gistfile1.md

ROGER EHRENBERG

• Do you want to go the 100m acquisition route, or do you want to go big? Dan: We want to be the second wave of ETF innovation
• In that case, always stay independent, and do not tie your brand to any other brand. Do not issue ETFs in close collaboration with a Blackrock or Prudential -- once you do that you're tied into their network and it will be very hard to get acquired. Doing such a deal would bring in faster revenue, but would also place you at the mercy of their restrictions and constraints. Doing it yourself will be more work, but will allow you to keep much much more upside.
• Similar with Markit -- don't mix in your brand with them too much. Try to get a revshare from them, but don't give them a revshare for us. Try to keep our platform separate.
• On focus: Roger literally says this: "Lack of focus will kill this company. People will have opinions about this, but I am right." He says to stay laser focused on the technology and the current product, and don't worry about the synthetic illiquids just yet. Doing that stuff is going to involve a lot of "plumbing" and operational work around regulatory work and legal work, and it will take a very serious team with a lot of management and we don't' have expertise there, but it will distract us from other things. We will have to do it in house. That team is going to be bigger than our current tech team, so we shouldn't get too distracted with that right now.
• These finance guys and banks don't have respect for things they don't pay for. Think "that's the guy we get free product from" vs "that's the guy we pay millions of dollars a year for." Don't become the former. As soon as you can, do something like talk to Goldman and get them to pay 100k for the next three months. Tell them that we are just looking for them to commit to something, it's not a long-term deal, and that we'll renegotiate at the end. Making a 100k deal is going to have to get sign off from a MD or even get run by the CFO of Goldman -- a VP isn't going to be able to cut that deal. This is important to validate for ourselves that they are actually really want our product and think it's valuable.
• Nononononoono on David Eisen. Big head shake. Big meh. Just money -- not worth it.
• Palantir sells $5M/yr to Bridgewater and others
• One of his portfolio companies that does in-memory analytics just closed a deal that will bring them in $2M/yr (highly confidential)
• Pricing: We can think about a grassroots sales strategy similar to asana: Price $25-50k per seat, and once there are like 20 of those in a company, the CFO will have to pay attention and make a high level institutional deal, probably in the single digit millions.

Blackstone: Bill Murphy

• "You would be surprised by how unsophisticated sophisticated investors are" Most people won't be able to appreciate the sophistication of our tools.
• Capital IQ was very dependent on their data sources. Competitors would sometimes acquire a data source, and then shut down the contract or increase prices exorbitantly.
• Look into Novus: $30M invested from Bain (capital?) I can't remember -- maybe Bill said that they have data?
• Bill recommends to do a clear market sizing -- there may not be very many people who will want to use our tools because they are too advanced.
• MSIR just wants to write good research reports -- that's their main incentive
• Bill I think expressed uncertainty that all hedge funds may want our tools, because many of them may be too unsophisticated
• Don't be just a feature of Bloomberg -- or else they could undercut you by simply adding in that feature.
• The data is crucial -- and it must be correct. Fundamental data, EBITDA, etc…
• Provide full audit-ability all the way down to the numbers, so that the customer can trust the computed numbers. If there's a computed value like an average or something, allow the ability to drill down to the raw data, such that the customer can even average it themselves if they want.
• One possibly play: license the computation to firms like Capital IQ and Bloomberg. To get this play, first sign on a small number of clients like Morgan Stanley -- they have them make the introduction. It'll be much more compelling for Capital IQ if they are approached by one of their major customers to ask for additional computation a la Kensho, as opposed to us directly approaching them.
• He suggested we might be worth $1k/mo/seat
• He also talked about the ROI framework. Cost reductions: How much in salary can you save from cutting research analysts? Increased returns: How much additional return do you think you can get if you can ask more questions more quickly?
• Must walk through ROI before introducing pricing to clients. Otherwise they might just say, you have 1/1000th the pages that the Bloomberg terminal has -- therefore we will pay you 1/1000th the amount.

Kensho team discussion afterward: Bill focused a lot on data, because CapitalIQ succeeded through that play. But Kensho is going to focus on the computation side of things and avoid getting tangled up in the data game by having strategic partnerships, like through Markit!