Lots of you are familiar with the work I do in the Palantir community. I basically dig into link data and update my sources every day to check new companies that are working within the Foundry environment and, some days, other entities that are developing their work on top of Gotham or other government-tailored Palantir solutions.
I also produced the “Dossier on Palantir” updates that, by the way, will get a new release this year, with some twists and changes. I’ll let everyone know some weeks in advance.
What those that decide to support my Substack know, through their access to DataBase, is that there are more than 1,000 Foundry entities (considering every Foundry link + every separate UK / NHS access page). In between those links, there are some names that have appeared in recent times but I’ve decided not to share with everyone, because you never know with Palantir links. There are some names that become inactive after some weeks and would become former good news, if you know what I mean.
That’s why I come up with a link activity rate, that I update every 3 months, approximately, and that I recently shared with my paying subscribers. The number remains stable, compared to what it was a quarter ago, more or less, and that allows us to come up with some deductions.
The first one is that Palantir is maintaining their client adoption rate, and this is good news. This means that even if Palantir loses more pilots compared to, say, the last 6 months, those newly added clients make up with the potentially lost revenues and the % of clients added v. those lost remains the same. This is great when the total number of entities produced grows, as has been the case in the past 6 and 9 months, when we saw extraordinary growth.
So, how do pilots work? How does a company decide to adopt Palantir and how do we see if they drop Foundry or decide to become paying clients?
First, we take notice of a newly produced entity, and I can try and decode the codename or see if the acronym or the name belong to a shared company webpage or a name that obviously makes reference to an established company. Then I share the data with paying subscribers through the DataBase and if it’s important enough I make a quick update through Substack notes. I then associate the new entity with a green color and keep it in there as it is. If after a few months it stays green, meaning the link stays active through manual checking, I then look at the industry it belongs to.
When I check this metric, I look at Foundry_Gap data inside DataBase, because inside of it I gather the time passed between link activation and a Palantir official confirmation, and every industry has their own time lapses. If the time passed is greater than the average, and Palantir has not made the news public, then I’ll decide to share it through my Twitter. This process is very accurate because I check everything in a daily basis, so I can pin down the exact days in every industry.
Of course, Foundry_Gap data grows larger every quarter and especially with each AIPCon, when Palantir goes public with a big chunk of companies that have been using Foundry and we knew of before anyone else. Subscribers know what I’m talking about.
Needless to say, if the link goes down I count the company as a lost client and take it as a lost case. Only in very rare instances has the company decided to join back, as Coca-Cola and GM did not so long ago.
So, how does the process of becoming a paying client work and how can we take this data into account in predicting future revenues?
This is the greatest thing we can come up with and that I’m actively seeking every month, when we gather more data. And when Palantir releases their quarter results, I’m closely looking at new data to feed DataBase and the Foundry_Gap section to enlarge their capabilities. There’s a lot of data and I’m digesting it as much I can.
Predicting when clients added will impact revenues is the next chapter, and I’m already working with
from to produce a Foundry_Gap 4.0 with these new metrics inside of it to let our supporters and ourselves know when these new revenues will come in.If the pilot links that we have in the DataBase drop and become inactive, I mark them as red and let them out of the equation. They remain inside the total entities list, but out of the active ones, becoming an important statistic and checking manually if they become active again, as has been the case in some instances.