Analogous

Selling leads is a lucrative business. You convince buyers that you got their prospective buyers in a proprietary database “They’re high quality, they will convert” you proclaim. The times are good and business activity is high, but you’re a cloud-based platform, you have features, you’re not the Yellow Pages. So you convince your buyers to sell them per-seat licenses and your buyers get an army of sales reps to spam prospects for 8 hours a day. Response rates are not great, and conversion is low-ish. But that's the cost of cold-calling and what are the alternatives? Build better products and foster client relationships? No… it’s spam, that’s the way.When the times are not good and interest rates are higher, your customers start cutting back and you can no longer defend the hefty per-seat price tag.

From Zoominfo’s Q3 report:“Beginning in the third quarter of 2026, we introduced a hybrid pricing model consisting of an annual platform fee, pre-purchased credits, and other flexible arrangements, in addition to our traditional seat-based packages”Sure it’s a way to preserve your client base and offer them more pricing flexibility, but you give up the cashflow and potentially down the line… revenue too.Getty Images went the other way, they used to sell their images using pre-purchased credit packs. The only problem is when customers are in a rut, they’re careful with their credit spend and at renewal they sign up for lower credit packages.From Getty Images annual report in 2024:“In recent years, we have shifted revenues toward subscription products including annual subscription products to drive revenue growth and durability. As of December 31, 2024, annual subscriptions represented more than half of total revenue”That obviously didn’t work though, the stock got wiped in the years after and their subscriber revenue retention is at ~88%.

Whichever way the wind blows, subscriptions are becoming hard to defend when budgets are tighter and customers see less value in the product. And hey… there are also these mystical agents from AI labs that are dissolving the need for certain classes of software, and data players might be one of them.

I have a book in PDF that I wanted to read on my Boox, the ebook reader for non-Kindle users. It's perfectly capable of viewing PDFs, but we're in the age of intelligence, and settling for a non-native format seemed like a tradeoff I wasn't willing to make.

I thought it was the perfect straightforward task to delegate to my agent. "Please convert this PDF to EPUB" I prompted confidently. Little did I know it was far from it. The pages were all scanned, which is more likely for older books with no native digital versions. The agent pressed on, ignored all the diagrams and outputted a digital book with garbled text. With further direction from me, it was able to identify and extract all diagrams and place them in their correct position. But the entire process dragged on for over two hours, yielding uneven fonts, awkward spacing, missing captions, and miserable formatting throughout.

Yes, I went in not knowing there's no linear path to converting scanned books to digital versions, and maybe that's on me. But thanks to reinforcement learning, these models are trained to keep digging in their heels and go on an endless quest for a solution. I would've just settled for "Listen, that's not possible, and my best efforts would render something resembling a digital book, but not quite readable by anyone. Would you like to proceed anyway?" Instead, it was a performative circus, lighting tokens on fire.