Splunk is now a long distance runner, according to CEO Godfrey Sullivan, who explains that four years ago, prior to the firm’s IPO, a poster went up in all the company’s offices with a simple message: Mile Three:
That message to our employees was that our IPO was not a destination, but just the mile three marker on the long run to build a lasting company and brand. The day after the IPO we took down all the mile three posters and put up our mile six poster with our corporate initiatives.
And just last month, three years later, we tore down the mile six posters and replace them with mile eight. The mile eight poster has four words on it enterprise, partners, solutions and cloud. These words represent the core initiatives that we are pursuing on behalf of our customers.
These four words say a lot about our next phase of growth and true to Splunk’s open culture, we publish them for everyone to see. Just like mile three and mile six, mile eight is a multiyear run ahead of us.
The emphasis on solutions is significant. Sullivan says:
Solutions, means apps, content and driving insights and business outcomes for our customers.
Splunk is not just about indexing and searching machine data. We are becoming a solutions company. A solutions company that supported by an enterprise grade platform with apps that we build and our partners build. Cloud, this may seem obvious, but here it means cloud first that is designing everything we do to be easy, to be offered as a service and to provide even shorter deployment cycles for our customers.
You'd be surprised how many customers start a conversation with us saying, ‘I have this business requirement’. They don't say we should come in and analyze my machine data for them, ‘Can you ingest our logs?’. They actually have a business problem they are trying to solve and that’s most often represented by an app or solution of some kind.
It’s important for us to become a solutions company and as we do so, those solutions that are not what we build ourselves are a perfect opportunity to go build a partner ecosystem.
There’s much for Sullivan to be pleased about. Recently released first quarter Q116 numbers saw a 46% year-on-year jump in revenues to $125.7 million. Other stats that cheer:
- 450 new customers in the past three months.
- Total worldwide customer base of 9,500.
- 226 orders greater than a $100,000 during the past quarter.
- The first 7-figure enterprise solution order to date, with “a customer who was moving 90% of all data to AWS”.
Enterprise customers include some big name brands. For example, Sony Playstation Network signed up for Splunk Enterprise, while Adobe, one of Splunk’s longer-standing customers, also signed for one of the firm’s new enterprise agreements.
On the cloud front, AOL will use a large terabyte Splunk cloud instance for real time performance monitoring, log analysis and dev ops. Meanwhile the city of Los Angeles purchased Splunk cloud and the app for enterprise security to correlate cyber threat information with other governments and monitor and analyze network traffic to identify discrepancies that indicate malicious attacks.
There’s a pattern here, with the majority of Splunks business coming from existing customers upgrading or upscaling. Sullivan explains:
Typically a customer goes from departmental over a period of years, goes from departmental to multi-departmental and at that point in time is when they're more interested in standardizing on a large-scale license.
If you look at enterprise software over the many, many years I've been in this business, there have been some fairly steady percentages between the percentage of customers who want to buy commercial off-the-shelf software, because they don't want to build anything. That’s about 60% of total spend and about 40 years of customers who were using a whole variety of build technologies to actually build custom enterprise applications.
So you have that kind of look at the market through those two lenses and say, ‘Which customer type is it?’ If it’s a customer who wants to buy software and not build it, the open-source thing never comes up.
Then are customers who want to build their own, he suggests, such as banks:
Building big custom enterprise apps, often large development staffs and a high propensity to build things, they will always go look at the next open-source technology to say that’s something that provides value. But when you get there Hadoop and ELK are two very different animals. So Hadoop is really good at cheap batch storage, ELK really good at solving very narrow questions, but you have to apply about five times the amount of hardware to get it to work.
That means higher bills, he warns:
It’s really expensive. I mean I know everybody says it's free, but it’s really expensive to build on ELK, both from our hardware standpoint, storage and cost of ownership and maintenance and it’s not a very flexible system.
We have a lot of customers who wander often try to build that stuff and they come back bruised and bloody and say ‘Boy that was hard and we didn’t get very far and every time we want to make a change we had to go back and then open up the system’.
So there are customers who want to buy software because they don’t want to build stuff and we never see them and it’s only really in the accounts that have a higher propensity to build where the stuff comes up and they usually wander off in the woods for a while and they come back.
An increasingly longer term vision for what is still a very young company. The next phase will go far in determining just what kind of company Splunk ends up being.