Earlier this week I spent a couple of days at Cloudera’s 3rd annual Analyst Day. The team put together a strong program and it was great to get such a condensed update on the company’s progress and strategy.
Strategy
Speaking of strategy, we spent a lot of time on company strategy at the event and covered several different views into their strategy, i.e. company, product, go to market, partner, etc. perspectives. A couple of highlights here:
Open Source. The company emphasized its commitment to open source through its hybrid (ie open-core) business model. This has been a point that the competition, namely Hortonworks, has hit them on, but they seemed to have learned how to tell this story and back it up with evidence of community contributions and commitment.
Hybrid open source is powerful way to sell sw. 'You already use it, we can help secure and manage it.' #ClouderaAD pic.twitter.com/9tKGRSokU1
— Sam Charrington (@samcharrington) March 21, 2016
Business Value. Cloudera believes that Hadoop is now mainstream, meaning their target market of Global 8,000 enterprises are moving beyond Hadoop science experiments and are ready to have business-level discussions about the transformative opportunity around big data. They’re focusing their efforts on three “board-level” value drivers: customer 360, product/service delivery, and risk.
#Hadoop in 2014: Talking zoo animals. 2015: Elevated discussion to data lake. 2016: Business value. #ClouderaAD pic.twitter.com/pBeoYYq4Do
— Sam Charrington (@samcharrington) March 21, 2016
Land & Expand. The companies go-to-market approach is based on getting a foot in the door with smaller opportunities (including the aforementioned science experiments, typically under USD100K), and expanding from there. They shared a fair amount of detail under NDA suggesting that this strategy is working very well.
More on the @cloudera land & expand model. #ClouderaAD pic.twitter.com/kNkWplG7gD
— Sam Charrington (@samcharrington) March 21, 2016
Product Packaging. Consistent with their belief in a mainstream Hadoop market, the company has plans to shift the way it talks about and sells its product. Unfortunately specifics can’t be shared here but will be made public soon.
Intel Partnership
It’s been almost two-years to the day since Cloudera announced Intel’s USD740M investment in the company. The company presented a number of data points suggesting that they “partnership” aspect of the investment is more than empty talk. The company now has extensive visibility into what’s coming down the pike in hardware, and has oriented its R&D; roadmap around those advances that will provide impact. Examples include the work in hardware-based encryption acceleration (present), support for Intel’s 3D XPoint tech (future, see below), pushing machine learning functions from Spark into silicon (future). In addition, the two are collaborating on some interesting projects in the cybersecurity space which will be announced soon.
.@Intel partnership gives @Cloudera 5 yr view into hardware pipeline and unfair advantage among peers #ClouderaAD pic.twitter.com/gdtq91tYZ6
— Sam Charrington (@samcharrington) March 21, 2016
Industry Insight
Also supporting Cloudera’s desire to elevate customer conversations is their investment in an internal team of vertical subject matter experts: Health and life sciences, telecom, retail, manufacturing, financial services and insurance were represented on a panel that was held; the discussion was generally very insightful.
.@Cloudera developing an industry vertical focus with consultative selling & engagement. #ClouderaAD pic.twitter.com/SzH4VlFUpH
— Doug Henschen (@DHenschen) March 21, 2016
Cloud
The company launched Cloudera Director a couple of years back (my coverage here), and while it was a positive step at the time, the product had seemed to languish a bit… until now. Cloudera is doubling down on the cloud. In the past 18 months, the company says, they”’ve seen growing enterprise acceptance of the public cloud, to the point that 15% of deployments are now cloud based.
Cloudera turning focus to the cloud: Drivers: 1. Data gravity 2. Flexible resourcing 3. Enterprise acceptance w/in past 18 mo click reference. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
With Cloudera Director as a starting place, which it will reinvigorate its investment in, the company is stepping back and tacking stock of the entire platform and how it needs to evolve to fully take advantage of cloud deployment.
Architecting #hadoop for on prem is different than architecting for the #cloud. #ClouderaAD pic.twitter.com/HYqyCMIqGh
— Sam Charrington (@samcharrington) March 21, 2016
Cybersecurity
The company sees a large opportunity in cybersecurity and has been, and will be investing in that area. Cyber threats, data loss, etc. represent such a huge risk to C-suite executives, helping them combat these risks has proven to make the checkbooks fly open (in the words of a partner).
#Cybersecurity a 2nd big investment area for @Cloudera. #Hadoop offers broader solution. @mikeolson #ClouderaAD pic.twitter.com/T9HiykgNlv
— Sam Charrington (@samcharrington) March 21, 2016
Notably, Cloudera is not doing the domain-specific heavy lifting by themselves—Intel, Accenture, and other partners are playing a large role in this.
Data Science
The company hopes to make it easier for data scientists (i.e. non-Java developers using languages like Python and R) to use the platform. Watch this space for more!
Update 24 Mar 2016: That was quick! Sense announces their acquisition by Cloudera.
Summary
Cloudera is very well positioned moving forward. They’re seeing strong results/metrics in the core business, are funded through to cash-flow breakeven, and have solid/productive partnerships in place. Even if they’re a bit ahead of the market in declaring Hadoop “mainstream,” what they’ve done with that–build out the industry team, tighten up the product definition, start to shore up the cloud story, and launch forays into a first application area–all bode well for the company.
If anything, the risk they face is having too many strategies. I suggested this to one company executive and he volunteered that the company does seem to have a “three-dimensional” strategy at times. The risk here is complexity, confusion in the ranks–especially the field–and faltering execution.
If they can simplify the strategy a bit and drive coherent execution I’m sure we’ll be seeing even better metrics this time next year.
More Via Twitter
Strategy & Business Overview
The only thing that could disrupt Hadoop is another open source project. That's why they embrace ecosystem. Tom Reilly #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera now >850 enterprise subscription customers (>60% growth y/y). Strong growth in vertically-focused partners. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Many @Cloudera customers living in #hybrid world. Investing in #cloud is a high priority. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
40% performance tax to encrypt in software, 3% to do in hardware. Intel investment helped here. #EncryptAllTheData #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
In next few years, pay-when-you-drive to be part of all insurance policies. Pay how you drive not far behind. Tom Reilly #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera CFO Jim Frankola reviewing the business. Unfortunately no pics or tweets. Or financials for that matter. #ClouderaAD #NDA
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera biz model based on Land & Expand strategy. Expansion occurring particularly quickly w/in target accounts. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera: Roughly 15% of use cases #cloud based. More on this to come. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Keys to Cloudera's success/growth in China: Inherited @Intel distribution's customer base and Shanghai-based team. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Top 3 concerns for @Cloudera CFO #1: Maintaining culture as the company grows, opens new offices. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Top 3 concerns for @Cloudera CFO #2: Managing transition fr land to expand over time eg sales/tech/partner ratio, channel comp #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Top 3 concerns for @cloudera CFO #3: Managing transition to #Cloud computing. We'll hear more on this later. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Now up: @Cloudera Strategy Update with @mikeolson. 90% of action now not MapReduce/HDFS #ClouderaAD pic.twitter.com/vwNNw40idT
— Sam Charrington (@samcharrington) March 21, 2016
Strategy sum-up from @mikeolson: If we can make #hadoop platform usable and consumable, it will become invisible & ubiquitous #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera following Spark since '10. Creator interned there while deving. No one sells more $ or or has more customers @mikeolson #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera product strategy – hasn't changed over past year. #ClouderaAD pic.twitter.com/CKjiqY7QcM
— Sam Charrington (@samcharrington) March 21, 2016
Customers don't have different standards (security, reliability, stability) for #opensource software. @kestelyn #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera tracks "boiling cauldron" of ~100 open source projs. About 1 in 4 make it into #CDH. #ClouderaAD pic.twitter.com/pzdhzukRAV
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera taking math functions from Spark used in #machinelearning and pushing them into silicon to drive performance gains #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Cloudera CTO @awadallah on 3 key trends @Cloudera investing in (roughly 70% of engineering resources) #ClouderaAD pic.twitter.com/G3REJGlJvb
— Sam Charrington (@samcharrington) March 21, 2016
Cloud increasingly important for @cloudera. Big message at #ClouderaAD pic.twitter.com/kNSbS90KOu
— Sam Charrington (@samcharrington) March 21, 2016
.@cloudera expects @intel's #3DXPoint to be inflection point in memory. Servers will have TBs memory #ClouderaAD pic.twitter.com/x3zh9Xarfc
— Sam Charrington (@samcharrington) March 21, 2016
EMC's #DSSD also a fundamental innovation that will impact hardware arch for #Hadoop @awadallah #ClouderaAD pic.twitter.com/ANgguJ0DnG
— Sam Charrington (@samcharrington) March 21, 2016
Executive panel getting kicked off at #ClouderaAD. What questions do you have? pic.twitter.com/xyNrDXKOKm
— Sam Charrington (@samcharrington) March 22, 2016
Ppl thought Spark was end of @Cloudera, now they're largest distrib. Question is what's next? What will Spark Spark? @mikeolson #ClouderaAD
— Sam Charrington (@samcharrington) March 22, 2016
Spark, like MapReduce before it, was a wonderful thing, but not everything. @mikeolson #ClouderaAD
— Sam Charrington (@samcharrington) March 22, 2016
Cloudera spending 30% of presales time in architecture discussions. Customers need help making sense of options. #ClouderaAD
— Sam Charrington (@samcharrington) March 22, 2016
As # of components and complexity of interactions has grown, Cloudera has become much more conservative in what is labeled GA. #ClouderaAD
— Sam Charrington (@samcharrington) March 22, 2016
Cloudera Labs is their bazaar next to the cathedral. If you want a view into what projects they think are coming, check there. #ClouderaAD
— Sam Charrington (@samcharrington) March 22, 2016
"Cloudera's partner momentum a leading indicator of where market is going" Tim Stevens #ClouderaAD pic.twitter.com/tKiuBlFSAm
— Sam Charrington (@samcharrington) March 21, 2016
#IoT a good use case for @Cloudera's #Kudu project. Hi rate ingest plus analytics on that data required. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Industry Perspectives Panel
"Telcos traditionally have been data rich but insight poor." Big opportunity for Hadoop. Apps: risk, churn, custsrv Vijay Raja #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Usage based insurance has been discussed for decade. What's diff today is ability to blend telematics w/ backend data #hadoop #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
"The customer is no longer king, they are omnipotent gods" @cloudera customer Marks & Spencer re Customer 360 in retail #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Customer 360 in 2016: Use the customers behavior in the moment to drive engagement in the same moment. #hadoop enbles. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
2/2 But to your pt @holgermu if they're smart they'll focus innovation horizontally & leave vertical innov to partners. #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Customer Panel
"A staggering amount of US food supply chain runs on FoxPro DBs. #Hadoop a flexible way to integrate their data." @OneLineage #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Getting a refresher in combinatorics at #ClouderaAD. Point is, some types of analytics are hard to push to end users via #selfservice.
— Sam Charrington (@samcharrington) March 21, 2016
"Best part of working with Cloudera services was that the engagement ended" Customer fr use case panel #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Data ingest the next big challenge to take on to enable enterprise data hub @mikeolson #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Thomson Reuters migrating a bunch of #machinelearning to #Spark. Before Spark had no way to scale it on #Hadoop. #ClouderaAD #ML
— Sam Charrington (@samcharrington) March 21, 2016
"Average time to understand an attack is 200 days. Getting down to a few days is huge." @accenture #cybersecurity #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
"All of @Accenture's clients know they've been hacked. Just don't know how or how bad." #ClouderaAD #cybersecurity https://t.co/MEuWoMMs4Z
— Sam Charrington (@samcharrington) March 21, 2016
Argyle Data Use Case
Mobile fraud & cybersec: "Truth is in the packets. If u can look at firehose u can see everything." #ClouderaAD pic.twitter.com/gkG8aRmNjW
— Sam Charrington (@samcharrington) March 22, 2016
If you're looking for fraud/attack patterns, need at least 13 months of data. Old world only stores 30 days. #ClouderaAD
— Sam Charrington (@samcharrington) March 22, 2016
Community
What's next for #Hadoop? w @cutting at #ClouderaAD pic.twitter.com/3FtDgtGA4d
— Sam Charrington (@samcharrington) March 21, 2016
Limited edition replica of @cutting's son's original #hadoop doll courtesy @cloudera #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Doug Cutting. @Cloudera not religious about: Tech stack, Open Core biz model. Is religious about open source, customer value #ClouderaAD
— Sam Charrington (@samcharrington) March 21, 2016
Recent #opensource milestones for @cloudera #ClouderaAD pic.twitter.com/3S3BCqZZde
— Sam Charrington (@samcharrington) March 21, 2016
Leave a Reply