Blog > Three Investment Marketing Trends Now that AI Can Write
Financial marketing in the age of trigger-happy content
In a sense, the inflows an active manager attracts are the result of two factors: how well the manager acts on changes in the world, and how well it explains its understanding to clients.
Yes, of course things like operations matter. But this big-picture view is helpful for spotting where new value can be created for clients in the coming year.
Back in 2004, Bill Gates had this to say on the matter of spam to the crowd at the Global Economic Forum:
Two years from now, spam will be solved.
It’s not without irony, then, that Microsoft is moving to own OpenAI, which unleashed chatGPT onto the world (as well as its sibling, powerful large language models).
But generative AI is also reducing the barrier for genuine creation. Even without any spam, we’re all still about to drown in a sea of AI-assisted content. Getting clients to care about your views will become, yet again, harder.
Marketers explaining how a manager sees the world will need to do it by delivering new types of value to clients. In 2023, we see this value-creation playing out in three ways:
In our conversations with financial-services companies, we’re seeing lots of interest in large language models, but also hesitation.
For one, some companies have strict rules on posting company information onto external services. Also, firms question the copyright status of generated content: is it theirs to use? Do they need to disclose generating it with AI?
There’s also a psychological element. Will your boss consider it cheating? Can employees use AI tools at their discretion? Do they need to tell anyone? Operations do matter, and these are all process questions. Right now, nobody’s sure what’s allowed.
As a result, we’re seeing three things happening. For a start, some companies are burying their heads in the sand, pretending this has nothing to do with them.
Others, currently the minority, are grabbing the bull by its horns and working on how they can safely use generative AI in their day-to-day. (Disclaimer: we provide training to companies on working with large language models.)
Lastly, and most interestingly, we see employees using it on their own initiative. They do so mostly privately. And they use it to cut down the workload, or tackle urgent, unexpected tasks.
This year, we’ll see some companies move from the first group to the second, and a whole lot of companies – unwittingly – move to the third.
Some stories remain hidden until you collect a bunch of information and look at underlying trends.
For example, consider this asset quilt, showing returns across asset classes over the 2008-2022 period. What makes it interesting?
Sure, the colours catch your eye. But when you look closely, the interest stems not from the chart’s columns (individual years), nor its rows (who performed best), nor even its colouring (how any specific asset class fared).
The value is in the mismatch between what you expect to see happening compared to what real life delivers. Like how some asset classes (particularly REITs and EMs) repeatedly flip from being the best performers in one year to the worst the following one.
The same idea works for investment comms:
The more content to make sense of, the less any individual piece matters. When Warren Buffet says something, his is just one voice in a sea of information. This raises the value of making sense of the conversation.
As the value of any single article falls, the value of what the mass of information says increases. Sentiment analysis, text analytics, pattern spotting…are all ways of thinking about text in bulk.
Just like with the asset quilt, new patterns begin to emerge. New insights, which couldn’t be gleaned from any single piece of content.
When the FT opened 2023 by stating Asset managers brace for tough year of cost-cutting in 2023, its story featured an assortment of statements and quotes.
But a different way to tell this story is to look at the mix of headlines in fund-management publications and consider their sentiment.
Here, for example, is a sample from a single day in December from ETF Stream. Any single headline tells a story. But combined, they project a negative mood. Tracked over time, this grouped-content analytics tells a powerful story on the state of the asset management industry.
On social media, users don’t get much choice on what they see. From the user’s perspective, the outcome of exponentially more content is simply a worse experience.
Email, on the other hand, gives users control. When content is dull or spammy, clients simply block it or unsubscribe. Generative AI increases the value of email, because its signal-to-noise ratio goes up.
Introducing innovation to email newsletters, by niching down or introducing an innovative angle, will likely pay off handsomely.
The most amazing thing about large language models like chatGPT is, five years ago this technology hadn’t existed at all. Yet already, these models are performing at a level previously considered science fiction.
Ours is a rare moment in time, where we know something about the future for certain: cheap, automated content will be everywhere. Written, images, video. And the tell-tale signs of what’s fake will weaken as the tech improves. All the above trends converge into one, single conclusion:
Curation is the new black.
In 2023, value will stem not from simply generating more content, but from making it easier for clients to understand the world.