There’s been an interesting Youtube video doing the rounds of Tech and Business circles since Dec 12, 2024. This is the one where Satya Nadella, global CEO of Microsoft has been interviewed by two renowned Tech podcasters, Bill Gurley and Brad Gerstner, popularly called BG-2. This video took off because in this interaction, Nadella went on to announce the impending demise of SaaS applications like CRM and ERP. And Gen AI agents interacting with the database directly!!

Link to the Youtube video below :

I capture the key highlights of this video here. The podcast almost had two parts – for me, the lessons in leadership and strategy that Satya spoke about were equally exciting as his predictions on AI.

Part 1 – LESSONS IN LEADERSHIP AND STRATEGY

1. Tenure of Satya Nadella in Microsoft

• Joined Microsoft in 1992
• Took over Online business in 2007
• Launched Bing Search in 2009
• Launched Azure in 2011
• Became global CEO in 2014 (Aug 2013 Steve Balmer had decided to retire).
• In 2014, Satya wrote a 10-page essay called “Irrelevance of Microsoft” for the CEO selection process :
  • Satya had forecast an impending environment of Ambient Intelligence and Ubiquitous Computing, which was later called, Mobile-first, Cloud-first. And he posited that MSFT had to earn the right to win in this changing space.
  • The environment and business has pretty much played out as per his essay; MSFT went gangbusters into Cloud through Azure – less margin, higher TAM. Later on, they have been the first movers in AI with their OpenAI investment (reportedly 49% stake?).

2. Over this period, outputs for MSFT changed as follows :

• Azure annual run rate revenue changed from USD 1bn to 66bn
• MSFT Revenue grew to 2.5X
• Earnings to 3x
• Share price is 100x
• Added 3tn value to shareholders – by this metric, Satya could be the best professional CEO ever hired of all times in corporate history!!!

3. Microsoft – Technology changes grabbed and missed since 1992 :

Since 1992, Microsoft has been in and out of the five huge interventions in the industry :

1. Did well with Browser after initially missing it (to Netscape)…and later losing again to Chrome 😊
2. Missed Search (Google walked away with this) – “we recognised later that the organizing layer of Internet happened to be Search”
3. Smartphone – entered but iphone swamped everyone
4. We got the Cloud right…Azure, coming from behind AWS and GCP
5. Got AI right with Copilot…and more famously with their OpenAI investment.

4. Lessons in Strategy from the master strategist – largely verbatim

• Nadella views his Microsoft experience from 1992-2024 as “one continuous period for me, where I have tried to match patterns – where had we been successful and where all we were not – and then try to do more of the former than the latter”.
• Sometimes it is OK to Fast-follow; but you should not do anything out of envy. You should do it only because you have Brand permission to do it, and you think you can do it better than others – both these axes matter. (By brand permission, I am assuming that it aligns with our core strengths, our strategy moats, the larger business we are in).
• Jeffry Moore said once Why don’t you go and do things that your Customers love? Cloud was one such thing for Satya. Azure vs AWS : It was never a winner takes all for infrastructure.
• What I think I have done a reasonable job of is the areas MSFT has entered in and tried to build a good business in. For that, I have used a two-step framework for deciding what areas to work in :
• You recognise your structural position (strengths) and
• then find areas to win where you have permission from the brand, your customers, and your partners who want you to win.

The above could of course be called the basics of Strategy.

• A fit with our sense of purpose and mission, culture are of course also necessary pre-requisite conditions to do any of the above.

5. Advice for new CEOs to reboot culture and grow business / What’s the cultural change you brought about once you became CEO ?

• Changed hubris to empathy : From ancient Greece to Silicon Valley, one thing which has brought civilisations and companies down is…hubris. Had read a book by Carol Dweck – Growth Mindset (better parenting etc.). Microsoft was the most valuable tech company at one point of time and some hubris had set in. (https://www.goodreads.com/book/show/40745.Mindset)
• Encouraged his leaders to change from know-it-alls to learn-it-alls…and it’s a destination that you never reach…by definition. Because the day you say that I have a growth mindset, it means…you DON’T have a growth mindset; if you have one, you won’t talk about it, you will keep on learning…and growing.
• There’s not a single meeting that I do where I do not start with alluding to our Mission and our Culture – those are the two bookends.
• I have been super-disciplined in my Strategy framework over the past 11 years as CEO. The arc of thinking that I follow now is the same as what was written in the original 2014 memo. Three-step thinking :
  • Fit : Does this business align with our Mission ? Do we have the culture to support a big thrust into this ? Do we have the structural position of strength to conquer leadership here? Do we have the permission to win in this space ?
  • What’s our World-view about this space ?: Where is the world moving towards ? (eg ambient intelligence and ubiquitous computing). Does this new line of business stand a chance of succeeding in this new world ?
  • Build and execute like hell : Then build the specifics of the product strategy.

I pick every word in this framework, keep on repeating these three-steps in every meeting till everyone lives and breathes this, and is confident that we will success in this venture.

Part 2 – Satya Nadella on AI

1. Why did he invest in OpenAI or even bet on AI so big ?

• Bill Gates had the vision that there’s only category in Digital – Information Management, and entire MSFT was trying to schematise all information in the world. However, Satya later thought that a non-linear breakthrough will happen only if we start schematising information like the human brain does, possibly in natural language. The OpenAI and LLM thinking elegantly fit into to this theory. (He sneaks in an interesting anecdote that Sam Altman and Elon Musk had come from OpenAI to MSFT to negotiate for cheap Azure credits 😊)

2. How does he see the “AI Industry” shaping up ?

• There are three layers : the foundation LLM layer, the enabler (model) layer and then the Apps layer.
• There are going to be the Mag 7 companies +2-3 more smart new companies competing in each layer. OpenAI is clearly one of the 2-3 new companies in the foundation LLM layer.
• Apps are the ones which will have network effects

3. Search vs Answers (like ChatGPT)

Search is rapidly being replaced by Answers. Why would one go through multiple links and articles to get the answer when you can get the answer directly itself ? “Chat meeting Answers” is the way forward.

Distribution matters and Google has an enormous advantage. Because they are on all Android phones, and on a lot of iphones too. People still go to google.com to ask questions.

Currently, it’s an information-driven move from Search to Chat…once commercial intent moves (eg shopping, booking hotels, airlines etc) to chat, the dam may break towards chat.

4. The Biggest impending disruption…found at 46-51 mins of the video :

The biggest impending disruption are AI-first products :a first-principle AI app

SaaS apps or Business applications are essentially CRUD databases with a set of business logic. Now, of late, all the business logic is going into the Agents…into the AI tier, and then the business application may lose their raison d’etre. Business apps may collapse in the Agentic era.

People want more and more AI-native Business apps

This may impact CRMs and ERPs too (Finance and Operations).

The logic layer can be orchestrated by AI and AI agents.

5. Lean companies

Since 1990s onwards, 100-200bps margin has been added by Lean principles to disciplined industrial companies – increase value, reduce waste. Similarly, AI is the Lean for knowledge work!!! BPR of the 1990s is back, with people who can think end-to-end process flows. And then re-engineer those for process efficiency, using AI agents.

Nvidia says that they will increase revenue by 2x-3x by adding 25% employees.

An exciting world coming up…Marc Benioff, founder of Salesforce.com went on a podcast in Jan 2025, taking on Nadella’s take on future of agentic AI, and almost rubbished that as science-fiction…more on that soon 😊.