What video advertising covers
Video advertising is any paid message delivered as moving pictures, placed in or around other video or in feeds where people scroll. That covers a skippable spot before a tutorial, a vertical clip between friends’ posts, a thirty-second break in a streamed series and a short bumper before a clip. The common thread is that someone pays for attention that would otherwise go to something else.
Predicting where this goes is risky. What follows describes the main formats, the main ways space is bought, and the measurement questions that matter. Where something is a direction rather than a settled fact, the text says so.
Common ad formats
Formats are usually described by where and how they appear.
- In-stream ads play before, during or after the video a viewer chose. They may be skippable or not, and they usually borrow the length conventions of television.
- Out-stream and in-feed ads appear inside a page or a social feed, often starting silently and relying on captions until the viewer taps for sound.
- Short bumpers and brief clips are designed to deliver one idea in a few seconds, which suits places where people expect fast content such as short-form clips.
- Television-style spots run inside streamed programmes, including on connected TVs, where the screen is large and viewers are often seated and less likely to skip.
- Interactive and shoppable units add taps, polls or product links on top of the video. Shoppable video connects the ad directly to a purchase path, and interactive ads trade simplicity for engagement.
None of these is universally better. A format that suits a quick reminder on a phone may be poor for explaining a complicated product, and a long spot on a big screen may be wasted on a viewer who is half watching a feed.
How ad space is bought
Behind each format sits a buying model, and the model affects price, control and risk.
Direct deals. An advertiser agrees with a publisher or platform to buy a set amount of exposure in specific programmes or channels, often at a negotiated price. This offers predictability and brand-safe placement, but less flexibility.
Automated auctions. Much online video is sold impression by impression in real time, with the advertiser’s bid, targeting and the expected quality of the ad all influencing who wins. How video ad auctions and targeting work is a topic of its own, but the key point is that price moves with demand rather than being fixed.
Pricing units. Buyers may pay per thousand impressions, per completed view, per click or per desired action, depending on the goal. Each unit shifts risk differently: paying per view moves some risk to the advertiser, while paying per action moves more of it to the seller.
Ad-supported streaming. As more services offer cheaper or free tiers funded partly by commercials, ad-supported streaming and FAST channels have drawn more advertisers into television-like inventory delivered over the internet. How large this becomes is uncertain and varies by country.
Targeting and the privacy question
Advertisers have long wanted to show the right message to the right person. Targeting methods range from broad ones, such as the programme being watched, to ones based on past behaviour or declared interests.
Rules and platform policies on tracking have tightened in many places, and they differ by country. The direction of travel is often described as less reliance on individual-level tracking and more on context (what the surrounding video is about), on data a brand collects directly from its own customers, and on aggregated reporting. How far each of these will go, and how well it works, is still being tested. Anyone running campaigns should check the rules that apply where their audience lives.
Measurement trends
Measurement is where video advertising is most contested, because the numbers that are easy to collect are not always the ones that matter.
- Delivery metrics such as impressions and completed views show that an ad was served, not that it was noticed. Definitions of a “view” vary between platforms, so figures are not always comparable.
- Attention measures try to estimate whether a person was actually looking, using signals such as whether the player was visible. These are a developing area and methods differ.
- Outcome measures link ads to site visits, sign-ups or sales. These are closer to business results but harder to attribute, since people see several messages across several screens.
- Incrementality tests compare a group that saw the ad with a similar group that did not, to estimate what the ad added. This is often considered a more credible approach, though it takes planning and enough volume.
Cross-platform measurement is a long-standing difficulty. A person may see the same campaign on a phone, a laptop and a television, and no single platform sees all of it. For the practical side of reading results, the topic of marketing metrics explains which figures are useful and which can mislead, and video analytics covers how platforms report viewing.
Creative trends
Several creative directions appear repeatedly in discussion of where ads are going.
- Sound-off design. Many ads are viewed on phones with the sound muted, so captions and clear visuals matter.
- Shorter, faster openings. Because viewers can often skip, the first seconds usually carry the main message.
- Native-feeling content. Ads that resemble the content around them, such as creator-style clips and brand deals with creators, often blend into feeds more naturally, though audiences may react poorly if the commercial relationship is unclear.
- Personalisation. Varying names, offers or scenes per viewer is becoming more feasible, though it adds production and privacy work.
- Interactivity. Taps, polls and branching stories can raise involvement, but they need to be judged against simpler spots rather than assumed to be better.
What remains uncertain
Several questions do not have settled answers. It is unclear how quickly ad-supported tiers will grow compared with subscription video, how targeting will work as tracking rules change, whether attention metrics will become a common currency, and how automated tools will alter the cost and volume of creative work. Anyone planning spending should treat forecasts, including any in this article, as informed guesses.
A sensible approach is to match the format to the goal, test small before scaling, choose a measure that reflects the real aim, and revisit assumptions as platforms and rules change. Trends are useful as prompts for experiments, not as guarantees.