In product, time-to-value (TTV) is how long it takes for a customer to receive the benefit of a product after initially purchasing it. Products should aim for low TTVs to minimize user churn.

Search and LLM products have a special case of TTV that I call time-to-knowledge (TTK), which reflects how long it takes for you to get some piece of information that you want.

There are 2 main limits to how fast TTKs can be. The theoretical fastest TTK is using brain-machine interfaces. As soon as you want some information, merely think it and the computer delivers it to you.

The 2nd main limit, and the one we have today, is how long it takes you to type the query. For most basic facts, Google has already hit that second limit. For example, if I want to know the year in which Barry Bonds was born, I can get that information as quickly as I can type it and press ‘Search’ (the answer is 1964, if you were curious). This means that there’s little reason to switch to another search or AI product for those questions.

But for some more complex queries, Google struggles to quickly display an answer. For these, I’ve started to adopt Perplexity because of its faster TTK. For example, say I want to understand Samo Burja’s Empire Theory. Here is what I get when I query Google:

The first link is the correct one, but the preview doesn’t answer the question, requiring me to click on the link and read through the piece itself. Part 1 of the piece is multiple pages long, and it’d take me perhaps 20 minutes to get through all parts.

By contrast, here’s what I get when I give the same query to Perplexity:

Although certainly not comprehensive, this is an accurate and fair summary which doesn’t require the full read, but still directs me to further sources where relevant. And since LLMs almost always provide me with the right answer the first time, which they present as soon as I type the query, they’ve hit the same TTK limit as Google – but for a wider range of queries.

So long as Google can’t keep up, I’ll continue to be a happy Perplexity user.

In product, time-to-value (TTV) is how long it takes for a customer to receive the benefit of a product after initially purchasing it. Products should aim for low TTVs to minimize user churn.

Search and LLM products have a special case of TTV that I call time-to-knowledge (TTK), which reflects how long it takes for you to get some piece of information that you want.

There are 2 main limits to how fast TTKs can be. The theoretical fastest TTK is using brain-machine interfaces. As soon as you want some information, merely think it and the computer delivers it to you.

The 2nd main limit, and the one we have today, is how long it takes you to type the query. For most basic facts, Google has already hit that second limit. For example, if I want to know the year in which Barry Bonds was born, I can get that information as quickly as I can type it and press ‘Search’ (the answer is 1964, if you were curious). This means that there’s little reason to switch to another search or AI product for those questions.

But for some more complex queries, Google struggles to quickly display an answer. For these, I’ve started to adopt Perplexity because of its faster TTK. For example, say I want to understand Samo Burja’s Empire Theory. Here is what I get when I query Google:

The first link is the correct one, but the preview doesn’t answer the question, requiring me to click on the link and read through the piece itself. Part 1 of the piece is multiple pages long, and it’d take me perhaps 20 minutes to get through all parts.

By contrast, here’s what I get when I give the same query to Perplexity:

Although certainly not comprehensive, this is an accurate and fair summary which doesn’t require the full read, but still directs me to further sources where relevant. And since LLMs almost always provide me with the right answer the first time, which they present as soon as I type the query, they’ve hit the same TTK limit as Google – but for a wider range of queries.

So long as Google can’t keep up, I’ll continue to be a happy Perplexity user.