Share
Embed
Building a Tech Startup Outside the Silicon Valley
Fabian Lange
September 11, 2018
Transcript
75 slides
-
Preface • This talk is about our journey • Your
experience will be different • Of course its heavily survivor biased -
Agenda for today • Chronology of Instana • How we
evolved our engineering organisation to over 50 engineers • Ask me anything -
What problem did we want to solve? • Management of
IT systems is messy and no fun • In many companies several monitoring silos exist in operations • End User Monitoring • Infrastructure Monitoring • Cloud Monitoring • Application Monitoring • … -
That is not a new problem! • For a startup
its great if a problem is recognized already • Companies have already allocated budgets to solve this problem • Solving this problem is essential for companies that need IT to do business -
There must be competition!? • The monitoring market has many
solutions • We needed to understand most of them to know their strengths and weaknesses • Lots of one-trick-pony type of solutions show complexity of the problem, but also need for end-to-end visibility • Technology advances, and solutions quickly become legacy • Many solutions do have quality issues on stability or usability • Lack of automation and assistance • High dissatisfaction about the inability to solve real problems -
Our window of opportunity • We focused on containers first
• We also utilize modern technology to process data better • We deliver end to end visibility • We provide high quality and future proof technology • We automate root cause analysis • We solve actually day to day use cases of our customers -
Founding Team • A well rounded team • Pete built
sales, marketing and channel at similar startup before • Pavlo renown big data and stream processing expert • Mirko serial entrepreneur, networker • Fabian domain expert and experienced engineer • Avoiding the typical anti-pattern of “I have a great idea, now I need a team to do it” -
Business Validation • What is the total addressable market (TAM)
• What could be our market share • Who is our customer? • Mega tech corps like Google? • Local mom & pop shops? • Who is relevant competition? • How do we want to position ourselves? • A new product (AI Ops) • A better product (next gen APM) -
The Market * of companies Edge - Web Scale Companies
Highly individual & complex architectures Head - Enterprise Customers „New architectures“ - System of innovation Head - Enterprise Customers „SOA architectures“ - System of Records Long Tail - Small Enterprises Packages Apps, Standardized less than 1.000* about 50.000* more than 1.000.000* -
The Market Vendors & Tools * of companies Perfect Market-Segments
to place INSTANA signal fx -
$2.6B market The Market Top 15 APM Vendors 0
100 200 300 Revenue 2014 Dynatrace CA Techno- logies IBM Dell Microsoft Riverbed Splunk HP New Relic Oracle App Dynamics BMC Software Fujitsu Hitachi Nec Data Source: Gartner, Inc., „Market Share Analysis: Application Performance Monitoring, 2014“ by Federico De Silva, May 27, 2015 16% Y-2-Y growth -
Time for APM 3.0! 1990 1995 2000 2005 2010 2015
Client/Server Linux Open Source Web 3-Tier Java EE SOA Web 2.0 OSS Middleware Cloud Big Data µService Containers Lambda Arch. Tivoli BMC CA Nagios CA/Wily Dell/Quest Precise HP/Mercury dynatrace AppDynamics New Relic APM 1.0 APM 2.0 APM 3.0 -
Tech Validation • Prototypes for core ideas • Single agent,
auto update, easy install, high resolution • Container Ready • Stream processing • Algorithms and learning for data stream • Graph for understanding • “NoSQL” Storage • 3D UI and innovative visualisations -
Lots of talking • We talked to hundreds of potential
customers and partners • From previous life, network and cold calling • Practiced your pitching and marketing • Got feedback on ideas and prototypes • Narrowed down our target audience • Not trying to serve everybody • Narrowed down our solution • Ideally we would have done only 1 thing, but APM is a wide topic -
Let’s do this! • Market? Check! • Exists, big, underserved
• Ideas? Check! • Validated by externals • Team? Check! • We had everyone to do it ourselves • Tech? Check! • It actually works -
Founding the company • We named our company Instana, because
it is cool, but also has a meaning. And it was “free”, except on twitter :-( • Lots of legal stuff • We incorporated in the US, due to better VC market • For investments it is important to build a good CAP table for founders, VCs and employees • Filed for Patents, Trademarks • Secured initial financing, with a convertible note -
Product Market Fit • Selling the USPs and vision beyond
• Building a foundation, that is • good enough • operable • expandable • Hiring slowly for missing tech skills • Most of our first hires came from our network -
Sell something • We started with a reasonable price for
our product from day one • Find something people are willing to pay for • This is not about profit, or break even. It is about proof. • We build “infrastructure” only monitoring. Sold to “friendly” customers, but also cold called / cheap marketing • Deliver value (even when low initially), convince of future value and execute and deliver. Keep your promises. • Underpromise and overdeliver. -
Watching the feedback • Feedback is mostly positive • Feedback
will help you to polish, but not generate radical solutions • Remember Henry Ford… -
Pitching • Lots of VCs • Lots of Variants of
Pitches • Working Demos • Showing Proof: Customers • Showing Scale • Of tech • Of company • Spent weeks in silicon valley -
Weeks in valley • Talked to thought leaders • Lots
of great networking potential in the bay • Everybody is open for discussions • Do not pitch your idea, but rather show results and ask for feedback based on their expertise and experience • Still tough to get into VC Money. Little trust that a team outside the valley can have success... -
Weeks in Europe • Looked for a tech VC •
Still pitched the idea of getting an US VC down the line • And we found our investor -
Our Series A Lead Investor: Target Partners • Target Partners
focuses mostly on Early Stage “Hard Tech” startups in Germany. • Our partner Berthold has been amazing throughout the process -
Building the 1.0 Version • It was clear to us
that our vision is way too big. The domain requires this and we knew it would be a long run • We build small features providing an outline of what the later product will look like • Many small little MVPs that form a big picture • Make components scale and redundant (and document trade-offs) -
Building the 1.0 Team • We decided to grow the
team very slowly • Focused on experts working “alone” • Since experts could be anywhere, we went global • Started to hire Sales, Pre Sales, Marketing • Always keeping cost in mind -
How we found our B round investor • 2017 something
interesting happened. • We were growing the product and customer base and making more noise marketing wise • We were starting to think about the B financing • Then suddenly VCs approached us… • … and flew in people… • …who refused to leave until we sign -
Our Series B Lead Investor: Accel • Accel focuses on
SaaS solutions for customers and enterprise • Harry and Adrian flew in the next day • Laura and Andrei came as well, and refused to leave -
Accelerating Growth • Along our customer base we also started
to grow the company faster • Grow all departments • Cost control and planning becomes a challenge • Turn external services into internal functions Accounting, HR • Overcoming organisational challenges -
Our Series C Lead Investor: Meritech • If you are
watching my JavaZone vimeo, this is the thing I could not announce on stage :) vimeo.com/289698520 • Meritech is a late stage growth fund • Happy to have Alex on board now -
Local vs Remote • Instana is a remote by default
company • However, in engineering we prefer to have teams co-located • We have 4 main engineering offices • Solingen, Germany • München, Germany • Novi Sad, Serbia • Austin, Texas, USA -
Local vs Remote • As you could see, diversity is
an issue • Remote hiring tends to favour extroverts, meritocrats • We cannot take care of Juniors remotely • We build (currently) 4 local teams, where diversity is easier to achieve, and add a few remote people to the mix -
Engineers need to resolve Customer Issues • Customer Success team
deals with questions and comments • Engineering needs to understand and fix issues • We have currently 2 Support Engineer of the Week (SoW) • Responsibilities • Triage • Communicate (via Zendesk) • Help (Document or Implement) -
Major Releases vs Hotfixes • Releases contain new functionality and
major changes • Hotfixes contain bug fixes, improvements, small additions • We need to communicate around releases • Marketing • Customer Success • Sales • SaaS is “hot-fixed” almost daily • On Premises Releases trail SaaS releases -
Teams organise themselves • Teams have a purpose / goal
• Teams are staffed with cross- functional skills • Teams organise their working mode on their own -
Github • Following the Gitflow model https://nvie.com/posts/a-successful-git-branching-model/ • Using PRs
to discuss changes • Also covers our documentation • And customer sample apps • Like Robot Shop -
Slack • You want asynchronous communication that mimics office chatter
• Slack does this for us • Use private channels for “blameless” discussions • Star and mute channels! -
Zoom • High Quality Video Conferencing and Screensharing • Zoom
has great apps and even “Meeting Room” support -
Realtime Board • Teams do various meetings to discuss plans
and designs • Visualisation of ideas and interaction are hard remote • Realtime Board fixes this -
Trello • A lot of planning does not require
a high level of detail • Trello is good for getting a high level overview of progress • We use it for roadmapping and other organisational things -
Pivotal Tracker • Lower level issue and task tracking •
Iteration planing aid • Progress tracking • Chaotic at times -
Wiki / Docs • Boards and Trackers are often insufficient
for longer design and requirements work • We use Notion.so as internal documentation & knowledge base and for shorter lived working documents. • For drafting things, we also use Google Docs / Sheets