Why teams switch paid media platforms

Companies in these interviews change who runs their paid media more often than the platform it runs on, and about half of those changes bring the work in-house. Tool swaps are fewer and cluster in programmatic buying, where buyers cite fees, results they could not prove and complexity. The moves, the pressures, the events, and why many stay.

Based on verified interviews with the performance, growth, paid search and paid social leaders who run paid media, at DTC and enterprise brands. Buyers are anonymized before publication; vendor names and views are reported as given. No vendor paid to appear or could edit this page.

Teams switch paid media platforms less often than they change who runs the work, and when the platform does change, it is often because the arrangement around it did. Three kinds of change are easy to confuse, so this page keeps them apart. A tool change replaces or drops a programmatic buying platform or a management layer for bidding, operations or creative analysis. An operating-arrangement change moves the work between agencies and internal teams, or from one agency to another, and may change the tools along the way. A vendor discontinuation can force a replacement or set off an urgent evaluation. The three describe the form of a decision and what prompted it.

The tools differ in kind. A DSP (demand-side platform) such as The Trade Desk or Display & Video 360 buys programmatic inventory. Management platforms run operations across the channels they support: Search Ads 360 manages search campaigns across engines, and Smartly builds and runs ads across social, connected TV and other channels. Motion focuses on analysing which ad creative performs. Replacing one is not always like for like: a company may narrow its channels, move to a managed service, consolidate, or stop a capability altogether. Moving budget between channels or publishers without retiring a tool is a channel decision, not a software switch, and stopping a whole channel and dropping its dedicated tool is a different case from replacing that tool with a competitor. Advertising on retailer sites is covered on why brands switch retail media tools. How companies organise paid media in the first place is on what paid media management software is, and the companion read is what paid media buyers wish they’d known, where the ratings live.

About half
Of completed changes in who runs paid media here brought the work in-house; about a third replaced an agency. Shares of the changes observed, not industry rates.
Fewer
Completed swaps or drops of a buying or management tool than changes in who runs the work. Most are in programmatic.
7.5/10
The Trade Desk’s buyer rating. It is the platform named most often on both sides of the programmatic moves here.
What changed in paid media moves, by kind of move Bringing paid media in-house is the most common completed change, followed by moving to a new agency and then by swapping or dropping a buying or management tool. Handing in-house work to an agency and moves set off by a vendor discontinuing a product or announcing a wind-down are the least common. Agency reviews make up the largest group of moves still planned or under evaluation. KIND OF MOVE HOW OFTEN, RELATIVE TO THE OTHERS Brought the work in-house search, social, programmatic, or all Moved to a new agency often one agency in place of several Swapped or dropped a tool mostly DSPs, some management layers Handed the work to an agency from an in-house team Set off by a vendor’s notice Xandr’s DSP, Marin Software Completed Under way, planned or under evaluation WHO RUNS THE WORK CHANGES MORE OFTEN THAN THE PLATFORM
Each bar’s length shows roughly how many companies are classified under that kind of change relative to the others, each counted once by its latest interview. The three bars on who runs the work count each company under its main change, so they do not overlap; the tool bar is counted separately, so a company whose tools changed with its agency can sit there too. Dark is completed; light is under way, planned or under evaluation, and a search for a replacement with no destination chosen sits in the light segment, not among completed switches. These are observed decisions, not churn rates or market shares.

Do teams switch the platform, or who runs it?

Who runs it, more often. Among companies describing a completed change in who runs their paid media, about half brought some or all of the work in-house, about a third changed agencies, and a smaller group handed internal work to an agency. That is the mix of changes these companies made, not the share of advertisers that make them in a year. In-housing is rarely all or nothing: a regional health system says “about three years ago, we had all of our performance media outsourced with an agency of record, we brought that in-house,” and keeps a budget for one-off agency work; a media and entertainment company has absorbed agencies but still uses some “on the tactical side,” for hands-on campaign setup. A multi-brand home-furnishings retailer in-housed programmatic first, replacing partners it calls “more black box, fully managed service,” and “Did the same thing for paid social a few years later.”

The agency moves often consolidate. Several brands moved to one agency covering what several had, or to a larger agency for more staff and data capabilities, and one global brand moved its agency of record to another office of the same agency because it was cheaper. The move in the other direction comes from capacity, not dissatisfaction with an in-house team’s results: a travel company moved paid search to an agency because of “staff challenges, single points of failure, and just skill gaps that meant we were missing out on opportunities.” Several more companies are mid-review, and an agency review is the most common decision still open here.

The pressure that builds: why teams switch paid media platforms

Completed tool moves are few enough here to describe one by one rather than rank. Three pressures recur among them and among the evaluations still open. A complaint that sits beside a move did not necessarily cause it, and one decision can carry several.

Slow-burn · #1

Fees out of proportion to the spend

The fee complaint is about spend set against the cost of reaching and running a platform, and that cost comes in layers. A DSP can charge for the platform itself, for data and measurement, and for a managed service; an agency or reseller can add its own fee or markup; and a platform that is economical at one budget may not be at another. The buyers here are reacting to different layers, not reporting that any one platform is expensive in general. A digital health subscription company left a TV ad-buying service because “the fees were really large for us in comparison to what we were spending,” and says running MNTN itself instead meant “We saved a ton and we were much faster, much more efficient.” A DTC apparel brand walked away from The Trade Desk because “I can’t spend enough on one to make it worth it,” across the inventory sources it opens up, and works directly with Tatari for TV. Some fees are not the platform’s: a global spirits company is “looking at any type of tool that an agency is using at the moment that we’re paying for, for a markup that we could bring in house,” The Trade Desk among them, though it says it is not actively doing so yet. What the DSP, the agency and any reseller each charge is lesson seven on the lessons page.

Named in this context: MNTN · Tatari · The Trade Desk

Slow-burn · #2

Results the buyer could not prove

Some departures followed a buyer’s inability to show enough value, which is not the same as a controlled finding that the platform had no incremental effect. An online jewelry retailer left The Trade Desk because “we were unable to prove” that its heavy spend there was bringing incremental value in awareness and in reaching customers earlier. A global automaker offboarded Amazon DSP after early successes: “It’s just not a mature platform if you compare it to DV360 and Meta’s platform, etc.,” and the overall return “wasn’t worth it.” The same company tests smaller programmatic partners and drops the ones “not delivering the results that we needed.” A luxury specialty retailer applies the test to agencies instead: “we’ve recently continued to fire our paid agencies because we don’t think they’re truly moving the needle.” These are the reasons the buyers gave, from their own reading of their own results, not independent measurements.

Named in this context: The Trade Desk · Amazon DSP

Slow-burn · #3

A platform too complex for the team running it

A DSP needs an operator, and complexity becomes a reason to switch when the team running it lacks the capacity, expertise or support, whether the gap is in the product’s usability, the managed service or the staffing. The jewelry retailer above says “The Trade Desk’s platform is very complicated for any new user and even for experienced users,” and expected the managed service to make up for it; its suggestions were “very generic.” A B2B office-technology company took the opposite route: “as my paid media expert became more familiar with our strategy, with our tools, etc., we offloaded our agency,” and it is now evaluating a DSP to support that one specialist. Search tools draw the same question from the other side, whether a layer on top of Google Ads still earns its cost; that story is on what paid media management software is, along with buyers leaning toward dropping Search Ads 360 or Skai and one that left Smartly. Those are leanings and a single departure, not a trend.

Named in this context: The Trade Desk · Search Ads 360 · Skai · Smartly

That's the average buyer. How does your stack measure up?A free, peer-powered read of your tech stack, with a shortlist and your next steps.
Get my stack read →

The event that fires the decision

Several completed moves name an event rather than a complaint about the platform. The counts are small, and an event can accompany a move without having caused it.

Trigger · #1

A new agency brings its own tools

When the agency changes, the tools often change with it, because the agency holds the seats. A travel company that moved paid search to an agency says “they use Search Ads 360, so we use that.” A home-wellness brand trialled Motion and liked it, but did not subscribe because it “partnered with a new agency that has their own Motion, shares reports with us.” A global hotel group lost something in its move: “We were previously working with a different agency where we had complete access to our Facebook Ads Manager, but now we do not have that access,” and sees a dashboard instead. A distributor’s move from Display & Video 360 to StackAdapt came with a change of agency too; it is told on the lessons page. Whether tools follow people more broadly is on do marketing tools follow people?

Named in this context: Search Ads 360 · Motion · Meta Ads · StackAdapt

Trigger · #2

The work comes in-house, and the stack is rebuilt around the team

In-housing creates tool decisions of its own. The regional health system above says executing media itself meant going out for “probably eight tools,” among them Google Marketing Platform and The Trade Desk. A DTC beauty brand brought media buying in-house and restarted Motion, because “in the past agencies didn’t use it at all, so it’s kind of wasteful.” A DTC footwear brand says “we’ve recently brought media buying in-house, saved us a ton of money,” and runs it with an AI media-buying tool at what it puts at “a tenth of the cost we were paying before,” rating that tool six and a half. A national specialty retail chain gives a reason that is about data, not cost: an agency once used, by accident, an audience it was sent with another client, and the company then brought “all of our media buy in house.”

Named in this context: Google Marketing Platform · The Trade Desk · Motion

Trigger · #3

A vendor ends a product, or says it will

Two vendor events here prompted a rethink, and neither is a competitive win for a destination. One ended in a completed DSP replacement. Microsoft announced on May 14, 2025 that it would stop supporting media buying through its Invest DSP, the former Xandr DSP, from February 28, 2026, and a media company that buys programmatic for its clients says its new DSP “replaced Xandr when Xandr got decommissioned.” The other set off an urgent search that the interviews do not show finished. Marin Software’s board approved a plan to dissolve the company in April 2025, subject to a stockholder vote, and a life insurer that used Marin for search was looking for a replacement weeks later: “They have ceased operations, so they’re shutting down in 10 days. So, it’s caused us a little bit of work.” That was the buyer’s understanding at the time, and its interviews show no replacement chosen. Marin did not dissolve. It filed for Chapter 11 in Delaware on July 1, 2025, its board abandoned the dissolution plan, and the company said it remained fully operational and expected no impact on its customers. Its reorganization plan was confirmed on August 29, 2025 and took effect on September 5: the old shares were cancelled, the new ones went to Kaxxa Holdings, the plan sponsor, and to the lender that financed the case. The insurer’s search shows what an announced wind-down sets off among customers, not that Marin stopped serving them.

Named in this context: Xandr · Microsoft Invest · Marin Software

Where do teams switch to?

The Trade Desk appears on both sides of the programmatic moves. A biopharma company says “We consolidated our programmatic media buying platform to The Trade Desk. We were previously on multiple platforms,” a telecom brand chose it over Google’s display and video buying, and a specialty insurer moved to it from an older DSP a few years ago. That insurer is now weighing whether to “move away from The Trade Desk” for “lower fees and a better algorithm,” possibly through an RFP, while keeping Amazon as its secondary DSP; nothing has been decided. The jewelry retailer and the apparel brand above left it. That makes it the platform named most often in these moves, not one gaining or losing share, and its users rate it 7.5 out of 10.

The other destinations are specialised arrangements, often chosen to cut cost: MNTN run self-serve for connected TV, Tatari bought directly for TV, StackAdapt for programmatic. The interviews do not show them as equivalent to what they replaced in inventory, function or terms. And some moves have no platform destination at all. A specialty fashion retailer says “we used to use MNTN as our CTV provider, and we don’t anymore,” because it stopped running connected TV, not because it chose another tool. A consumer-electronics maker that “internalized all paid media efforts” changed who runs every channel without a platform change being named. For how the programmatic platforms rate against each other, see the lessons page.

Want your own stack read the same way?

Get mine →

The stories behind the switches

An online jewelry retailer bought through The Trade Desk with its managed service and found the help thinner than it expected: the service did what it was told but brought no ideas of its own about what worked for other customers, on a platform the buyer found hard even for experienced users. The bigger gap was evidence. It had hoped the platform would show that its spend reached customers earlier and brought them back to buy later, and it could not show that. It stopped using The Trade Desk. Senior director, performance marketing · online jewelry retailer
A travel company had run a large paid search program in-house and moved it to an agency, and the agency’s tools came with it. The reason was the pace of the work: “the difficulty of keeping in-house staff up-skilled to the required level was basically that the industry was moving too quickly.” It now runs Search Ads 360 and Display & Video 360 because the agency does. Senior performance marketing manager · travel company
A DTC apparel brand used The Trade Desk for connected TV, display retargeting and a test of programmatic audio. What worked best was retargeting, which it thinks it could run itself, and paying extra for viewability measurement was “another cost we would have to incur.” So it moved TV to Tatari directly and audio to a small podcast agency: “I don’t really need Trade Desk right now to have access to all these different inventory sources because I can’t spend enough on one to make it worth it.” Director of paid media and retention · DTC apparel brand

The counter-current: moves considered and dropped

Several companies weighed a move and kept what they had. A national credit union looked at a measurement tool to stand in for work its agency does and stayed with the agency, because “Even if we get the tool in-house, we’ll have to give access to the agency to use it.” A B2B software company looked at moving from the Google Display Network to a full DSP, but The Trade Desk’s fees were “just, I think, too much,” and a change in budget made a big DSP pointless. A faith-based nonprofit is still deciding: it is in an early evaluation of The Trade Desk to see whether it should bring buying in-house, a decision it expects to run into 2027. These are reasons to stay or to wait, not a measure of loyalty to any vendor.

What this means if you’re weighing a switch

In these interviews the decision before a paid media platform change is usually who will run the work, and several tool changes followed an agency change or a move in-house rather than a complaint about the tool. Where a complaint came with a completed tool move, it was fees against spend, results the buyer could not prove, or a platform too complex for the team. The Trade Desk sits on both sides of the programmatic moves. Microsoft’s end to its Invest DSP forced one DSP replacement, and Marin’s announced wind-down sent another buyer searching before Marin reorganized under Chapter 11 instead. What the evidence does not show is that in-house teams outperform agencies or the reverse, that any vendor is in decline, or that a higher-rated platform is safer from departure. These are small counts of completed moves, and results are the buyers’ own reading of their own programs.

What follows is our recommendation, not a practice buyers describe.

  1. Settle the operating arrangement first. Decide what stays with an agency and what moves in-house, who owns the ad accounts and platform contracts, who runs campaigns day to day, and who owns reporting, measurement and quality checks. A hybrid is a legitimate answer.
  2. List what has to transfer, and who owns it. Ad account access, campaign history, conversion events and pixels, audience definitions, creative, report exports and contractual access to the platforms. Separate what belongs to you from what belongs to the agency or reseller, and note what will have to be rebuilt. Test permissions and exports, not just dashboard access.
  3. Price the whole operation, not the platform. Media spend, DSP or management fees, data and measurement charges, agency and managed-service fees, any reseller markup, and the staffing, training and transition cost of the change. Where a platform does not suit your spend, price a narrower or managed arrangement beside it.
  4. Agree the success measure before moving budget. Use the same outcomes and definitions on both sides, and a holdout or geographic test where you can run one. A before-and-after return on ad spend is distorted by changes in budget, targeting, creative, season and channel mix, so write down what the comparison cannot show.
  5. Plan the handover. Settle who pauses and who launches, how tracking stays intact, when access changes hands, and how the two sides reconcile during the overlap. If a vendor is ending the product, confirm the end date and what you must export before you lose access.

For the ratings, see what paid media buyers wish they’d known; for the causal question, what incrementality testing is.

Common questions

Why do companies switch paid media platforms?

In these interviews the reasons buyers give for completed tool moves are fees out of proportion to their spend, results they could not prove were incremental, and platforms too complex for the team running them. Several moves follow an event instead: a new agency that brings its own tools, work coming in-house, or a vendor discontinuing a product. Completed tool swaps are few, so these are cases, not a ranking of causes.

Should we bring paid media in-house or use an agency?

Both moves happen. Among completed changes in who runs paid media in these interviews, about half brought the work in-house, about a third moved to a new agency and a few handed in-house work to an agency. Companies that came in-house cite control of data and strategy, cost and agencies that were not moving results; the ones that went to an agency cite skill gaps and the scale of day-to-day management. These are observed decisions, not evidence that either arrangement performs better.

Why do companies leave The Trade Desk?

The departures here give three reasons: an online jewelry retailer could not prove its spend brought incremental value and found the platform complicated, a DTC apparel brand could not spend enough in any one inventory source to justify it, and a specialty insurer is weighing an RFP for lower fees and a better algorithm. The Trade Desk also gained companies, through consolidation onto one platform. It rates 7.5 out of 10 among the buyers who rate it.

Does changing agency mean changing ad tools?

Often, in these interviews. A travel company that moved paid search to an agency took on the agency’s Search Ads 360, a home-wellness brand dropped its own creative-analysis trial because the new agency shares reports from its own seat, and a hotel group lost direct access to its Meta ads account under a new managed agency. A brand that brought buying in-house restarted a tool its agencies never used. Who holds the seat decides what stays when the agency changes.

What happened to Marin Software and Xandr?

Each was a vendor event, not a competitive loss. Microsoft said in May 2025 that it would stop supporting media buying through its Invest DSP, formerly Xandr, from February 28, 2026, and a media company here replaced it with another DSP. Marin Software’s board approved a plan to dissolve the company in April 2025, and a life insurer in these interviews began looking for a replacement for its search tool; the interviews do not show one chosen. Marin did not dissolve. It filed for Chapter 11 on July 1, 2025, abandoned the dissolution plan and said it remained fully operational, and its reorganization plan took effect on September 5, 2025, cancelling the old shares and issuing new ones to its plan sponsor, Kaxxa Holdings, and its lender. Neither move is a vote against the product by its users.

Your stack, read by the buyers who've run it

Peer-powered: built from hundreds of verified buyer interviews, and fitted to the tools you already run. You get a shortlist for your stack and your next steps.

Get my stack read

Free · no password · your answers are anonymized before they inform research like this.

Want to come back to this?We'll email you a link to this page, so it's easy to find later or forward to your team.
Methodology. Alium conducts verified interviews with software buyers — the marketing, growth, ecommerce and media leaders who run these platforms. This page aggregates what buyers in that corpus, through October 2026, say about replacing, dropping or adding a programmatic buying platform, a campaign management, bidding or creative-analysis tool, and about changing who runs paid media: a new agency, a move in-house, or a move from in-house to an agency. Moves were read one company at a time in the buyer’s own words, with the interviewer’s question above each answer, and each company is counted once, by its latest interview. In the changes of who runs paid media, each company is counted under its main change, so those three shares add up; tool changes are counted separately, so a company whose tools changed along with its agency can appear in both. A completed move is one finished at the buyer’s current company; moves at a previous employer are left out. Completed moves, migrations under way, open evaluations, plans, evaluations that kept the incumbent, and moves prompted by a vendor discontinuing a product or announcing a wind-down are kept apart; a search for a replacement with no destination chosen is not a completed switch. A company choosing its first platform is a new selection, not a switch. Budget moved between ad platforms, such as from Meta to TikTok, is a channel decision and is outside the scope, as is advertising on retailer sites, which has its own page. Shares are computed over completed changes in who runs paid media; tool moves are too few for shares and are described as cases. Reasons are editorial codings of what buyers said, and a move can carry several or none. Vendor facts come from the vendors’ own announcements and product pages and from Marin Software’s SEC filings, checked in November 2026. The one rating is a buyer-satisfaction average from published transcripts; the full table is on the paid media lessons page. Buyer identities are verified at interview time and anonymized before publication; vendor names are reported as given. No vendor paid to appear or was able to edit this page.