In the third and final part of our interview with Tomasz Tunguz, we get his take on…
- The value of basic estimation. If you know lots of people will use your feature, that’s a good place to start.
- Why transparent culture in SaaS is necessary, and how to create transparency when the data is delicate (e.g. in the case of user metrics).
- How SaaS is shaking up the Fortune 100, and why it’s only just beginning…
Also, Tom shares a brief overview of his upcoming book Winning with Data (co-authored with Frank Bien, CEO of Looker), available on June 17th. It will be a deep dive on how companies create sustainable competitive advantages with, you guessed it, data.
Matt Pasienski: Is it going to be 100% of the users? Is it going to actually go out and triple our total [impact 00:00:04]? These are questions that can be answered. What is your normal method of suggesting to get to that point?
Tomasz Tunguz: You just need a basic estimate. You can look at similar features. That's a really easy way of doing it. You often know ... Let's take an example. Let's say you were working on a word processing document, and you were going to build a new feature that allowed you to .... Let's say you really wanted to have alternating rows of a table different colors. That is not 100% adoption feature, right? You might have like 5%. Not everyone is going to need it. You can look at the usage data to try to figure out what percentage of documents actually have tables and back into it that way. The other thing, if you are going to build a print function, you kind of know lots of people are going to use this. What you need to do is ... It's just basic estimation. I don't think you need to do anything more than that.
Matt Pasienski: Awesome. You told me you were creating a book.
Tomasz Tunguz: Yeah.
Matt Pasienski: Can you tell me about what the book is and the data-driven approach to forming businesses.
Tomasz Tunguz: Yeah. We wrote a book. It's called Winning with Data. It's available on June 17th. It's available on June 17th, and it is basically a collection of interviews and case studies about how companies create sustainable competitive advantages using data. A big part of it is here is the history of data. Here's why we are where we are. The second part is really about this is what success looks like. Let's take a look at what it really means. We interview companies like Zendesk and ThreadUp and Toyo. This is what success really looks like. This is what it means for their business. The third part of the book is really all about how do you get there? How do you replicate something similar in your business both by cultural change but also technical change.
Matt Pasienski: I'm obviously a big fan of this kind of interview-based approach to learning about running businesses, building products, using data. There might be master equations for each business, but there is not necessarily going to be one right way of doing this -
Tomasz Tunguz: No.
Matt Pasienski: And going in and looking at these are the specific, often times very detailed, small problems that people face and are overcoming but have massive impacts to their business. What were some of the things that you found interviewing that surprised you and that you put in the book?
Tomasz Tunguz: I think one of the biggest challenges was really getting access to the right data at the right time. We interviewed Zendesk and there was this case where the product management team had really wanted to launch a new feature. They were on the cusp of launching it. The product marketing team was being asked to allocate a substantial amount of resources with financial and also time to promoting this feature. They scheduled a meeting to have this debate because the product marketing team wasn't necessarily in favor. There was this one younger product marketing manager who was toiling away in the back part of the room for about ten minutes. In the middle of the meeting, she interrupted, and she said, "Excuse me. The total number of users this feature is actually going to impact on the order of like 100. By the way, we have 10,000 users." The meeting just stopped. People realized it didn't matter.
Matt Pasienski: It's so funny that you said the part of marketing manager because I found that is one of those roles in an organization that I feel doesn't get a lot of love, but it's the transmission mechanism between the sales team and the product team. If that's not working. It's a younger person, and they aren't the highest paid person in the room.
Tomasz Tunguz: Exactly.
Matt Pasienski: They need to have the data and the ability to come in. The fact that she said, "There is 100 people going to use this." She now has a stake in that conversation.
Tomasz Tunguz: That's exactly right. Just having access to that data and people being aware that that data exists ... There is obviously a technical component, which is how do we access provision and permission people to actually accessing the data. The second is an education. People have to be aware that they can access the data and process it. The third is just training. If you were to ask this question, "How are you going to do this?"
Matt Pasienski: The fact that one person knew, but it wasn't pervasive, what other things are they missing? What other types of data are they missing? SaaS is very much about transparency.
Tomasz Tunguz: Absolutely, yeah.
Matt Pasienski: A lot of companies haven't gone to that level of creating a transparent culture, especially around stuff like how many people are using our product. It's incredible.
Tomasz Tunguz: Yeah. I'm a big fan of transparency. I think that [sunlight 00:04:44] is the best disinfectant, right? We all need to know the problems facing the business if we are all to put our best efforts into solving it. Obviously, there is going to be permissioning. You can only access certain subsets of data. My co-author in the book, his name is Frank Bien. He's the CEO of Looker. They really champion this idea of a data driven company, which means everybody has access to the data that matters to them when they need it.
Matt Pasienski: How does Looker do it? There is that tension, which is ... Look, we want to be open, but a lot of stuff is inappropriate, or we are not legally allowed to disclose. How does Looker take that approach to creating transparency without necessarily causing more problems than it solves?
Tomasz Tunguz: What we do is we really work with the data teams and the executive teams to really try to figure out. Is there social security numbers? Is there personally identifiable information? Do you need to hash that data in some way? There is permissioning that goes along with the structure. It tends to be pretty basic. It tends to be pretty easy to solve that problem. The bigger challenge in most of these businesses is that there is a data team that gets lots of requests for data and they're totally overwhelmed and they are overworked. The problem that Looker tries to solve and does a really remarkable job for companies is enabling that data team to scale. What they do is they allow people within the company to manipulate data as if they knew everything that the data team knows. That's the big advantage. The data team says, "Here is how the data is structured." People can cut the data the way they wish.
Matt Pasienski: Instead of needing a data scientist to write SQL or do queries or anything like that, you are just going in and directly manipulating this. One last topic. I know you have a tremendous ... I'm always listening to you in your talks and things. You have a tremendous breadth of understanding of how SaaS is changing, the conferences that you participate in, you have your office hours that you run, how do you see SaaS starting to affect bigger, older businesses? It's very easy when you start a company to go in and sell your SaaS product to other tech companies because they have fast adoption cycles, but when you start getting to the point where they are looking to go in much bigger companies, the Fortune 100 companies of the world, how is that kind of transparency that SaaS enables going to transform those companies over the next five or six years?
Tomasz Tunguz: I think the big change is happening is everyone is becoming a software company. Automobile companies need to build software in order to build autonomous vehicles. It's happening much faster than I think any of them would have anticipated.
Matt Pasienski: Or liked.
Tomasz Tunguz: Consumer package goods companies are now starting to use data to try to figure out ... They've obviously used data for a long, long time, but they will continue to use data. In space. Software is starting to impact lots and lots of different industries and classic industries. What those businesses really need to do is they are looking to startups and try to figure out what disciplines and how these startups actually do this. There will be a cultural change, maybe it's a necessary cultural change for all those businesses to survive. I think in large part, they are going to do it with software as a service. Just to give a sense of where we are in the evolution, we ran a calculation that 2% of all IT is in this cloud.
Matt Pasienski: That's going to be -
Tomasz Tunguz: That's 2% of the market caps. If you think about it, even if you assume that two thirds of the market cap is serving regulated industries like government, finance, which are going to take forever, you're still have more than a 10x growth in the amount of value that will be created in SaaS.
Matt Pasienski: We're just starting.
Tomasz Tunguz: Just starting. Yeah. It's really, really, really [nice 00:08:34]. It's a very exciting time.
Matt Pasienski: Awesome. I think that is a good place to end it. Book coming out. What was the title again?
Tomasz Tunguz: Winning with Data.
Matt Pasienski: Winning with Data. I love that. Thanks a lot.