Will Perplexity AI's Integration of Ads Kill Demand or Allow it to Grow?

Revenue generation comes into focus in Generative AI at the intersection of BigTech investments.

Hey Everyone,

Where does the rubber meet the road for Generative AI? It appears to be how these hyped startups actually generate revenue, how they make money. And well, this is going going to be a problem in contrast to the Venture Capital boom around the space. Perplexity AI going into Ads is a major case in point.

As Generative AI startups struggle big-time (outside of OpenAI) to generate real revenue that would allow them to keep going, Perplexity is going to be one to watch.

Casey Newton’s Platformer, that was originally about Platforms (presumably dying social media ones!) has mysteriously turned into an AI Newsletter and recently covered this topic. In the headline he makes an important point, for chatbots like ChatGPT or Perplexity, Ads could easily kill momentum and lower user trust.

It’s one thing for Google to have Ads, it’s a search engine. But for a chatbot like how Perplexity AI works, Ads would feel a lot more intrusive. It’s not as if Perplexity has that many free or paid users.

Ads are likely not the Golden Ticket Any Longer for Consumer AI Products

According to the latest data, Perplexity AI has approximately 10 million active monthly users. So yikes, that’s not very money for the hype they get on X or LinkedIn. A self-proclaimed Unicorn? How? How Big Tech and their funded startups make revenue is going to be a problem, and it’s been apparent with Stability AI and it’s apparent with the likes of Cohere and even Anthropic.

Startups like ElevenLabs or Midjourney who serve a specific function and are leaders in their field are a different story. Perplexity will likely get acquired just as Neeve was, there’s no real way around this conclusion. But this is no longer a matter of mere speculation:

Generative AI search engine Perplexity, which claims to be a Google competitor and recently snagged a $73.6 million Series B funding from investors like Jeff Bezos, is going to start selling ads, the company told ADWEEK.

If you know Perplexity, it’s a search engine but more using LLMs, actual citations and a chat like interface. A bit similar to You.com. But nobody really wants to pay for search and with the (inevitable) introduction of Ads it’s a big test of their product-market fit which might break their entire business model.

Perplexity, founded in August 2022, has a much smaller user base (a self-reported 15 million monthly active users) than the likes of Google, Microsoft Bing and AI rival ChatGPT. However, Google, OpenAI and other AI-powered search engines have yet to figure out how to reliably monetize AI search.

At least with Google I can trust that they will reincarnate fresh takes on how Ads might appear in search experiences, not longer just a static interface but a more interactive approach. Just like firms like Roblox is reinventing Ads in communities and in gaming.

How will Ads be Implemented?

It appears according to Adweek and Search Engine Land that since Perplexity gets a lot of related questions, which include links to sources, that account for 40% of Perplexity’s queries. So Perplexity plans to add relevant, related brand-sponsored questions along with organic questions.

I’m not sure from a UX standpoint how that does not kill the experience though? I personally find Perplexity’s interface very ugly and off-putting.

When a user delves deeper into a topic, the AI search engine might offer organic and brand-sponsored questions. That might feel a bit like an LLM hallucination, but it’s not - it’s a Native Sponsored Ad! Chatbots are already not so popular with consumers and Consumer AI is a bit of a game of musical chairs as a16z’s deep dives have shown us.

Consumer AI is Unprofitable for a Reason!

a16z’s deep dives shows a quickly changing Consumer AI ecosystem to the point of very few winners.

See September 2023:

To March, 2024:

Notice anything in particular? The rise of Claude and a lot of change on the leaderboards.

• Perplexity moved up from 10th to 7th

• Claude moved up from nowhere to 10th (that’s actually deceptive people have been using it via Poe)

• ElevenLabs moved up from 24th to 11th, so a major winner.

• But outside of QuillBot, Consumer AI is a constantly shifting affair.

You basically need to be a category winner to take any major spoils that might translate into revenue, like ChatGPT has or like Character.AI does. After Claude 3 Opus I expect Claude to move even higher on this list with complements on X. However Perplexity having 15 million users doesn’t really move the needle in the future of Search.

“Perplexity was founded on the belief that searching for information should be a straightforward, efficient experience, free from the influence of advertising-driven models.”

However it appears that Perplexity always saw Advertising as the way out of trying to hook consumer and build a real business model.

In an interview with Wired, Perplexity founder and CEO Aravind Srinivas discussed ads, said he wasn’t against advertising. In fact, his vision for advertising on the platform is to help advertisers understand who is searching and then bid on the most “high-value traffic.” I guess that is also another interesting thing about giving multiple citations for a query, there might be some Advertising advantages involved in offering that level of transparency to consumers.

Sam Altman says he doesn’t love Ads. He has even called internet advertising a “momentary industry.” But he does seem to like Nuclear Fusion for energy. So Perplexity is at a very interesting window in its AI startup journey where it has to become a bit more like BigTech to survive as an AI startup.

Perplexity’s roster of investors is also highly fascinating including:

• Databricks

• Nvidia

• Here’s a recent list: NEA, Databricks Ventures, former Twitter VP Elad Gil, Shopify CEO Tobi Lutke, ex-GitHub CEO Nat Friedman and Vercel founder Guillermo Rauch. Other participants in the round included Nvidia and — notably — Jeff Bezos. 

Since Snowflake acquired Neeva, it would be pretty hilarious if Databricks acquired Perplexity, but also Amazon or even Apple might make sense. Consumer AI needs to exit or die, even Midjourney has slipped in the a16z rankings that we see above. It’s tough out there.

The Chief Business Officer of Perplexity, Dmitry Shevelenko, spoke with Adweek to reveal the company's plans to sell ads. It’s not clear if Perplexity can find a great monetizing angle with Ads or if they can not be disruptive for their power users. In early 2024, they raised over $70 million in their Series B.

What’s impressive is their average site duration of each visit. But they would need to vastly increase their revenue just to survive the next few years without getting acquired. Ads might not be the golden ticket and I’m not sure what else they can do to monetize.

Perplexity’s Subscriptions are Expensive

Perplexity Pro costs a stunning $20 per month or $200 per year. I’m not sure how they reached that price point for their premium product?

I’ve heard people who really like it, but they tend to be technical people who work in technology for whom $200 a year is nothing.

Many Trending Consumer AI Products are Temporary Fads

When we look at which new tools to the leaderboard are sustainable, uncensored ‘spice’ chatbots are trending. Is using Perplexity a trend that will stick among its active users?

Perplexity AI has some Momentum

• Perplexity AI has around 10 million active monthly users.

• In 2023, there were over 500 million requests on Perplexity AI

• More than 50 million Perplexity AI visits were recorded in February 2024

• Approximately 1 in 4 Perplexity users are based in Indonesia

• Perplexity AI is currently valued at over $500 million

The two faced narrative of the CEO and branding though might be a turn off for some customers. While Perplexity touts on its site that search should be “free from the influence of advertising-driven models,” advertising was always in the cards for the company. What’s a search company in the skin of a Generative AI company to do?

Perplexity is s powered by both OpenAI’s GPT model and its proprietary AI model. It’s a good test pilot for OpenAI who is rumored to be building a similar product with the help of Microsoft Bing. How Advertising is implemented will be the major product-market fit issues of the company, and will likely decide how soon it exits and in what manner.

Is Perplexity Engaging? Ads might be a Kill-Joy

I’m not sure how fun Perplexity can be. It’s not Suno.AI or Sora by OpenAI. It’s search with a lot more citations and somewhat hit-or-miss answers that don’t usually do a stellar job of summarizing things if you are looking for niche quality answers.

Unless you use Search nearly constantly for your job, paying $200 for Search seems like a rather far-fetched idea like its former competitor Neeva quickly found out. In the end Snowflake acquired Neeva to accelerate search in the Data Cloud through generative AI. At least that was the public story, in other words it was an acquihire.

Ads will definitely make Perplexity less engaging and fun and stress-free.

Perplexity was founded by Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski — engineers with backgrounds in AI, distributed systems, search engines and databases. Srinivas, Perplexity’s CEO, previously worked at OpenAI, where he researched language and GenAI models.

As per the report, the plan is to integrate ads in a way that is not obstructive or intrusive to the user. For that, Perplexity will show ads in its related questions section that appear when a user makes a primary query on the platform.

But Ads as Sponsored Questions? That’s a bit far fetched. Brands are also wary of brand safety and transparency. They don’t want the related questions they sponsor to show up against queries they don’t want to be associated with. That won’t be easy to sort out.

I don’t find questions and answer typically very engaging, I mean I did spend some time on Quora years ago. But those were fascinating because they were human experts and included human insights.

Per the a16z deep dive, Search accounts for just 4% of Generative AI consumer products.

At least Perplexity stands on its own, it’s not such a crowded field as you might think. Neither is You.com a very strong competitor. I might personally use both services maybe a few times a month each as it stands today. Both with extremely mixed results in early 2024, though I am not a paid user.

I’ll be watching how Ads feel in Sponsored Questions on Perplexity.

Perplexity’s unique value proposition might just be its lack of serious competitors outside of Google and Bing. Even a tiny sliver of the lucrative Search Ad market would be enough to make Perplexity profitable enough to be able to continue to refine its LLM search experience. Ad spending in the Search Advertising market market worldwide is forecasted to reach US$304.90bn in 2024.