Skip to main content

For most streamers, the numbers just aren’t adding up.

For years, Netflix promised to never use advertising. And yet, here they are adding a cheaper, ad-fueled pricing tiers in hopes that it will not only attract more viewers who cannot afford premium prices but will also earn more revenue from advertisers. A recent bump in subscriptions and stock price seems to endorse the decision– but this uptick hasn’t erased the steep 2022 drop.

While some analysts bet Netflix’s move into advertising will be enough to put the business in the black, the consensus still leans cautious. And Netflix isn’t the only streaming giant facing problems with profitability. Disney’s streaming service Disney+ announced it would be adding an ad tier, as well, and Hulu recently increased monthly prices for both ad-free and ad-supported price tiers. The question many experts are asking now is…

Is adding an ad-tier enough to fix profitability?

The answer is probably not. The use of ads to offset growth stagnations is viewed by many as the ultimate answer to their issues. However, there are mixed views from experts; some have grown increasingly critical of the business model, pointing out genuine issues with the new decision to include ads.

The first issue for streamers is ad spending minimums. Netflix currently requires a $20 million minimum ad spend from advertisers, severely limiting their number of potential partners– especially on a global scale. Temporarily, this may not be an issue: even the now-defunct Quibi secured $100 million in ad commitments before launching. But even if Netflix has no issue finding advertisers, this minimum may limit the company’s ability to scale its ad revenues in the long run.

A look at today’s most successful digital companies reveals that small businesses are critical for growth. Meta and Google– companies notorious for their sizable advertising clientele– have not only attracted a great diversity of advertisers, but the bulk of these advertisers are small and local businesses. For Netflix, then, requiring a large minimum ad-spend could directly threaten one of the most significant bases for scale and profitability.

Another issue is retention. Research shows that viewers struggle with advertisements on streaming services. One survey found that 69% of US adults find streaming services ads repetitive, and 79% are bothered by the streaming ad experience.

It’s therefore no surprise that some experts have already predicted the demise of ad-fueled tiers. Pivotal Research principal and senior analyst Jeff Wlodarczak states Netflix is taking a significant risk with the new tier. The average revenue per user (ARPU) and product perception could suffer for a prominent streamer and is higher risk than reward. In Wlodarczak’s words, Netflix programming is simply “not optimized” for ad breaks.

Even though ad-fueled tiers appear promising on the surface, consumer disapproval and a lack of inclusivity for small business advertisers signal major headwinds for organizations looking to change their business model. Such a pivot may end up doing streamers more harm than good.

Is SVOD Dead Forever?

Subscription streamers’ catastrophic losses this year have made one thing resolute: “growth” is not a sustainable model in the wake of market saturation, increased competition, and a vulnerable economy. But for streamers relying solely on subscriptions– referred to in the industry as “SVOD’s”– the only familiar alternative rev-stream seems to be commercial ads. It’s a Catch 22: “growth”-strategies (usually in the form of costly pushes to create more content) are plateauing as subscriptions start to saturate, but commercial ads, with their heavy threats of churn, offer little shelter from the storm.

Is this the end of SVOD? Are we all doomed to interruptive commercials on our beloved streaming apps? Will 2022 mark the year Streamers abandoned the notion of truly continuous content?

Can Shoppable Ads Fix Streaming?

There does seem to be an alternative SVODs have yet to consider. Shoppable advertising, also known as contextual advertising, is not a new concept. For decades, TV corporations have sought to find a way to embed relevant ads into the viewing experience without the interruption of commercials. It makes sense: TV has long influenced shopping habits and continues to drive consumers to purchase. Most recently, an episode of the popular show Euphoria led to a 900% increase in web searches for “black cut out dress” after one of the characters was shown in such an outfit. Product placement, though, is a somewhat limited and unreliable execution on the concept of Shoppable video. While at best the effect is potent, at worse it’s unnoticeable.

Still, seamless, shoppable advertising seems to promise immense value. Across industries, companies that have successfully integrated this guiding principle have reaped massive gains. Amazon is one of the most prominent examples of targeted, seamless marketing: its ability to offer relevant product ads without interrupting the shopping experience enables it to generate $27.94 billion in ad revenue alone.

If Netflix could deploy a similar solution, their SVOD ails would largely disappear. The solution, might be for streaming services to offer information on products found in video scenes whenever viewers press the pause button. Advertisers would then pay for placement of their own, similar-looking products during these pauses, creating a completely new revenue stream for Streamers. This would be an even more ideal execution for seamless advertising on TV than pure product placement, since viewer’s purchase intent is built in– it’s inherent to the very behavior of pausing.

Recently, there have been a number of shoppable advertising efforts by streamers:

Despite all of these efforts, shoppable advertising has only achieved minor and sporadic successes, as streaming giants are still unable to expand shoppable ads across their entire video libraries.

Barriers to Shoppable Video Ads

If shoppable ads have such promise, why are they not a reality yet?

The engineers at CONTXTUAL have been working on solutions in this area for several years, and point out several key technical challenges that need to be solved to make shoppable ads ubiquitous in streaming. CONTXTUAL has been developing AI to analyze and identify products in mainstream streaming video to deliver shoppable ads that match products to in-scene items when viewers click pause.

According to CEO, Gregory Casalino, there are several reasons why shoppable ads have not had widespread success. Most important, though, is that broad scalability is hindered because the life span of typical e-commerce products are short-term. Product “matches” frequently become unavailable, and this requires a way to constantly re-match new items that are still in-stock, all whilst maintaining the same standard of similarity to the in-scene item as the last match.

Why is this difficult?

One problem is that E-commerce products are constantly in flux. Products sell out, become outdated, or shift from their original link. This is frequently a problem for Pinterest users, who, upon clicking “Buy” on their pins are instead rerouted to an irrelevant page. Hard-links require constant updates to ensure that they continue to work, but manually updating ads that become outdated, unavailable, or broken is nearly impossible because of the sheer volume of changes that are needed. Still, working links are fundamental to the value proposition of seamless contextual ads: advertisers and consumers would stop using it if the links do not work.

The next problem is that dynamic ad matching, which seems like an obvious fix for broken links, relies largely on metadata – which is notoriously unreliable. It seems simple… program a system that constantly tests links and then switch in similar products when they become unavailable. The problem is beyond some basic categorization, retail metadata is largely unstructured– especially in the crucial attributes that distinguish key visual details.

For example, while “leather jackets” are a very specific type of clothing, they can show massive variety in their styles and attributes even within the same taxonomies of “Moto,” or “Bomber,” for example. It doesn’t help that these taxonomies, too, are unstandardized. Thus, when a computer system uses dynamic ad matching to compare products by metadata, many of the matches, albeit in the correct product category, look significantly different from the in-scene product due to differences in the the finer attributes like buttons and zipper placements, collar shapes, textures, etc.

Casalino refers to this range– where the ad is close enough to meet the typical viewer’s expectations of a visually similar product– as the Visual Zone of Tolerance (VZOT). Dynamic ad matching, being so reliant on unstandardized metadata, typically results in poor matches on the VZOT. But in an age of automated intelligence, neural networks, and machine learning, why can’t AI and Computer Vision match products to scenes and provide contextual ads? The challenge with AI, says Casalino, is that products in scenes are often partially obscured by other objects in a scene, and/or affected by variables in lighting and angles, or manipulated in someway that makes their identity unclear. This could be something as simple as rolling the sleeves up on a button-down shirt, says Casalino. These different product contexts ultimately create exponential training demands for AI to accurately identify specific products that fall within the VZOT.

Thus, while many have tried for years to crack the code on shoppable video, there have been some serious technical challenges to overcome.

Is there hope for profitability?

While the move to new ad-tiers may give streamers a short term fix, the benefits may still not be enough to cover the costs that could come with alienating viewers. Consumers who left cable companies to their streaming services may feel it is a bait-and-switch and turn on them.

As Netflix and other large streaming services try to find a new way to do business, they have to consider the potential drawbacks of fundamentally changing their models.

Meanwhile, companies like CONTXTUAL are breaking down the technical barriers that have kept shoppable ads out of the market. Non-interruptive, relevant, and useful to consumers, these ads seem to represent a novel revenue stream that could provide a way out for struggling SVODs, increasing revenue drastically with a much lower likelihood for subscriber backlash. CONTXTUAL’s estimates suggest they could add anywhere from $5-20 in monthly ARPU for streaming companies. For a company like Netflix, with a current subscriber base of 220 million, that would mean $1.93 billion in new revenue per month.

Neither avenue is without its challenges, but regardless of the hypothetical losses or gains it is worth noting that the problems associated with commercial ads are fundamentally existential in nature: ad-free Streamers switching over to ad-tiers must ask themselves if they think they’ll be able to successfully grow their consumer base albeit reneging on one of their primary value offerings. The unfortunate irony here is that service-providers’ solution to garner more business is to offer less service.

The problems with shoppable ads, on the other hand, seem merely technical: fix broken links, and train a hyper-accurate AI. And maybe that’s not easy, but considering recent technological feats (e.g., self-driving cars, gene editing, a rover on Mars, etc.) this hurdle feels more than surmountable. So for all of the TV-loving, commercial-loathing folks around the planet, couldn’t we rise to the occasion and, to quote our favorite martian, “science the shit outa this“. The team at CONTXTUAL certainly thinks so!

Team Contxtual

Author Team Contxtual

More posts by Team Contxtual