As I understand it IBM now pulls in a team of domain experts to help tune their algs and even markets this as an advantage over their competitors. That makes t

As I understand it IBM now pulls in a team of domain experts to help tune their algs and even markets this as an advantage over their competitors. That makes the MD Anderson case more interesting as it mostly means that MD Anderson spent $60 million to help IBM figure out the importance of that. It begins to seem obvious how IBM stays so impervious despite being so sluggish. If your marketing and sales teams can subsidize the engineering and design side so heavily by wooing the executives that actually sign the contracts, you can spin it as just "where the technology was at that time". Selling on hype gets you in the door and establishes you as player, and eventually the tech will catch up. Being 1-2 years behind blurs when you're signing 5 year enterprise-scale deals that took a year to organize and negotiate. And if your products really are a dead end, you've got enough of a war chest to just aquihire an entire branch of your company, like IBM did buying Softlayer to just become IBM Cloud Services.