Sep 30, 2016 · 29m · mad
Big Data at Slack // Noah Weiss, Slack (Data Driven NYC / FirstMark)
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
At a Data Driven NYC event hosted by FirstMark's Matt Turck, Noah Weiss, VP of Product at Slack, discusses Slack's rapid growth, the consumerization of enterprise software, and how his Search, Learning, and Intelligence team is leveraging data, search algorithms, and machine learning to make workplace communication smarter.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 13.5% of the talking time here. How this is scored →
Matt as informed peer 3.0 Guest teaching 3.3 Guest disagreement 0.1 Matt pushing back 1.0
speaking balance: gold is Matt, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 3:38 Correcting Company History Sequence
In an overall highly genial conversation, Noah gently corrects Matt when Matt confuses Tiny Speck as the first company rather than the second venture after Ludicorp.
Hardest push from Matt ▶ 14:46 Pressing for Product Launch TimelineMatt refuses to accept a high-level product description without concrete milestones, explicitly pressing Noah on whether launch is months or years away.
Biggest teaching moment ▶ 17:20 Explaining Enterprise Search AmbiguityNoah thoroughly educates the room on why enterprise search is fundamentally different from Google web search due to contextual query ambiguity.
Matt holds his own ▶ 11:20 Citing Historical Knowledge Management VendorsMatt demonstrates deep enterprise IT domain knowledge by naming specific pioneer vendors from the early knowledge management and search industry.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Matt as informed peer | Guest teaching | Guest disagreement | Matt pushing back | Why |
|---|---|---|---|---|---|---|
| Noah Weiss's Background and Career Journey | 2 | 4 | 0 | 0 | Matt prompts Noah for his background and briefly clarifies his role, while Noah shares his extensive career path across Google Knowledge Graph and Foursquare. | |
| The Evolution, History, and Growth of Slack | 4 | 3 | 1 | 1 | Matt draws parallels between Slack's origin and Twitter's pivot from Odeo, showing background knowledge, while Noah corrects Matt on the order of Tiny Speck and Ludicorp. | |
| Defining the Search, Learning, and Intelligence Group | 3 | 3 | 0 | 1 | Matt asks why Slack succeeded where older tools failed, and Noah explains the market macro-trends and the internal backronym history of Slack. | |
| Reimagining Enterprise Search at Slack | 6 | 3 | 0 | 2 | Matt demonstrates clear domain authority by listing historical enterprise search vendors like Verity, Autonomy, Fast, and Endeca, contextualizing Noah's explanations. | |
| Mitigating Information Overload with Smart Prioritization | 3 | 2 | 0 | 3 | Matt pushes Noah for concrete release dates on features, asking if the rollout is three months or two years out. | |
| Understanding Slack's Unique Data Dynamics and Personalization | 2 | 5 | 0 | 0 | Noah educates the audience on data dynamics in enterprise search, contrasting clear consumer search intent with ambiguous enterprise queries. | |
| The Slack Platform, App Ecosystem, and Bots | 2 | 3 | 0 | 1 | Matt prompts Noah for specific examples of bot applications, leading Noah to explain developer motivations versus end-user needs. | |
| The Consumerization of Enterprise Software | 3 | 4 | 0 | 0 | Matt frames Slack as the prime example of enterprise software consumerization, and Noah details the bottom-up viral adoption pattern inside companies. | |
| Establishing and Growing Slack's New York Presence | 3 | 3 | 0 | 2 | Matt questions whether opening an NYC office was strategic or purely circumstantial around Noah, prompting Noah to analyze tech talent distribution across regions. | |
| Building the NYC Search, Learning, and Intelligence Team | 2 | 3 | 0 | 0 | Matt asks about team composition before opening questions to the audience regarding security and authentication layers. |
Statements from this episode (17)
Weiss: Continuously relearning web data for search is brittle
“Relearning all this information from the web over and over again actually turned out to be quite brittle, and a lot of this information is actually quite evergreen.”
Noah Weiss: Slack has passed 3 million daily active users
“Our public stats are that we have passed over three million daily active users, so people are using the product literally every day.”
Weiss: Slack users spend over two hours a day active foregrounded
“Those daily active users are using it for about two hours, over two hours a day, like, actually foregrounded on their phone or on their desktop. Ah, they send over 60 messages a day.”
Weiss: Many companies using Slack send millions of messages per day
“When you look at companies that are using it many are sending even millions of messages a day now,”
Weiss: Slack stands for Searchable Log of All Conversation and Knowledge
“So the backronym that they came up with was that Slack stood for a searchable log of all conversation and knowledge.”
Weiss: Surfacing individual experts satisfies 80% of workplace search needs
“Unlike Google, if you give someone as a result, an individual person often that can be 80% of what someone's actually looking for.”
Weiss: Integration of third-party workflows drives long-term Slack retention
“People come to Slack initially because it's a great tool for collaborating with your team. They usually stick around because it's a way of integrating all these different workflows that you have and the dozens of third party services that you use into one plac…”
Weiss: Slack median session length on mobile is 30 seconds
“Especially on mobile where, like, the median session length is 30 seconds.”
Weiss: Slack search lacks a head query distribution across teams
“When you think about it broken out by team there's no head, right? Like, if you append kind of team equals Slack or team equals first mark, you know, there is no head of that query stream across all teams. Whereas Google, on the other hand, you know, they have…”
Noah Weiss: Almost every search query on Slack is inherently ambiguous
“On Slack, we kind of view it as almost every query is inherently ambiguous because the lens that you're looking through is quite different.”
Weiss: About 600 apps have been published in Slack's App Directory
“We launched in December kind of our version of an app store, and since then about 600 apps have been published, largely all kind of bots of some sort.”
Weiss: Most early Slack bots are still in an 'interesting toy phase'
“Quite frankly, I think a lot of them are still in more of the, like, interesting kind of toy phase in terms of the sophistication. They're not kind of fully robust tools.”
Weiss: Developer hype for bots stems from app store distribution frustration
“I think a lot of the excitement about, ah, bots in general, not just on Slack, on Messenger, on all these other services, is developers are really fed up with trying to actually get adoption and distribution for the service that they're building.”
Weiss: Companies love Slack most when piping internal backend tools into it
“I think the companies that wind up loving Slack the most, they basically take all these really ugly internal tools that they have that no one ever wants to log into anyway, And they start just kind of connecting those pipes into Slack and that's something that…”
Weiss: Roughly 90% of Slack employees previously built consumer software
“Most people at Slack have never worked at enterprise software before. If I had to guess, I'd say 90% of people had only worked on consumer software like Flickr, or Facebook, or Foursquare, Twitter, Google, whatnot.”
Weiss: Most Slack deployments start bottom-up via personal credit cards
“Most of Slack's deployments don't start with a CIO doing a security review and saying, I approve Slack for deployment. It normally is the opposite, which is a bunch of groups are using it. They're using their personal credit cards.”