Magnite, which operates sell-side advertising platforms for publishers and streaming services, published a survey of 883 adults in Sweden in early October 2026, reporting that 99% engage with streaming, mobile apps, digital audio or news at least twice a week and that 54% pick product relevance as a reason they would recall an ad, the most-chosen of eight factors. The English-language materials give no fieldwork dates, sampling method or margin of error, and the survey was run by a company whose platforms sell advertising inventory in the channels it measured.

In Short

Magnite, a company that helps websites, apps and streaming services sell their ad space, asked 883 people in Sweden how they watch, listen, play and read online, and what makes them remember an ad. Advertisers and publishers are the audience for the findings, which Magnite frames as a case for buying across streaming TV, apps, audio and news together - and Magnite sells ad space in those same places. The published files do not say when the survey ran or how large the margin of error is, so the results show what these respondents said about themselves rather than a precise count of all Swedes.

What Magnite published and how the sample was built

Magnite posted the study, titled Decoding the Connected Consumer: Sweden 2026 and subtitled Understanding Omnichannel Ad Engagement and Impact, on its website in early October 2026. A 22-slide report accompanies the landing page, with English and Swedish versions listed. Neither document carries a publication date.

According to Magnite, the fieldwork was done with Bounce Insights, which it describes as a global research technology firm. Respondents were adults aged 18 and over who completed a 15-minute online survey delivered in Swedish. To count as digitally connected, a respondent needed internet access and at least one media-related activity from a list covering linear television, streaming video, social media, gaming, streaming audio, mobile apps and digital news or magazines. The sample was therefore screened for media use at the outset.

The respondent count appears in two forms. The landing page and the appendix slide give 883; the second slide says over 800. The appendix age bands - 9% aged 18-24, 25% aged 25-34, 22% aged 35-44, 21% aged 45-54 and 27% aged 55 and over - add up to 104%, while gender splits 49% male and 51% female. The report says Bounce's platform "ensures high-quality, reliable data from engaged, representative audiences", yet it describes no quotas, weighting, panel source or field dates, and no margin of error appears in either document.

The design follows a pattern. Magnite's Japan study, covered on November 26, 2025, surveyed 800 consumers aged 16 and over with a different research partner, Milieu. The Spain edition was fielded in December 2025 among 872 adults who watch ad-supported streaming. Later work covered 920 adults in Hong Kong and, in a live-streaming study, 835 adults in the United States, 884 in Australia and 899 in New Zealand. Sweden's 883 sits inside that band of 800 to 920.

What the 99% figure measures

The key-takeaways slide states that 99% of Sweden's digitally connected population "engages with streaming, mobile apps, digital audio, or news on a regular basis". The landing page words it differently: 99% "uses a combination of streaming, mobile apps, digital audio, and digital news at least twice per week". The first version describes at least one channel; the second implies several.

The channel table on slide 7 points firmly to the first reading. Penetration among internet users runs from 62% for digital news to 75% for streaming video, so a figure of 99% can only be a union of channels rather than a single-channel reach. Slide 6 reproduces the question behind it, which asks which activities respondents "frequently engage in for entertainment at least twice a week". A footnote adds that mobile apps include gaming and service apps such as finance, shopping, food, communication and delivery. With communication and finance apps in the count, the threshold for being active is not demanding.

Slide 5 scales the result to a national number: 9.9 million Swedes described as being on the open internet, a term Magnite defines as active use, twice a week or more, of streaming, gaming, audio and digital content. The footnote labels the figure a Magnite estimate built from the Bounce survey, a population estimate from Census.gov and internet penetration rates from Statista. The arithmetic is not shown.

The report then sets the 99% against two walled gardens measures: 79% active on social media and 52% active on video user-generated content on YouTube and TikTok. Its recommendations slide describes unified activity across premium environments as "significantly outperforming the reach of single-channel activations such as social or UGC". The comparison places a union of several channel groups against single categories, and no significance test is reported. On Magnite's own table, no single open-internet channel exceeds 75%, which is below the 79% recorded for social.

Time across channels

Slide 7 supplies the channel-level data, reproduced here as Magnite presents it. Penetration carries the label household penetration, although the survey interviewed individuals, and hours are among users of each channel.

ChannelPenetration (among internet users)Hours per day (among users)Using more than last year
Streaming video75%2.548%
Gaming apps69%2.640%
Mobile service apps71%2.1not given
Streaming audio73%Music 2.1, podcasts 1.448%
Digital news62%1.240%

Adding the per-channel hours gives 11.9 a day. That points either to heavy overlap between activities - listening while commuting, say, or playing while a programme runs - or to generous self-reporting; the report does not address it. The landing page condenses the table into two lines: more than two hours a day on gaming apps, streaming video, mobile service apps and streaming music, and more than one hour on podcasts and digital news. The final column records only one direction of change, and the share using less is not given.

Time of day and place

Slide 9 asked respondents to pick up to three active periods for each channel. Streaming video concentrates in the evening, at 73% for 5pm to 9pm and 53% after 9pm, against 8% before 9am. Music peaks at 38% in the early evening and 34% in the afternoon. Digital news has the strongest morning showing of any channel, at 30% before 9am, while podcasts are flatter, between 16% and 24% across the day. Magnite lists the living room and the bedroom as the main places for watching, and commuting and the kitchen as the main places for listening; mobile apps and digital news, according to the report, fill the gaps in between.

Shared screens and second devices

Two slides carry headlines asserting that co-viewing amplifies the impact of ads shown on CTV and that second screeningenhances their effectiveness. The data beneath them record attitudes and reported habits, not measured ad outcomes.

Co-viewing

On the first of those slides, 77% of respondents say they are more likely to co-view on a larger screen, 67% more likely to co-view television and films than user-generated video, and 51% more likely to co-view live content than on-demand. Among those who do co-view, the average is one other person. Some 38% say they discuss ads with others while co-viewing and 25% say ads feel more relevant. The deck does not state whether those last two shares are based on all respondents or on co-viewers only.

Viewing groups are larger elsewhere in Magnite's research. The Spain study found 74% of connected TV streamers always or often watching with others, with an average of three co-viewers per session. In Hong Kong, 72% co-viewed with an average of three other people, and the US figure in the live-streaming study was 69% with three additional viewers on average. Sweden's average of one other person is a third of those numbers, although question wording and bases differ across the studies. In Japan, 42% of co-viewers discussed ads and 33% found them more relevant, against 38% and 25% in Sweden.

Second devices

On the second slide, 59% say they are always or often using a second device while streaming on CTV. The landing page describes the same 59% as viewers regularly using a smartphone. Social media (68%), messaging (52%) and internet browsing (48%) lead the activities, and 25% are online shopping. The landing page frames that 25% as a share of the 59%, which would equal roughly 15% of all viewers; the deck's slide names no base. Magnite presents the shopping share as an opening for shoppable ads.

The comparison markets point in a similar direction. In Hong Kong, 56% used a second device and 37% shopped online during viewing, and the US live-streaming figure was 55% on a second screen.

Reported actions by channel and ad format

Slide 8 maps a purchase funnel for each channel: discovery (searched or discussed a product), consideration (saved a product or clicked a link) and action (added to cart or purchased). The base is described as those who have engaged with each ad format; the question wording is not shown.

ChannelDiscoveryConsiderationAction
Streaming video57%32%26%
Gaming apps37%37%22%
Mobile service apps46%35%24%
Streaming audio44%28%19%
Digital news47%32%20%

Streaming video leads on discovery and action. Gaming apps lead on consideration, where the 37% matches their own discovery rate, so the funnel does not narrow between its first two stages there. Digital audio gets a separate breakdown on slide 18: music at 36%, 23% and 15%, podcasts at 41%, 24% and 16%. The combined audio row above sits higher than both, slide 18 states no base, and the deck does not explain the gap.

Emerging formats

Slide 15 covers three newer formats: shoppable ads, pause ads and tile ads. Shoppable ads record 33%, 26% and 19% across the three stages, pause ads 34%, 23% and 12%, and tile ads 31%, 23% and 13%. A note says the results rest on all streaming video viewers, including those not yet exposed to each format, and that "greater exposure may increase reported actions". The base differs from slide 8, so these rows and the channel-level streaming video row (57%, 32%, 26%) are not like for like.

What respondents say drives ad recall

On slide 10, respondents chose from a list of factors that make an advertisement more likely to be recalled. Relevant products took 54%, personalisation to the respondent's interests 45% and entertainment value 41%. Brand familiarity, being shown across multiple platforms and including audio each took 25%; being shown across multiple devices took 16% and interactivity 11%. The shares add up to 242%, so each respondent could pick several, and the numbers describe stated belief rather than measured ad recall. The key-takeaways slide words the 54% as the factor most likely to make consumers remember an ad; because the question allowed multiple answers, the figure shows that relevance was chosen most often, not that it was chosen alone.

The ranking sits awkwardly beside the report's recommendations, which stress activity across platforms and, in the second-screen slide, interactive units. Both items rank low on respondents' own list. Magnite's earlier studies reported recall as a rate rather than a ranking: 87% of CTV viewers in Spain could recall details from recent sessions, and 66% of US live-streamers could recall an ad from their latest session. The Swedish deck reports no equivalent rate.

The seller behind the survey

Magnite sits on one side of the market it describes. The report closes with four recommendations: activate across video, audio and mobile apps; invest in premium large-screen streaming environments; use mobile and audio to sustain engagement; and reach audiences through what it calls its unified platform.

Magnite's results show how central streaming video has become to the business. In the first quarter, reported on May 6, 2026, CTV passed 51% of contribution ex-TAC for the first time. In the second quarter, reported on August 5, 2026, CTV contribution ex-TAC rose 36% to $97.1 million and international revenue grew 23% to $50.4 million. The same coverage records Dentsu and Magnite extending an EMEA partnership into Sweden, building on the July 2025 deployment of SpringServe across dentsu Total TV.

The Swedish market has its own figures. IRM data prepared for IAB Sweden put programmatic spending at 4.7 billion kronor in 2024, up 12%, with video at 2.4 billion kronor overtaking banners at 2.3 billion for the first time. Projections ran to 5 billion kronor in 2025 and 5.4 billion in 2026, with video at 52% of programmatic spending by 2026. Podcasts show a different pattern: programmatic podcast spending was 25 million kronor in 2024, 6% of total podcast advertising, with a projection of 48 million kronor for 2026. Set against the 1.4 hours of daily podcast listening that Magnite's respondents report, that 6% indicates that most podcast advertising in Sweden is still traded outside programmatic channels.

For the trade, the Swedish edition adds a data point rather than a verdict. It documents behaviour that buyers and sellers largely assume already - heavy evening streaming, widespread second-device use, a stated preference for relevance - in a sample screened for media activity, with self-reported measures and, in several places, bases that are unstated or inconsistent. Whether 99% reach across the open internet translates into delivered impressions in Sweden is not something a survey of this kind can establish.

Timeline

  • 2024 - IRM data for IAB Sweden put programmatic video spending at 2.4 billion kronor, ahead of banners at 2.3 billion for the first time (source).
  • July 2025 - Dentsu and Magnite deploy SpringServe across dentsu Total TV in European markets (source).
  • November 26, 2025 - Magnite's Japan study of 800 consumers aged 16 and over is covered (source).
  • December 2025 - Fieldwork for Magnite's Spain study among 872 adults; results were covered on March 10, 2026 (source).
  • May 6, 2026 - Magnite reports first-quarter results, with CTV above 51% of contribution ex-TAC for the first time (source).
  • August 5, 2026 - Magnite reports second-quarter results, with CTV contribution ex-TAC up 36% to $97.1 million and the Dentsu EMEA partnership extended into Sweden (source).
  • August 20, 2026 - Magnite's live-streaming study across the United States, Australia and New Zealand is covered, with US respondents accepting 8.7 minutes of ads per hour (source).
  • Early October 2026 - Magnite publishes Decoding the Connected Consumer: Sweden 2026, based on 883 respondents and a 22-slide report.

Summary

  • Who: Magnite, which operates sell-side advertising platforms, with research partner Bounce Insights; 883 digitally connected adults in Sweden were surveyed.
  • What: Magnite published Decoding the Connected Consumer: Sweden 2026, reporting that 99% engage with streaming, mobile apps, digital audio or news at least twice a week, that streaming video users spend 2.5 hours a day, that 59% use a second device while streaming on CTV, and that 54% select product relevance as a recall factor, the most-chosen of eight. The materials give no fieldwork dates, sampling method or margin of error.
  • When: Early October 2026, according to when the files appeared on Magnite's website; neither document carries a publication date, and the fieldwork dates are not stated.
  • Where: Sweden, with the report and landing page published on Magnite's website in English and Swedish versions.
  • Why: According to Magnite, the aim was to understand how digitally connected consumers engage with ad-supported media and advertising across streaming video, mobile apps, digital audio and news. The report's recommendations point to cross-channel buying, and Magnite sells inventory in those channels, which is the context for reading the figures.