporting are often priced separately from campaign management. This is the line most likely to be assumed in and billed out.

Get each confirmed in the written proposal rather than in the meeting. Scope written down is what you can hold people to at renewal, and renewal is where the money is.

Read the contract before you read the credentials deck

Every article about agency pricing compares fees. Almost none of them discuss the document that determines what your second year costs, which is a strange omission given that the second year is where most of the money is.

Four clauses do the work: term, auto-renewal, notice period, and what you keep on exit. A published example from this category, cited neutrally because a real one is easier to reason about than a hypothetical: Quartile's terms and conditions, read 20 August 2026, set an initial term of "one year from the Effective Date", automatic renewal "for successive one-year periods", non-renewal by "written notice of non-renewal at least sixty (60) days prior to the expiration", fees "due in advance of Services rendered", and no entitlement "to any refund in connection with any termination". Month-to-month is available only as a written exception.

Nothing there is improper. It is an ordinary enterprise agreement and versions of it sit behind most vendors on your list. The finding is that it is published, decisive, and cited by no one ranking for these keywords.

For an agency specifically, add four more:

• Who owns the Amazon DSP advertiser seat and the ad account permissions — you, or them?
• Who owns the Amazon Marketing Cloud instance and the queries written in it?
• Do you keep historical reporting and audience definitions after offboarding?
• Will they give written notice of fee changes, with how many days? That is a small ask and it protects you against the most common surprise in this business — and it is a fair thing to ask us for too.

Ownership in this category has moved a great deal recently, which makes change-of-control and price-protection language reasonable rather than paranoid. Ask what happens to your rate and your team if the firm is acquired mid-term.

The awkward one: they already sell our differentiator

The measurement argument is the centre of what we sell, so it would be dishonest to write a page about this particular agency without naming the obvious.

Tinuiti publishes a measurement practice called Bliss Point, with a productised Incrementality Lab inside it — real-time experimentation to isolate causal lift, sitting alongside media mix modelling and holistic ROAS. That is substantially the argument we make about display, from a much larger firm, and a well-prepared buyer will put it to us. They should.

Our position does not need them to be wrong. It rests on this: last-click attribution cannot prove incrementality and never could, holdouts and matched controls can, and the industry is better off the more firms say so. What is left to compete on is narrower and more useful to a buyer than any claim of uniqueness:

• Where does the measurement sit? Is the test designed and computed by the party being measured, and if so, what process manages that conflict? Ask who builds the holdout, who holds the query, and who sees the result first.
• Is it in the statement of work? A capability described in a pitch and a deliverable with a named cadence and an owner are different purchases. Ask for the clause.
• Does it span both channels? Sponsored ads and DSP claim the same orders constantly. Reconciling them in Amazon Marketing Cloud is what stops you adding two overlapping numbers and calling the total performance.

Ask those three of every agency on the shortlist, in the same words, and compare the answers rather than the decks. Ask them of us first.

The question nobody raises on the call: are you a big account here?

Tinuiti is one of the largest independent performance marketing agencies in the United States, and that size is a genuine advantage. Bench depth, platform relationships, people who have seen your problem at ten times your scale, and the ability to staff a launch in weeks rather than quarters — none of that can be faked by a small shop.

It also means the operating model is built around large budgets. If your Amazon spend would place you in the bottom quartile of their book, you are likely to be staffed accordingly, and the senior people in the pitch are unlikely to be the ones in your account every week. That is not a criticism. It is arithmetic that applies to every large agency, including firms we admire and including the parts of our own group that have grown.

So ask it plainly, and ask it of everyone:

• Where would our budget sit in your client base — top, middle, bottom?
• Who specifically works on this account, and what else do they hold?
• How many accounts does that person carry?
• What changes if our spend halves for two quarters?

A good agency answers all four without flinching, and the flinch is the information.

How we are priced, and what the fee covers

reMKTR is narrower on purpose. We run Amazon DSP as a managed service on our own seats and charge a percentage of media spend that is stated before you sign. We are part of Full Circle, a full-service Amazon management company with $500M+ in managed spend across 100+ brands, and we hold 109 live Amazon DSP advertiser seats.

The scoped evidence, from a live Amazon DSP API pull across 30 of those advertisers during July 2026: 6.04x return on ad spend for the set, 78.4 million impressions at a $4.00 CPM, a blended $1.42 cost per click, $5.49 cost per acquisition on 57,137 attributed purchases, and 20.1% of those purchases from shoppers new to the brand. Thirty advertisers, one month, labelled as such rather than presented as the whole book.

For context on what else you could buy: Amazon's own DSP page states that its managed-service option "typically requires a minimum investment of USD 50,000", with self-service available if you employ a trader. An agency holding seats is the middle path, generally with more flexible minimums and someone accountable weekly.

Two redirects we would make on a call. If the work is broader than advertising — listings, catalogue, retail operations, the things that decide whether ads can work at all — Full Circle is the parent and takes the whole scope. And if the margin problem is stockouts, fees and reimbursements rather than media efficiency, no fee structure fixes that; Dr. Stock is the right instrument.