Google flexes its Arm; Metronome runs the meter
Today: Google's first custom Arm server processor is now available, Metronome's new tool could help SaaS companies switch to usage-based pricing, and the quote of the week.
Smita Hashim's new job as Zoom's chief product officer is to help it become a broader player in enterprise software. In her view, this new era of hybrid work is opening up new product possibilities built for teams with a mixture of employees in the office and at remote locations.
After more than 20 years shaping video communication products for Microsoft and Google, the two juggernauts of office software, Smita Hashim is now trying to catapult Zoom into their category.
Zoom became a household name in March 2020 after the world shut down to try to prevent the spread of Covid-19. Sales surged — as did its stock price — as Zoom's video meeting product became indispensable to businesses, schools, and even friend groups trying to stay connected during those harrowing months.
Fast-forward more than three years, however, and Zoom's stock price and revenue growth have returned to their pre-pandemic levels. What looked like a clean shift into an era of remote work has become much cloudier as companies enforce return-to-office mandates and re-emphasize in-person collaboration, and lots of businesses are using Google Meet and Microsoft Teams as good-enough video communication tools that work seamlessly alongside their widely used email, calendar, and collaboration software.
As of February, Hashim's new job as chief product officer is to help it change that trajectory by becoming a broader player in enterprise software. In her view, this new era of hybrid work is opening up new product possibilities built for teams with a mixture of employees in the office and at remote locations.
"I'm a huge believer in hybrid work," Hashim said in a recent interview. "I think hybrid work is an opportunity; it's the good thing that came out of the pandemic, that people can work in flexible ways." In that interview, Hashim also touched on Zoom's AI strategy and its own hybrid work plan.
This interview has been edited and condensed for clarity.
Zoom obviously had one of the most incredible, surreal experiences of any company during the pandemic. How did that inform the product strategy over the last couple of years before you got there? And now that you're here, what are some of the things that you and your team have learned from that experience that you're bringing forward with product strategy?
Things were pretty crazy in video communication and Zoom was really at the forefront of it. So I think the way the team responded was just a massive, massive product build-out. I think last year, for example, Zoom launched 1,500 features.
We are not just a single-product company, we are a portfolio company. So as we think of multiple products, and bringing them together for users, we're going to make sure that we really take this portfolio approach; that the products connect to each other nicely in ways that feel fun again and comfortable.
Your background is interesting that way, given your time at Google and Microsoft. Zoom has talked a lot about building an office-suite type of product over the last couple of years, but I personally haven't seen many examples of people who are using Zoom as a central hub, the way they use Workspace or the way they use Teams and Office. Can you give me a sense of how that is playing out right now?
From our product portfolio point of view, where we see Meetings as the product from Zoom which is used by pretty much all our customers, then the (other) product that I see being used a lot is Zoom Phone. We called it out in our earnings call; it's now 10% of our revenue and is getting really good adoption. Our Rooms product, which we call spaces, is a product that continues to get a lot of adoption, (and) we look at products like Whiteboard, which is also becoming very good for us.
We are completely happy with a coexistence strategy. A lot of our customers, they love the Zoom meetings. They want to continue to use it, they want to use products like our Meeting Summary, which we just announced, they like being able to chat outside the meeting with continuous meeting chat. But we make sure that our products are also accessible (to other enterprise software products), that our data is accessible.
Most people I talk to feel Zoom as a video meeting product is a better experience than you can get from Workspace or from Slack or from Teams. But a lot of those other products have your calendar and email, and tools like that are extremely sticky. Does Zoom want to move into that space a little bit more?
I think a lot of our customers will continue to use Microsoft and Google Calendar and email. But if you think of email, you can think of the client and the service. And so the service can continue to come from Microsoft or from Google, but from a client perspective, you can deflect the email and calendar into your Zoom client. And then we can provide capabilities, like we can show you who has joined your Zoom meeting, for example.
Providing a great experience for people who are really Zoom Meetings users, without that toggle tax, is where we are going with the product and that's where we are getting traction.
You started off talking about how one of the things that brought you to Zoom was to be able to play this role in making hybrid work and remote work better. The sense I'm getting is that a lot of the remote work things that people were talking about two years ago are really changing. People are coming back to the office, companies are forcing people to come back to the office. How does Zoom evolve along with that shift? It feels like two years ago we were all talking about how remote work was the future and it's clearly not going to work out that way.
Even Zoom is going back; I'm part of product and engineering and people who are local, once a week we are going back into the office. I met a lot of customers yesterday as part of an advisory we do with them, and many of them are coming back into the office. But what we're hearing often is it's two days a week, three days a week. And it's still settling in; customers are beginning to say "I don't want remote people," but they are not sure if they can actually enforce it because the talent is so spread out. We let anyone go remote if they want to, but if you are local, we are asking you to come in one day a week.
I think actually the world is even more challenging now with hybrid work, because hybrid means that you have to somehow build products which accommodate this hybrid, distributed time zone, asynchronous world. An example would be if some people are in a meeting room and some are on video — which is almost always the case now with remote distributed teams — how do you bring the people in the room to the forefront so you don't get this bowling alley effect?
So there we have a product called Intelligent Director, which is about to launch, which takes three cameras and really does a good job disambiguating people. If people work across time zones, things like our Meeting Summary helps people who were not there catch up very, very easily.
One question I've been asking everyone that I talk to about generative AI is if everyone is embracing the same basic foundation for generative AI — really just OpenAI's tools — how do you as a product person differentiate from everyone else who's using the same underlying technology?
So from an AI point of view, one thing which we talk about is our federated approach. What federated means over here is that on the back end, we are deciding to use different models. I know Microsoft is doing all this on OpenAI, and that's the one they are going with. But for us, that's not the case.
We are going to have OpenAI; we are using OpenAI for one of the features which we launched, which is the Zoom IQ Chat Compose. For Meetings we are using our own language model, which is Zoom's own proprietary language model. And then we also announced a partnership with Anthropic; we are going to use their Constitutional AI to do more feature work for the Contact Center product, where we think that rule-based approach is going to be interesting because you have to be very specific in the answers you give.
We're designing our systems to be able to plug and play into these various types of models to be able to pick the best model. My personal belief is that smaller models, which are more specialized, will actually do very well.
There may not be a good answer to this question, but how do you know how small or large a model needs to be?
I think there's a lot of learning going on right now. That's another reason why the federated approach makes a lot of sense, because the technology is evolving so fast, you want that flexibility in your products to be able to do work with a variety of models.