Join over 11,000 Growth Memo readers. Get a free weekly email in your inbox or subscribe to the most thoroughly researched deep dives.
Artificial intelligence
How can you possibly start predictions without AI these days? 2023 was an overwhelming year for people and companies alike. The sequel, 2024, is a story of acceleration. Technology will jump another huge leap forward and leave us wondering where the future goes and how it will change our lives.
One point all tech veterans agree on is that you can never predict exactly how things play out. When the first browser, Mosaic, came out in 1992, nobody could have foreseen that it would provide a platform for companies like Google and that Google would provide a platform for hundreds of thousands of companies. AI is no different. We’re scratching the surface.
1/ Prediction: expect a wave of infrastructure improvements based on AI
I spent a significant amount of time thinking about generative AI and its impact on everything we do. My conclusion so far is that AI makes work more effective and efficient, especially creative and clear tasks.
Casey Winters brought it to a point when he wrote that generative AI platforms like Chat GPT, Claude or Bard are not new distribution platforms but shifts in infrastructure.1
Many marketers are desperately waiting for new distribution platforms to capitalize on low competition and high attention, but AI is not that. Instead, it makes many things we do faster and (increasingly) better.
In 2024, a lot of companies will adopt the following infrastructure improvements:
Companies will significantly improve onsite search (like Zillow or Spotify2).
Zillow, for example, enhanced its search function with AI to understand prompts instead of relying on exact keyword matches. The feature is live in its native app but might soon come to the web version as well.
Companies will use AI to personalize user experience, for example, through remixing.
Redfin is fighting back with Redfin Redesign, a feature that allows users to redesign walls, floors, countertops and more with AI.17
Redesign doesn’t just enable Redfin to improve the user experience and increases the chances of someone buying a home when they can better imagine what it looks like with changes. The feature also allows Redfin to collect valuable data about user preferences.
Software will take commands in prompt form. Hypothetical example: If I want to download a set of keyword positions, I’ll be able to prompt the tool and get a csv for download instead of having to navigate to a specific section and filter a table.
Conversational AI will help me with inspiration and brainstorming.
AI will answer user questions on websites and in apps (example: Arc browser3).
The Arc browser answers questions with Chat GPT
AI will summarize content from data and provide context for users.
Course Hero leverages AI to create document summaries - a prime use case for large language models.
Ebay uses AI to create text descriptions from images (source)
2/ Prediction: we’ll see at least 5 big publishers make deals with AI companies
Publishers and the media industry are not in a good place. AI has a chance to improve that, but not the way you think.
Publishers have what LLM makers need: well-written text. OpenAI & Co can crawl the web, but not all content is equal.
At the same time, publishers need better income streams. The advertising model paired with content is mot enough to run a news media organization. Plus, attention is moving to YouTube and TikTok.
Media companies and LLM makes have three options:
Users can pay extra to include content from publishers if they have a subscription.
Publishers sue AI companies until their content is no longer included in the output of generative AI tools.
LLM companies strike a licensing deal with publishers.
#3 is the best option because it gives both sides the best outcome in a feasible manner.
Apple has already started to negotiate with publishers like Conde Nast:
The technology giant has floated multiyear deals worth at least $50 million to license the archives of news articles, said the people with knowledge of talks, who spoke on the condition of anonymity to discuss sensitive negotiations. The news organizations contacted by Apple include Condé Nast, publisher of Vogue and The New Yorker; NBC News; and IAC, which owns People, The Daily Beast and Better Homes and Gardens.4
OpenAI has formed an alliance with Axel Springer.5
3/ Prediction: Google or Microsoft will build powerful AI agents based on your emails
One of the most powerful formats AI can take is that of an agent who knows you and can serve you in different ways.
“Nobody has figured out yet what the data structure for an agent will look like.” (Bill Gates)6
Google has tried to integrate Bard with Gmail and Google Docs and failed. But the idea has legs. Emails contain powerful context about our communication style, work, purchases and interests. I had my Gmail account for over 15 years, so there is a lot of context for an agent.
I see Apple in a good position to develop a very strong AI agent and join the race next to Alphabet, Meta and Microsoft. Apple already has incredibly strong capabilities: text-to-speech in iMessage and image recognition in photos are superb.
Microsoft or Apple could take market share from Google if Outlook or Apple Mail came with a powerful AI agent. All three companies have a big enough ecosystem of products to feed enough context to an LLM that accompanies you everywhere and anytime.
Organic Growth
1/ Prediction: Google continues to prioritize human content and launches an update to the Hidden Gems algorithm that improves the quality of forum content.
Unless LLMs stop evolving or Google figures out a way to reliably detect AI content and filter it out, companies will use more AI to create content for search. We’ve already seen struggling publishers sacrifice editorial standards to publish faster with AI.
Two, SI using AI to create buying guides is a perfect example of a shift from differentiated content (journalism) to arbitrage (affiliate marketing) and commodity content.
TV made news a commodity, except for the new breaker. AI does the same to generic content. A Tweet about traffic stealing recently attracted 6.7m impressions on X alone. The idea is to recreate content from a competitors site with AI and use it to beat them in organic search.
The best remedy I see today is prioritizing content that’s most likely created by humans. At the same time, Google will replace non-informational queries with products: shopping searches turn into a marketplace and local searches into Google Maps.
2/ Prediction: Google doesn’t launch SGE unless someone else launches a powerful AI search engine
Everyone wonders whether SGE will launch or not. At this point, I expect Google to treat SGE like a reservist:
keep SGE in beta until OpenAI, Bing or a new player puts something on the map that could endanger Google Search as it stands today
continue to develop and refine SGE behind the scenes to have it “ready” when they need it
SGE today doesn’t provide much value:
local searches show a map with local results that Google was showing already
shopping searches show a list of products, similar to organic and paid Google Shopping carousels
Informative searches show basically an expanded featured snippet
The upgrade to Gemini has improved the output somewhat, but the experience is still the same. It’s a design problem.
Launching SGE right now could cut into Google’s profit margin. Why take the risk when you don’t have to? Besides, Google can watch Microsoft and OpenAI get sued by media companies and creators from the sidelines.
However, this strategy also means Alphabet is playing defense and could be caught off guard again.
3/Prediction: SEO tools will add more AI features that deliver insights, simplify keyword research and help with content creation
It’s a no-brainer for SEO tools to add AI features that help with keyword research, insights, and reporting. As a cocktail of art and science, SEO is very sensitive to AI. I expect a steeper improvement in AI writers, but the next step in AI SEO isn’t to create content (art). It’s to get more out of the data and technical work (science).
4/ Prediction: higher focus on Organic Growth
In the decade before the pandemic, startups were funded with so much money that they could buy themselves into markets with crazy ad spend. A few companies are still able to do that, like TikTok or Shein and Temu. But moving forward, funding rounds outside of AI will be smaller and fewer.
The result is a higher focus on organic channels like SEO and social. Many companies haven’t developed the mind-muscle connection to run effective SEO strategies. Programmatic AI content looks very attractive, and I expect a lot of companies to make use of it. Some of them will succeed, and some will embarrass themselves in the process.
As a community, we’ll develop a better understanding of when to use human vs. machine input for content.
5/ Prediction: More creators will make a (good) living from creating content
For 2023, I predicted “Fewer crypto millionaires, more creator millionaires” but underestimated just how many people made a lot of money from the digital currency. However, I still see a lot of signs indicating that good content creators will make a lot more than minimum wage on the internet.
A lot of attention has been moving from companies to creators, and I expect that trend to accelerate in 2024. I’m not just talking about the new SaaS company that gets zero engagement on its LinkedIn company profile. I’m talking about journalists leaving publishers to start their own Substacks, inhouse operators making a living from cohort-based courses and teachers starting Youtube channels. People follow people.
Founder-led marketing, which is company founders posting on social media to get attention and business, is part of the same trend. Founders have special status because they have the courage to start a company, a 360-degree view of the company and deep insights about the industry they operate in. We rightfully glorify founders who take a huge risk by building a company, often around products will little chance of success.
Winners and losers
1/ Loser: Meta, Winner: Microsoft
Microsoft has a market capitalization of $2.7 trillion. At the end of 2024, I expect them to add another trillion and reach $4 due to a better economy and strong AI tailwinds.
Meta, on the other hand, will have a tough year with the exception of Threads. In October 2023, 33 attorneys general filed a lawsuit against Meta for harming millions of young users under 13 years. We’re finding increasing evidence that social media in its current form does more harm than good, and content platforms need to invest more money into moderation. After an impressive year of efficiency, Meta needs to make bigger changes to Facebook and Instagram.
At the same time, WhatsApp will get a lot more attention and play a bigger role in the lives of marketers due to its newsletter-like capabilities. People share more on messengers than on social feeds, and Meta knows it.
2/ Loser: X, Winner: Threads
Twitter’s downfall continues in 2024. Fidelity recently marked the company down by 71.5%. Musk’s vision for the Everything app won’t become reality.7
After Elon Musk tore advertiser trust apart like a paper plane, the only way forward is a subscription for everyone, which could weed out bots and toxic people.
Threads will surpass X in MAU (monthly active users) after adding Europe and potentially other geographies, plus an open API.
3/ Loser: Quora, Winner: Reddit
Google’s Hidden Gem update seems to have tipped Reddit over into profitability. Either way, after overcoming protests in 2023 and developing a new API pricing model, the company sets sail toward IPO.8
Quora also gained a lot of traffic from Google’s update but provides such a horrible user experience that Google will reverse the bump.
My prediction is that Quora will sell to an LLM maker in 2024, either OpenAI (Adam D’Angelo already sits on the board) or Anthropic.
4/ Losers: traditional media, Winner: creators
In 2024, we’ll see 1-3 big publishers sell to private equity companies for cents on the dollar and make deals with LLM makers.
The 2024 presidential election marks the final break of trust between broader society and traditional media. We’ll wonder whether traditional media companies really have higher editorial standards than TikTok & Co.
Meta is already trying to get ahead of fake AI content around the 2024 election by restricting advertisers and enforcing transparency about AI.9 But TikTok, the source of information and news for most younger generations, doesn’t. Expect to see more polarizing content on TikTok, like in the Israel-Gaza war. If you thought Facebook in 2016 was bad, you haven’t seen TikTok in 2024.
On the other side of the coin, creators see an influx of cash into the economy. 44% of advertisers plan to increase creator investment by an average of 25% in 2024.10
Question marks for 2024
Some of the most exciting questions for this year:
Will the open web celebrate a return? Will people care about and adopt ActivityPub?11
How will Google fight AI content? What does the ad model for LLMs look like?
Will OpenAI release ChatGPT 5? Will LLM makers be able to defend themselves against copyright lawsuits or settle?
Will the economy recover to pre-COVID conditions? Will startup funding follow suit?
How will content platforms and marketplaces change as a result of AI?
What do you think I missed? What’s important in 2024? Leave your answer in the comments!
THANK YOU. Those cornballs at Gartner saying SGE would reduce clicks 50% isn’t based on reality (at all). Love your take on Google only moving if forced.
THANK YOU. Those cornballs at Gartner saying SGE would reduce clicks 50% isn’t based on reality (at all). Love your take on Google only moving if forced.